1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
9 PP/02 Make -n do something, by making it not do something.
10 When combined with -bP, the name of an option is not output.
12 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
15 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
16 $sender_host_name and config options to manage this, and basic check
19 PP/05 DSCP support for outbound connections and control modifier for inbound.
21 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
22 (Only plugin which currently uses this is kerberos4, which nobody should
23 be using, but we should make it available and other future plugins might
24 conceivably use it, even though it would break NAT; stuff *should* be
25 using channel bindings instead).
27 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
28 name; added for Sendmail compatibility; requires admin caller.
29 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
30 just ignore it); requires trusted caller.
31 Also parse but ignore: -Ac -Am -X<logfile>
34 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
36 TL/02 Add +smtp_confirmation as a default logging option.
38 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
39 Patch by Magnus Holmgren from 2007-02-20.
41 TL/04 Bugzilla 1281 - Spec typo.
42 Bugzilla 1283 - Spec typo.
43 Bugzilla 1290 - Spec grammar fixes.
45 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
47 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
49 JH/02 Support "G" suffix to numbers in ${if comparisons.
51 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
53 NM/01 Bugzilla 1197 - Spec typo
54 Bugzilla 1196 - Spec examples corrections
56 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
58 PP/09 Add gnutls_enable_pkcs11 option.
60 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
61 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
63 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
64 "acl {{name}{arg}...}", and optional args on acl condition
67 JH/05 Permit multiple router/transport headers_add/remove lines.
69 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
71 JH/07 Avoid using a waiting database for a single-message-only transport.
72 Performance patch from Paul Fisher. Bugzilla 1262.
74 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
77 JH/09 Add $headers_added variable, with content from use of ACL modifier
78 add_header (but not yet added to the message). Bugzilla 199.
80 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
81 Pulled from Bugzilla 817 by Wolfgang Breyha.
83 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
86 JH/11 Add A= logging on delivery lines, and a client_set_id option on
89 JH/12 Add optional authenticated_sender logging to A= and a log_selector
96 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
98 This, or similar/improved, will also be change PP/11 of 4.82.
104 PP/01 Handle short writes when writing local log-files.
105 In practice, only affects FreeBSD (8 onwards).
106 Bugzilla 1053, with thanks to Dmitry Isaikin.
108 NM/01 Bugzilla 949 - Documentation tweak
110 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
113 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
115 PP/02 Implemented gsasl authenticator.
117 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
119 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
120 `pkg-config foo` for cflags/libs.
122 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
123 with rest of GSASL and with heimdal_gssapi.
125 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
126 `pkg-config foo` for cflags/libs for the TLS implementation.
128 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
129 properties get this fed in as external SSF. A number of robustness
130 and debugging improvements to the cyrus_sasl authenticator.
132 PP/08 cyrus_sasl server now expands the server_realm option.
134 PP/09 Bugzilla 1214 - Log authentication information in reject log.
135 Patch by Jeremy Harris.
137 PP/10 Added dbmjz lookup type.
139 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
141 PP/12 MAIL args handles TAB as well as SP, for better interop with
142 non-compliant senders.
143 Analysis and variant patch by Todd Lyons.
145 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
146 Bug report from Lars Müller <lars@samba.org> (via SUSE),
147 Patch from Dirk Mueller <dmueller@suse.com>
149 PP/13 tls_peerdn now print-escaped for spool files.
150 Observed some $tls_peerdn in wild which contained \n, which resulted
151 in spool file corruption.
153 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
154 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
155 or write after TLS renegotiation, which otherwise led to messages
158 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
159 as a tracking header (ie: a signed header comes before the signature).
160 Patch from Wolfgang Breyha.
162 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
163 comma-sep list; embedded commas doubled.
165 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
167 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
169 Report and patch from Dmitry Banschikov.
171 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
172 Removed SSL_clear() after SSL_new() which led to protocol negotiation
173 failures. We appear to now support TLS1.1+ with Exim.
175 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
176 lets Exim select keys and certificates based upon TLS SNI from client.
177 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
178 before an outbound SMTP session. New log_selector, +tls_sni.
180 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
181 NULL dereference. Report and patch from Alun Jones.
183 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
184 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
185 Not seeing resolver debug output on NetBSD, but suspect this is a
186 resolver implementation change.
188 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
189 Left warnings. Added "eximon gdb" invocation mode.
191 PP/21 Defaulting "accept_8bitmime" to true, not false.
193 PP/22 Added -bw for inetd wait mode support.
195 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
196 locate the relevant includes and libraries. Made this the default.
198 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
199 Bugzilla 1246, report and most of solution from Tomasz Kusy.
201 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
202 This may cause build issues on older platforms.
204 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
205 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
206 gnutls_require_mac & gnutls_require_protocols (no longer supported).
207 Added SNI support via GnuTLS too.
208 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
210 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
212 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
213 Applied second patch from Janne, implementing suggestion to default
214 multiple-strings-in-record handling to match SPF spec.
216 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
218 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
219 read-only, out of scope).
220 Patch from Wolfgang Breyha, report from Stuart Northfield.
222 PP/29 Fix three issues highlighted by clang analyser static analysis.
223 Only crash-plausible issue would require the Cambridge-specific
224 iplookup router and a misconfiguration.
225 Report from Marcin Mirosław.
227 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
229 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
230 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
231 As part of this, removing so much warning spew let me fix some minor
232 real issues in debug logging.
234 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
235 assignment on my part. Fixed.
237 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
238 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
239 Janne Snabb (who went above and beyond: thank you).
241 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
242 string otherwise requires a connection and a bunch more work and it's
243 relatively easy to get wrong. Should also expose TLS library linkage
246 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
247 64-bit ${eval} (JH/03).
249 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
250 GNU libc to support some of the 64-bit stuff, should not lead to
251 conflicts. Defined before os.h is pulled in, so if a given platform
252 needs to override this, it can.
254 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
255 protection layer was required, which is not implemented.
256 Bugzilla 1254, patch from Wolfgang Breyha.
258 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
259 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
260 tls_dhparam take prime identifiers. Also unbreak combination of
261 OpenSSL+DH_params+TLSSNI.
263 PP/39 Disable SSLv2 by default in OpenSSL support.
269 PP/01 Solaris build fix for Oracle's LDAP libraries.
270 Bugzilla 1109, patch from Stephen Usher.
272 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
274 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
277 TF/02 Fix a couple more cases where we did not log the error message
278 when unlink() failed. See also change 4.74-TF/03.
280 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
281 lock up or crash if it happened to be inside a call to libc when it
282 got a SIGUSR1 from exiwhat.
284 The SIGUSR1 handler appends the current process status to the process
285 log which is later printed by exiwhat. It used to use the general
286 purpose logging code to do this, but several functions it calls are
287 not safe for signals.
289 The new output code in the SIGUSR1 handler is specific to the process
290 log, and simple enough that it's easy to inspect for signal safety.
291 Removing some special cases also simplifies the general logging code.
292 Removing the spurious timestamps from the process log simplifies
295 TF/04 Improved ratelimit ACL condition.
297 The /noupdate option has been deprecated in favour of /readonly which
298 has clearer semantics. The /leaky, /strict, and /readonly update modes
299 are mutually exclusive. The update mode is no longer included in the
300 database key; it just determines when the database is updated. (This
301 means that when you upgrde Exim will forget old rate measurements.)
303 Exim now checks that the per_* options are used with an update mode that
304 makes sense for the current ACL. For example, when Exim is processing a
305 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
306 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
307 must specify per_mail/readonly. If you omit the update mode it defaults to
308 /leaky where that makes sense (as before) or /readonly where required.
310 The /noupdate option is now undocumented but still supported for
311 backwards compatibility. It is equivalent to /readonly except that in
312 ACLs where /readonly is required you may specify /leaky/noupdate or
313 /strict/noupdate which are treated the same as /readonly.
315 A useful new feature is the /count= option. This is a generalization
316 of the per_byte option, so that you can measure the throughput of other
317 aggregate values. For example, the per_byte option is now equivalent
318 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
320 The per_rcpt option has been generalized using the /count= mechanism
321 (though it's more complicated than the per_byte equivalence). When it is
322 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
323 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
324 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
325 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
326 Note that using per_rcpt with a non-readonly update mode in more than
327 one ACL will cause the recipients to be double-counted. (The per_mail
328 and per_byte options don't have this problem.)
330 The handling of very low rates has changed slightly. If the computed rate
331 is less than the event's count (usually one) then this event is the first
332 after a long gap. In this case the rate is set to the same as this event's
333 count, so that the first message of a spam run is counted properly.
335 The major new feature is a mechanism for counting the rate of unique
336 events. The new per_addr option counts the number of different
337 recipients that someone has sent messages to in the last time period. It
338 behaves like per_rcpt if all the recipient addresses are different, but
339 duplicate recipient addresses do not increase the measured rate. Like
340 the /count= option this is a general mechanism, so the per_addr option
341 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
342 example, measure the rate that a client uses different sender addresses
343 with the options per_mail/unique=$sender_address. There are further
344 details in the main documentation.
346 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
348 TF/06 Removed a few PCRE remnants.
350 TF/07 Automatically extract Exim's version number from tags in the git
351 repository when doing development or release builds.
353 PP/02 Raise smtp_cmd_buffer_size to 16kB.
354 Bugzilla 879. Patch from Paul Fisher.
356 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
357 Heavily based on revision 40f9a89a from Simon Arlott's tree.
360 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
362 PP/05 Variable $av_failed, true if the AV scanner deferred.
363 Bugzilla 1078. Patch from John Horne.
365 PP/06 Stop make process more reliably on build failure.
366 Bugzilla 1087. Patch from Heiko Schlittermann.
368 PP/07 Make maildir_use_size_file an _expandable_ boolean.
369 Bugzilla 1089. Patch from Heiko Schlittermann.
371 PP/08 Handle ${run} returning more data than OS pipe buffer size.
372 Bugzilla 1131. Patch from Holger Weiß.
374 PP/09 Handle IPv6 addresses with SPF.
375 Bugzilla 860. Patch from Wolfgang Breyha.
377 PP/10 GnuTLS: support TLS 1.2 & 1.1.
379 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
382 PP/11 match_* no longer expand right-hand-side by default.
383 New compile-time build option, EXPAND_LISTMATCH_RHS.
384 New expansion conditions, "inlist", "inlisti".
386 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
388 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
390 PP/14 fix log_write() format string regression from TF/03.
391 Bugzilla 1152. Patch from Dmitry Isaikin.
397 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
399 PP/02 Harmonised TLS library version reporting; only show if debugging.
400 Layout now matches that introduced for other libraries in 4.74 PP/03.
402 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
404 PP/04 New "dns_use_edns0" global option.
406 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
409 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
410 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
412 TK/01 Updated PolarSSL code to 0.14.2.
413 Bugzilla 1097. Patch from Andreas Metzler.
415 PP/07 Catch divide-by-zero in ${eval:...}.
418 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
421 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
422 format-string attack -- SECURITY: remote arbitrary code execution.
424 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
425 time unintentionally subject to list matching rules, letting the header
426 cause arbitrary Exim lookups (of items which can occur in lists, *not*
427 arbitrary string expansion). This allowed for information disclosure.
429 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
430 INT_MIN/-1 -- value coerced to INT_MAX.
436 NM/01 Workround for PCRE version dependency in version reporting
439 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
440 This fixes portability to compilers other than gcc, notably
441 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
443 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
444 makefiles for portability to HP-UX and POSIX correctness.
446 PP/01 Permit LOOKUP_foo enabling on the make command-line.
447 Also via indented variable definition in the Makefile.
448 (Debugging by Oliver Heesakkers).
450 PP/02 Restore caching of spamd results with expanded spamd_address.
451 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
453 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
454 Improves build reliability. Fix from: Frank Elsner
456 NM/02 Fix wide character breakage in the rfc2047 coding
457 Fixes bug 1064. Patch from Andrey N. Oktyabrski
459 NM/03 Allow underscore in dnslist lookups
460 Fixes bug 1026. Patch from Graeme Fowler
462 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
463 Code patches from Adam Ciarcinski of NetBSD.
465 NM/04 Fixed exiqgrep to cope with mailq missing size issue
468 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
469 is logged, to avoid truncation. Patch from John Horne.
471 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
472 Patch from Jakob Hirsch.
474 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
475 SQL string expansion failure details.
476 Patch from Andrey Oktyabrski.
478 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
479 Patch from Simon Arlott.
481 PP/09 New lookups functionality failed to compile on old gcc which rejects
482 extern declarations in function scope.
483 Patch from Oliver Fleischmann
485 PP/10 Use sig_atomic_t for flags set from signal handlers.
486 Check getgroups() return and improve debugging.
487 Fixed developed for diagnosis in bug 927 (which turned out to be
490 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
491 Patch from Mark Zealey.
493 PP/12 Bugzilla 1056: Improved spamd server selection.
494 Patch from Mark Zealey.
496 PP/13 Bugzilla 1086: Deal with maildir quota file races.
497 Based on patch from Heiko Schlittermann.
499 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
500 Patch from Uwe Doering, sign-off by Michael Haardt.
502 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
503 variable declaration deep within a block. Bug and patch from
506 PP/15 lookups-Makefile IRIX compatibilty coercion.
508 PP/16 Make DISABLE_DKIM build knob functional.
510 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
511 Patch by Simon Arlott
513 TF/03 Fix valgrind.h portability to C89 compilers that do not support
514 variable argument macros. Our copy now differs from upstream.
520 TF/01 Failure to get a lock on a hints database can have serious
521 consequences so log it to the panic log.
523 TF/02 Log LMTP confirmation messages in the same way as SMTP,
524 controlled using the smtp_confirmation log selector.
526 TF/03 Include the error message when we fail to unlink a spool file.
528 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
529 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
530 for maintaining out-of-tree patches for some time.
532 PP/01 Bugzilla 139: Documentation and portability issues.
533 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
534 Handle per-OS dynamic-module compilation flags.
536 PP/02 Let /dev/null have normal permissions.
537 The 4.73 fixes were a little too stringent and complained about the
538 permissions on /dev/null. Exempt it from some checks.
539 Reported by Andreas M. Kirchwitz.
541 PP/03 Report version information for many libraries, including
542 Exim version information for dynamically loaded libraries. Created
543 version.h, now support a version extension string for distributors
544 who patch heavily. Dynamic module ABI change.
546 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
547 privilege escalation vulnerability whereby the Exim run-time user
548 can cause root to append content of the attacker's choosing to
551 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
554 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
555 If dropping privileges for untrusted macros, we disabled normal logging
556 on the basis that it would fail; for the Exim run-time user, this is not
557 the case, and it resulted in successful deliveries going unlogged.
558 Fixed. Reported by Andreas Metzler.
564 PP/01 Date: & Message-Id: revert to normally being appended to a message,
565 only prepend for the Resent-* case. Fixes regression introduced in
566 Exim 4.70 by NM/22 for Bugzilla 607.
568 PP/02 Include check_rfc2047_length in configure.default because we're seeing
569 increasing numbers of administrators be bitten by this.
571 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
573 PP/03 Bugzilla 994: added openssl_options main configuration option.
575 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
577 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
579 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
581 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
582 without a peer certificate, leading to a segfault because of an
583 assumption that peers always have certificates. Be a little more
584 paranoid. Problem reported by Martin Tscholak.
586 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
587 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
588 NB: ClamAV planning to remove STREAM in "middle of 2010".
589 CL also introduces -bmalware, various -d+acl logging additions and
590 more caution in buffer sizes.
592 PP/09 Implemented reverse_ip expansion operator.
594 PP/10 Bugzilla 937: provide a "debug" ACL control.
596 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
598 PP/12 Bugzilla 973: Implement --version.
600 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
602 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
604 PP/15 Bugzilla 816: support multiple condition rules on Routers.
606 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
607 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
608 ignore trailing whitespace.
610 JJ/02 prevent non-panic DKIM error from being sent to paniclog
612 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
615 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
616 Notification from Dr Andrew Aitchison.
618 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
619 ExtendedDetectionInfo response format.
620 Notification from John Horne.
622 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
625 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
626 XSL and documented dependency on system catalogs, with examples of how
629 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
632 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
633 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
634 configuration file which is writeable by the Exim user or group.
636 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
637 of configuration files to cover files specified with the -C option if
638 they are going to be used with root privileges, not just the default
641 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
642 option (effectively making it always true).
644 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
645 files to be used while preserving root privileges.
647 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
648 that rogue child processes cannot use them.
650 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
651 run-time user, instead of root.
653 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
654 Exim run-time user without dropping privileges.
656 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
657 result string, instead of calling string_vformat() twice with the same
660 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
661 for other users. Others should always drop root privileges if they use
662 -C on the command line, even for a whitelisted configure file.
664 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
666 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
672 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
673 $data_path, and $header_path variables; fixed documentation bugs and
676 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
677 exipick to access non-standard spools, including the "frozen" queue
680 NM/01 Bugzilla 965: Support mysql stored procedures.
681 Patch from Alain Williams
683 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
685 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
686 Patch from Andreas Metzler
688 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
689 Patch from Kirill Miazine
691 NM/05 Bugzilla 671: Added umask to procmail example.
693 JJ/03 installed exipick 20100323.0, fixing doc bug
695 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
696 directory. Notification and patch from Dan Rosenberg.
698 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
700 TK/02 Improve log output when DKIM signing operation fails.
702 MH/01 Treat the transport option dkim_domain as a colon separated
703 list, not as a single string, and sign the message with each element,
704 omitting multiple occurences of the same signer.
706 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
707 Bugzilla 985, 986. Patch by Simon Arlott
709 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
710 Patch by Simon Arlott
712 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
713 MBX locking. Notification from Dan Rosenberg.
719 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
721 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
723 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
725 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
727 NM/04 Bugzilla 915: Flag broken perl installation during build.
733 TK/01 Added patch by Johannes Berg that expands the main option
734 "spamd_address" if it starts with a dollar sign.
736 TK/02 Write list of recipients to X-Envelope-Sender header when building
737 the mbox-format spool file for content scanning (suggested by Jakob
740 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
741 (http://www.dcc-servers.net/) support via dccifd. Activated by
742 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
744 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
745 by Mark Daniel Reidel <mr@df.eu>.
747 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
748 When building exim an external PCRE library is now needed -
749 PCRE is a system library on the majority of modern systems.
750 See entry on PCRE_LIBS in EDITME file.
752 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
753 conversation. Added nologin parameter to request.
754 Patch contributed by Kirill Miazine.
756 TF/01 Do not log submission mode rewrites if they do not change the address.
758 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
760 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
761 log files in place. Contributed by Roberto Lima.
763 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
765 TF/03 Bugzilla 615: When checking the local_parts router precondition
766 after a local_part_suffix or local_part_prefix option, Exim now
767 does not use the address's named list lookup cache, since this
768 contains cached lookups for the whole local part.
770 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
771 Robert Millan. Documentation is in experimental-spec.txt.
773 TF/04 Bugzilla 668: Fix parallel build (make -j).
775 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
777 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
778 Patch provided by Jan Srzednicki.
780 TF/05 Leading white space used to be stripped from $spam_report which
781 wrecked the formatting. Now it is preserved.
783 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
784 that they are available at delivery time.
786 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
788 TF/08 TLS error reporting now respects the incoming_interface and
789 incoming_port log selectors.
791 TF/09 Produce a more useful error message if an SMTP transport's hosts
792 setting expands to an empty string.
794 NM/06 Bugzilla 744: EXPN did not work under TLS.
795 Patch provided by Phil Pennock.
797 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
798 Patch provided by Richard Godbee.
800 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
801 acl_smtp_notquit, added index entry.
803 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
804 Patch provided by Eugene Bujak.
806 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
807 accept(). Patch provided by Maxim Dounin.
809 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
810 Patch provided by Phil Pennock.
812 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
814 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
815 Patch provided by Brad "anomie" Jorsch.
817 NM/14 Bugzilla 622: Added timeout setting to transport filter.
818 Patch provided by Dean Brooks.
820 TK/05 Add native DKIM support (does not depend on external libraries).
822 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
823 Patch provided by Graeme Fowler.
825 NM/16 Bugzilla 851: Documentation example syntax fix.
827 NM/17 Changed NOTICE file to remove references to embedded PCRE.
829 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
832 NM/19 Bugzilla 745: TLS version reporting.
833 Patch provided by Phil Pennock.
835 NM/20 Bugzilla 167: bool: condition support.
836 Patch provided by Phil Pennock.
838 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
839 clients. Patch provided by Phil Pennock.
841 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
842 Patch provided by Brad "anomie" Jorsch.
844 NM/23 Bugzilla 687: Fix misparses in eximstats.
845 Patch provided by Heiko Schlittermann.
847 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
848 Patch provided by Heiko Schlittermann.
850 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
851 plus update to original patch.
853 NM/26 Bugzilla 799: Documentation correction for ratelimit.
855 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
856 Patch provided by David Brownlee.
858 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
860 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
862 NM/30 Bugzilla 888: TLS documentation bugfixes.
864 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
866 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
867 Unlike the original bugzilla I have changed all shell scripts in src tree.
869 NM/33 Bugzilla 898: Transport filter timeout fix.
870 Patch by Todd Rinaldo.
872 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
873 Patch by Serge Demonchaux.
875 NM/35 Bugzilla 39: Base64 decode bug fixes.
876 Patch by Jakob Hirsch.
878 NM/36 Bugzilla 909: Correct connect() call in dcc code.
880 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
882 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
884 NM/39 Bugzilla 911: Fixed MakeLinks build script.
890 TK/01 Add preliminary DKIM support. Currently requires a forked version of
891 ALT-N's libdkim that I have put here:
892 http://duncanthrax.net/exim-experimental/
894 Note to Michael Haardt: I had to rename some vars in sieve.c. They
895 were called 'true' and it seems that C99 defines that as a reserved
896 keyword to be used with 'bool' variable types. That means you could
897 not include C99-style headers which use bools without triggering
898 build errors in sieve.c.
900 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
901 as mailq or other aliases. Changed the --help handling significantly
902 to do whats expected. exim_usage() emits usage/help information.
904 SC/01 Added the -bylocaldomain option to eximstats.
906 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
908 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
910 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
916 PH/01 Another patch from the Sieve maintainer.
918 PH/02 When an IPv6 address is converted to a string for single-key lookup
919 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
920 dots are used instead of colons so that keys in lsearch files need not
921 contain colons. This was done some time before quoting was made available
922 in lsearch files. However, iplsearch files do require colons in IPv6 keys
923 (notated using the quote facility) so as to distinguish them from IPv4
924 keys. This meant that lookups for IP addresses in host lists did not work
925 for iplsearch lookups.
927 This has been fixed by arranging for IPv6 addresses to be expressed with
928 colons if the lookup type is iplsearch. This is not incompatible, because
929 previously such lookups could never work.
931 The situation is now rather anomolous, since one *can* have colons in
932 ordinary lsearch keys. However, making the change in all cases is
933 incompatible and would probably break a number of configurations.
935 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
938 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
939 conversion specification without a maximum field width, thereby enabling
940 a rogue spamd server to cause a buffer overflow. While nobody in their
941 right mind would setup Exim to query an untrusted spamd server, an
942 attacker that gains access to a server running spamd could potentially
943 exploit this vulnerability to run arbitrary code as the Exim user.
945 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
946 $primary_hostname instead of what libspf2 thinks the hosts name is.
948 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
949 a directory entry by the name of the lookup key. Previously, if a
950 symlink pointed to a non-existing file or a file in a directory that
951 Exim lacked permissions to read, a lookup for a key matching that
952 symlink would fail. Now it is enough that a matching directory entry
953 exists, symlink or not. (Bugzilla 503.)
955 PH/03 The body_linecount and body_zerocount variables are now exported in the
958 PH/04 Added the $dnslist_matched variable.
960 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
961 This means they are set thereafter only if the connection becomes
964 PH/06 Added the client_condition to authenticators so that some can be skipped
965 by clients under certain conditions.
967 PH/07 The error message for a badly-placed control=no_multiline_responses left
968 "_responses" off the end of the name.
970 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
972 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
973 (without spaces) instead of just copying the configuration text.
975 PH/10 Added the /noupdate option to the ratelimit ACL condition.
977 PH/11 Added $max_received_linelength.
979 PH/12 Added +ignore_defer and +include_defer to host lists.
981 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
982 way in which PCRE > 7.0 is built.
984 PH/14 Implemented queue_only_load_latch.
986 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
987 MAIL command. The effect was to mangle the value on 64-bit systems.
989 PH/16 Another patch from the Sieve maintainer.
991 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
993 PH/18 If a system quota error occurred while trying to create the file for
994 a maildir delivery, the message "Mailbox is full" was not appended to the
995 bounce if the delivery eventually timed out. Change 4.67/27 below applied
996 only to a quota excession during the actual writing of the file.
998 PH/19 It seems that peer DN values may contain newlines (and other non-printing
999 characters?) which causes problems in log lines. The DN values are now
1000 passed through string_printing() before being added to log lines.
1002 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1003 and InterBase are left for another time.)
1005 PH/21 Added message_body_newlines option.
1007 PH/22 Guard against possible overflow in moan_check_errorcopy().
1009 PH/23 POSIX allows open() to be a macro; guard against that.
1011 PH/24 If the recipient of an error message contained an @ in the local part
1012 (suitably quoted, of course), incorrect values were put in $domain and
1013 $local_part during the evaluation of errors_copy.
1019 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1020 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1023 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1024 issue a MAIL command.
1026 PH/02 In an ACL statement such as
1028 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1030 if a client was not listed at all, or was listed with a value other than
1031 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1032 the condition was not true (as it should be), so access was not denied.
1033 The bug was that the ! inversion was incorrectly passed on to the second
1034 item. This has been fixed.
1036 PH/03 Added additional dnslists conditions == and =& which are different from
1037 = and & when the dns lookup returns more than one IP address.
1039 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1040 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1042 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1043 FSYNC, which compiles an option called disable_fsync that allows for
1044 bypassing fsync(). The documentation is heavily laced with warnings.
1046 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1048 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1049 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1050 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1051 including adding "make clean"; (3) Added -fPIC when compiling the test
1052 dynamically loaded module, to get rid of a warning.
1054 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1055 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1056 = 0s. The bug is otherwise harmless.
1058 PH/07 There was a bug in the dovecot authenticator such that the value of
1059 $auth1 could be overwritten, and so not correctly preserved, after a
1060 successful authentication. This usually meant that the value preserved by
1061 the server_setid option was incorrect.
1063 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1065 PH/09 Installed PCRE release 7.0.
1067 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1068 run for batched SMTP input. It is now run at the start of every message
1069 in the batch. While fixing this I discovered that the process information
1070 (output by running exiwhat) was not always getting set for -bs and -bS
1071 input. This is fixed, and it now also says "batched" for BSMTP.
1073 PH/11 Added control=no_pipelining.
1075 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1076 patch, slightly modified), and move the expansion of helo_data till after
1077 the connection is made in the smtp transport (so it can use these
1080 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1082 PH/14 Added log_selector = +pid.
1084 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1086 PH/16 Add ${if forany and ${if forall.
1088 PH/17 Added dsn_from option to vary the From: line in DSNs.
1090 PH/18 Flush SMTP output before performing a callout, unless control =
1091 no_callout_flush is set.
1093 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1094 was true (the default) a successful delivery failed to delete the retry
1095 item, thus causing premature timeout of the address. The bug is now
1098 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1100 PH/21 Long custom messages for fakedefer and fakereject are now split up
1101 into multiline reponses in the same way that messages for "deny" and
1102 other ACL rejections are.
1104 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1105 with slight modification.
1107 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1108 draft, changing the syntax and factoring some duplicate code".
1110 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1111 for deliveries of the second and subsequent messages over the same SMTP
1114 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1115 ${reduce, with only minor "tidies".
1117 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1119 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1120 expansion side effects.
1122 PH/27 When a message times out after an over-quota error from an Exim-imposed
1123 quota, the bounce message says "mailbox is full". This message was not
1124 being given when it was a system quota that was exceeded. It now should
1127 MH/03 Made $recipients available in local_scan(). local_scan() already has
1128 better access to the recipient list through recipients_list[], but
1129 $recipients can be useful in postmaster-provided expansion strings.
1131 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1132 in the case of a MAIL command with additional options following the
1133 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1134 were accidentally chopped off.
1136 PH/29 SMTP synchronization checks are implemented when a command is read -
1137 there is a check that no more input is waiting when there shouldn't be
1138 any. However, for some commands, a delay in an ACL can mean that it is
1139 some time before the response is written. In this time, more input might
1140 arrive, invalidly. So now there are extra checks after an ACL has run for
1141 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1142 pipelining has not been advertised.
1144 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1146 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1147 returned in the SMTP response when smtp_return_error_details was false.
1148 This has been fixed.
1150 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1151 instead of the C I/O that was originally supplied, because problems were
1152 reported on Solaris.
1154 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1155 Exim which did not show up earlier: it was assuming that a call to
1156 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1157 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1158 was a macro that became an assignment, so it seemed to work. This has
1159 changed to a proper function call with a void return, hence the compile
1160 error. Exim's code has been fixed.
1162 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1165 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1167 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1169 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1170 because it turns out to be too big or there is a timeout) there is no
1171 "Completed" line in the log. When some messages of this type were
1172 selected by exigrep, they were listed as "not completed". Others were
1173 picked up by some special patterns. I have improved the selection
1174 criteria to be more general.
1176 PH/38 The host_find_failed option in the manualroute router can now be set
1177 to "ignore", to completely ignore a host whose IP address cannot be
1178 found. If all hosts are ignored, the behaviour is controlled by the new
1179 host_all_ignored option.
1181 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1182 homing or because of multiple MX records with /mx) generated more than
1183 one IP address, and the following item turned out to be the local host,
1184 all the secondary addresses of the first item were incorrectly removed
1185 from the list, along with the local host and any following hosts (which
1186 is what is supposed to happen).
1188 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1189 whoever called Exim into the -H file. In the case of the daemon it was
1190 behaving confusingly. When first started, it used values for whoever
1191 started the daemon, but after a SIGHUP it used the Exim user (because it
1192 calls itself on a restart). I have changed the code so that it now always
1195 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1196 message are rejected with the same error (e.g. no authentication or bad
1197 sender address), and a DATA command is nevertheless sent (as can happen
1198 with PIPELINING or a stupid MUA), the error message that was given to the
1199 RCPT commands is included in the rejection of the DATA command. This is
1200 intended to be helpful for MUAs that show only the final error to their
1203 PH/42 Another patch from the Sieve maintainer.
1205 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1206 Eximstats - Fixed some broken HTML links and added missing column headers
1208 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1209 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1211 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1213 PH/43 Yet another patch from the Sieve maintainer.
1215 PH/44 I found a way to check for a TCP/IP connection going away before sending
1216 the response to the final '.' that terminates a message, but only in the
1217 case where the client has not sent further data following the '.'
1218 (unfortunately, this is allowed). However, in many cases there won't be
1219 any further data because there won't be any more messages to send. A call
1220 to select() can be used: if it shows that the input is "ready", there is
1221 either input waiting, or the socket has been closed. An attempt to read
1222 the next input character can distinguish the two cases. Previously, Exim
1223 would have sent an OK response which the client would never have see.
1224 This could lead to message repetition. This fix should cure that, at
1225 least in a lot of common cases.
1227 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1228 advertised in response to EHLO.
1234 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1235 fixed by 4.65/MH/01 (is this a record?) are fixed:
1237 (i) An empty string was always treated as zero by the numeric comparison
1238 operators. This behaviour has been restored.
1240 (ii) It is documented that the numeric comparison operators always treat
1241 their arguments as decimal numbers. This was broken in that numbers
1242 starting with 0 were being interpreted as octal.
1244 While fixing these problems I realized that there was another issue that
1245 hadn't been noticed. Values of message_size_limit (both the global option
1246 and the transport option) were treated as octal if they started with 0.
1247 The documentation was vague. These values are now always treated as
1248 decimal, and I will make that clear in the documentation.
1254 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1255 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1258 MH/01 Don't check that the operands of numeric comparison operators are
1259 integers when their expansion is in "skipping" mode (fixes bug
1260 introduced by 4.64-PH/07).
1262 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1263 child addresses, Exim now panics and dies. Previously, because the count
1264 is held in a short int, deliveries were likely to be lost. As such a
1265 large number of recipients for a single message is ridiculous
1266 (performance will be very, very poor), I have chosen to impose a limit
1267 rather than extend the field.
1273 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1274 leftover -K file (the existence of which was triggered by #402).
1275 While we were at it, introduced process PID as part of the -K
1276 filename. This should rule out race conditions when creating
1279 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1280 processing considerably. Previous code took too long for large mails,
1281 triggering a timeout which in turn triggers #401.
1283 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1284 in the DK code in transports.c. sendfile() is not really portable,
1285 hence the _LINUX specificness.
1287 TF/01 In the add_headers option to the mail command in an Exim filter,
1288 there was a bug that Exim would claim a syntax error in any
1289 header after the first one which had an odd number of characters
1292 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1293 callout verification, Exim cached a "reject" for the entire domain. This
1294 is correct for most verifications, but it is not correct for a recipient
1295 verification with use_sender or use_postmaster set, because in that case
1296 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1297 case of MAIL FROM:<> rejection from other early rejections (e.g.
1298 rejection of HELO). When verifying a recipient using a non-null MAIL
1299 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1300 Whatever the result of the callout, the value of the domain cache is
1301 left unchanged (for any other kind of callout, getting as far as trying
1302 RCPT means that the domain itself is ok).
1304 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1307 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1308 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1309 session, the daemon ignores ECONNECT errors and logs others; it now
1310 ignores EPIPE as well.
1312 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1313 (quoted-printable decoding).
1315 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1316 later the small subsequent patch to fix an introduced bug.
1318 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1320 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1322 PH/08 An error is now given if message_size_limit is specified negative.
1324 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1325 to be given (somewhat) arbitrary names.
1327 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1330 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1331 miscellaneous code fixes
1333 PH/10 Added the log_reject_target ACL modifier to specify where to log
1336 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1337 hostname. This is wrong, because it relates to the incoming message (and
1338 probably the interface on which it is arriving) and not to the outgoing
1339 callout (which could be using a different interface). This has been
1340 changed to use the value of the helo_data option from the smtp transport
1341 instead - this is what is used when a message is actually being sent. If
1342 there is no remote transport (possible with a router that sets up host
1343 addresses), $smtp_active_hostname is used.
1345 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1346 tweaks were necessary in order to get it to work (see also 21 below):
1347 (a) The code assumed that strncpy() returns a negative number on buffer
1348 overflow, which isn't the case. Replaced with Exim's string_format()
1350 (b) There were several signed/unsigned issues. I just did the minimum
1351 hacking in of casts. There is scope for a larger refactoring.
1352 (c) The code used strcasecmp() which is not a standard C function.
1353 Replaced with Exim's strcmpic() function.
1354 (d) The code set only $1; it now sets $auth1 as well.
1355 (e) A simple test gave the error "authentication client didn't specify
1356 service in request". It would seem that Dovecot has changed its
1357 interface. Fortunately there's a specification; I followed it and
1358 changed what the client sends and it appears to be working now.
1360 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1363 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1364 address A is aliased to B and C, where B exists and C does not. Without
1365 -v the output is "A verified" because verification stops after a
1366 successful redirection if more than one address is generated. However,
1367 with -v the child addresses are also verified. Exim was outputting "A
1368 failed to verify" and then showing the successful verification for C,
1369 with its parentage. It now outputs "B failed to verify", showing B's
1370 parentage before showing the successful verification of C.
1372 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1373 look up a TXT record in a specific list after matching in a combined
1376 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1377 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1378 they consult the DNS. I had assumed they would set it the way they
1379 wanted; and indeed my experiments on Linux seem to show that in some
1380 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1381 To be on the safe side, however, I have now made the interface to
1382 host_find_byname() similar to host_find_bydns(), with an argument
1383 containing the DNS resolver options. The host_find_byname() function now
1384 sets these options at its start, just as host_find_bydns() does. The smtp
1385 transport options dns_qualify_single and dns_search_parents are passed to
1386 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1387 of host_find_byname() use the default settings of RES_DEFNAMES
1388 (qualify_single) but not RES_DNSRCH (search_parents).
1390 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1391 spool_read_header() do less string testing, by means of a preliminary
1392 switch on the second character of optional "-foo" lines. (This is
1393 overdue, caused by the large number of possibilities that now exist.
1394 Originally there were few.) While I was there, I also converted the
1395 str(n)cmp tests so they don't re-test the leading "-" and the first
1396 character, in the hope this might squeeze out yet more improvement.
1398 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1399 flag allowing group syntax was set by the header_syntax check but not
1400 turned off, possible causing trouble later; (2) The flag was not being
1401 set at all for the header_verify test, causing "group"-style headers to
1402 be rejected. I have now set it in this case, and also caused header_
1403 verify to ignore an empty address taken from a group. While doing this, I
1404 came across some other cases where the code for allowing group syntax
1405 while scanning a header line wasn't quite right (mostly, not resetting
1406 the flag correctly in the right place). These bugs could have caused
1407 trouble for malformed header lines. I hope it is now all correct.
1409 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1410 with the "reply" argument non-NULL. The code, however (which originally
1411 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1412 but it didn't always do it. This confused somebody who was copying the
1413 code for some other use. I have removed all the tests.
1415 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1416 feature that was used to support insecure browsers during the U.S. crypto
1417 embargo. It requires special client support, and Exim is probably the
1418 only MTA that supported it -- and would never use it because real RSA is
1419 always available. This code has been removed, because it had the bad
1420 effect of slowing Exim down by computing (never used) parameters for the
1421 RSA_EXPORT functionality.
1423 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1424 authenticator to fail if there's a tab character in the incoming data
1425 (there should never be unless someone is messing about, as it's supposed
1426 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1427 if the connection is using TLS or if the remote IP is the same as the
1428 local IP, and the "valid-client-cert option" if a client certificate has
1431 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1432 authenticators. This can be used for authorization after authentication
1433 succeeds. (In the case of plaintext, it servers for both authentication
1436 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1437 if any retry times were supplied.
1439 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1440 connection, where there is no client IP address to check. In this
1441 situation, the verify now always succeeds.
1443 PH/25 Applied John Jetmore's -Mset patch.
1445 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1447 PH/27 In a string expansion for a processed (not raw) header when multiple
1448 headers of the same name were present, leading whitespace was being
1449 removed from all of them, but trailing whitespace was being removed only
1450 from the last one. Now trailing whitespace is removed from each header
1451 before concatenation. Completely empty headers in a concatenation (as
1452 before) are ignored.
1454 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1455 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1457 PH/29 [Removed. This was a change that I later backed out, and forgot to
1458 correct the ChangeLog entry (that I had efficiently created) before
1459 committing the later change.]
1461 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1462 address errors (4xx response to RCPT) over the same connection as other
1463 messages routed to the same hosts. Such deliveries are always "forced",
1464 so retry times are not inspected. This resulted in far too many retries
1465 for the affected addresses. The effect occurred only when there were more
1466 hosts than the hosts_max_try setting in the smtp transport when it had
1467 the 4xx errors. Those hosts that it had tried were not added to the list
1468 of hosts for which the message was waiting, so if all were tried, there
1469 was no problem. Two fixes have been applied:
1471 (i) If there are any address or message errors in an SMTP delivery, none
1472 of the hosts (tried or untried) are now added to the list of hosts
1473 for which the message is waiting, so the message should not be a
1474 candidate for sending over the same connection that was used for a
1475 successful delivery of some other message. This seems entirely
1476 reasonable: after all the message is NOT "waiting for some host".
1477 This is so "obvious" that I'm not sure why it wasn't done
1478 previously. Hope I haven't missed anything, but it can't do any
1479 harm, as the worst effect is to miss an optimization.
1481 (ii) If, despite (i), such a delivery is accidentally attempted, the
1482 routing retry time is respected, so at least it doesn't keep
1483 hammering the server.
1485 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1486 in ${readsocket because some servers need this prod.
1488 PH/32 Added some extra debug output when updating a wait-xxx database.
1490 PH/33 The hint "could be header name not terminated by colon", which has been
1491 given for certain expansion errors for a long time, was not being given
1492 for the ${if def:h_colon_omitted{... case.
1494 PH/34 The spec says: "With one important exception, whenever a domain list is
1495 being scanned, $domain contains the subject domain." There was at least
1496 one case where this was not true.
1498 PH/35 The error "getsockname() failed: connection reset by peer" was being
1499 written to the panic log as well as the main log, but it isn't really
1500 panic-worthy as it just means the connection died rather early on. I have
1501 removed the panic log writing for the ECONNRESET error when getsockname()
1504 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1505 runs only) independently of the message's sender address. This meant
1506 that, if the 4xx error was in fact related to the sender, a different
1507 message to the same recipient with a different sender could confuse
1508 things. In particualar, this can happen when sending to a greylisting
1509 server, but other circumstances could also provoke similar problems.
1510 I have changed the default so that the retry time for these errors is now
1511 based a combination of the sender and recipient addresses. This change
1512 can be overridden by setting address_retry_include_sender=false in the
1515 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1516 remote server are returned as part of bounce messages. This was not
1517 happening for LMTP over a pipe (the lmtp transport), but now it is the
1518 same for both kinds of LMTP.
1520 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1521 in header lines that were in fact CNAMEs. This is no longer the case.
1523 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1524 and queue runs started by the daemon processed all messages. This has
1525 been fixed so that -R and -S can now usefully be given with -q<time>.
1527 PH/40 Import PCRE release 6.7 (fixes some bugs).
1529 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1531 PH/42 Give an error if -q is specified more than once.
1533 PH/43 Renamed the variables $interface_address and $interface_port as
1534 $received_ip_address and $received_port, to make it clear that these
1535 values apply to message reception, and not to the outgoing interface when
1536 a message is delivered. (The old names remain recognized, of course.)
1538 PH/44 There was no timeout on the connect() call when using a Unix domain
1539 socket in the ${readsocket expansion. There now is.
1541 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1542 be meaningful with "accept".
1544 SC/01 Eximstats V1.43
1545 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1547 SC/02 Eximstats V1.44
1548 Use a glob alias rather than an array ref in the generated
1549 parser. This improves both readability and performance.
1551 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1552 Collect SpamAssassin and rejection statistics.
1553 Don't display local sender or destination tables unless
1554 there is data to show.
1555 Added average volumes into the top table text output.
1557 SC/04 Eximstats V1.46
1558 Collect data on the number of addresses (recipients)
1559 as well as the number of messages.
1561 SC/05 Eximstats V1.47
1562 Added 'Message too big' to the list of mail rejection
1563 reasons (thanks to Marco Gaiarin).
1565 SC/06 Eximstats V1.48
1566 Mainlog lines which have GMT offsets and are too short to
1567 have a flag are now skipped.
1569 SC/07 Eximstats V1.49 (Alain Williams)
1570 Added the -emptyok flag.
1572 SC/08 Eximstats V1.50
1573 Fixes for obtaining the IP address from reject messages.
1575 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1576 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1577 whitesspace changes from 4.64-PH/27
1579 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1582 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1583 are found, allow negative numbers in numeric criteria)
1585 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1587 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1588 to match changes made in 4.64-PH/43
1590 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1592 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1593 "domain" is actually an IP address, and force a failure. This locks out
1594 those revolvers/nameservers that support "A-for-A" lookups, in
1595 contravention of the specifications.
1597 PH/48 When a host name was looked up from an IP address, and the subsequent
1598 forward lookup of the name timed out, the host name was left in
1599 $sender_host_name, contrary to the specification.
1601 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1602 restricted to single-key lookups, Exim was not diagnosing an error if
1603 * or *@ was used with a query-style lookup.
1605 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1607 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1608 long ago, but noone interested enough thought of it. Let's just say that
1609 the "1.1" means that there are some new functions that weren't there at
1610 some point in the past.
1612 PH/51 Error processing for expansion failure of helo_data from an smtp
1613 transport during callout processing was broken.
1615 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1616 tested/used via the -bh/-bhc/-bs options.
1618 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1619 bug, fixed in subsequent PCRE releases).
1621 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1622 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1624 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1630 SC/01 Use a glob alias rather than an array ref in eximstats generated
1631 parser. This improves both readability and performance.
1633 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1634 Don't display local sender or destination tables in eximstats unless
1635 there is data to show.
1636 Added average volumes into the eximstats top table text output.
1638 SC/03 Collect data on the number of addresses (recipients) as well
1639 as the number of messages in eximstats.
1641 TF/01 Correct an error in the documentation for the redirect router. Exim
1642 does (usually) call initgroups() when daemonizing.
1644 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1645 with consistent privilege compared to when running as a daemon.
1647 TF/03 Note in the spec that $authenticated_id is not set for local
1648 submissions from trusted users.
1650 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1651 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1653 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1654 by adding some example configuration directives to the default
1655 configuration file. A little bit of work is required to uncomment the
1656 directives and define how usernames and passwords are checked, but
1657 there is now a framework to start from.
1659 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1660 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1661 without this. I don't know how relevant this is to other LDAP libraries.
1663 PH/02 Add the verb name to the "unknown ACL verb" error.
1665 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1667 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1669 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1670 directory not expanded when it should be if an expanded home directory
1671 was set for the address (which is overridden by the transport).
1673 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1676 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1677 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1678 because it is too late at that time, and has no effect.
1680 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1681 security issue with \' (bugzilla #107). I could not use the
1682 PQescapeStringConn() function, because it needs a PGconn value as one of
1685 PH/08 When testing addresses using -bt, indicate those final addresses that
1686 are duplicates that would not cause an additional delivery. At least one
1687 person was confused, thinking that -bt output corresponded to deliveries.
1688 (Suppressing duplicates isn't a good idea as you lose the information
1689 about possibly different redirections that led to the duplicates.)
1691 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1692 systems where poll() doesn't work, in particular OS X.
1694 PH/10 Added more information to debugging output for retry time not reached.
1696 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1697 operations in malware.c.
1699 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1702 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1703 syslog_duplication set false, log lines that would normally be written
1704 both the the main log and to the reject log were not written to syslog at
1707 PH/14 In the default configuration, change the use of "message" in ACL warn
1708 statements to "add_header".
1710 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1711 not followed by a command (e.g. "seen endif").
1713 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1714 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1717 PH/17 Added extra conditions to the default value of delay_warning_condition
1721 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1722 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1723 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1726 The Auto-Submitted: and various List- headers are standardised, whereas I
1727 don't think Precedence: ever was.
1729 PH/18 Refactored debugging code in route_finduser() to show more information,
1730 in particular, the error code if getpwnam() issues one.
1732 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1733 This is apparently needed in addition to the PH/07 change above to avoid
1734 any possible encoding problems.
1736 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1737 but not after initializing Perl.
1739 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1740 output them only if debugging. By default they are written stderr,
1741 apparently, which is not desirable.
1743 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1746 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1749 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1751 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1752 authenticated or an ident call has been made. Suppress the default
1753 values for $authenticated_id and $authenticated_sender (but permit -oMai
1754 and -oMas) when testing with -bh.
1756 PH/24 Re-jigged the order of the tests in the default configuration so that the
1757 tests for valid domains and recipients precede the DNS black list and CSA
1758 tests, on the grounds that those ones are more expensive.
1760 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1761 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1762 This bug exists in every version of Exim that I still have, right back to
1765 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1766 However, an attempt to turn on case-sensitivity in a regex key by
1767 including (?-i) didn't work because the subject string was already
1768 lowercased, and the effects were non-intuitive. It turns out that a
1769 one-line patch can be used to allow (?-i) to work as expected.
1775 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1776 other effects) broke the use of negated acl sub-conditions.
1778 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1781 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1782 "Deny" causes Exim to reject the incoming connection with a 554 error.
1783 Unfortunately, if there is a major crisis, such as a disk failure,
1784 tcp-wrappers gives "deny", whereas what one would like would be some
1785 kind of temporary error. A kludge has been added to help with this.
1786 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1787 554 error is used if errno is still zero or contains ENOENT (which occurs
1788 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1791 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1793 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1794 errors. Otherwise a message that provokes a temporary error (when other
1795 messages do not) can cause a whole host to time out.
1797 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1798 addresses were routed directly to files or pipes from a redirect router.
1799 File deliveries just didn't batch; pipe deliveries might have suffered
1802 PH/06 A failure to get a lock for a hints database would erroneously always say
1803 "Failed to get write lock", even when it was really a read lock.
1805 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1806 of 0600. This has been changed to use the value of the lockfile_mode
1807 option (which defaults to 0600).
1809 PH/08 Applied small patch from the Sieve maintainer.
1811 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1812 folder from quota calculations, a direct delivery into this folder messed
1813 up the contents of the maildirsize file. This was because the regex was
1814 used only to exclude .Trash (or whatever) when the size of the mailbox
1815 was calculated. There was no check that a delivery was happening into an
1816 excluded directory. This bug has been fixed by ignoring all quota
1817 processing for deliveries into excluded directories.
1819 PH/10 Added the maildirfolder_create_regex option to appendfile.
1825 PH/01 The code for finding all the local interface addresses on a FreeBSD
1826 system running IPv6 was broken. This may well have applied to all BSD
1827 systems, as well as to others that have similar system calls. The broken
1828 code found IPv4 interfaces correctly, but gave incorrect values for the
1829 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1830 that it would not match correctly against @[] and not recognize the IPv6
1833 PH/02 The ipliteral router was not recognizing addresses of the form user@
1834 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1836 PH/03 Added disable_ipv6.
1838 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1839 decoded form, because it is most often used to construct To: headers
1840 lines in autoreplies, and the decoded form may well be syntactically
1841 invalid. However, $reply_address has leading white space removed, and all
1842 newlines turned into spaces so that the autoreply transport does not
1845 PH/05 If group was specified without a user on a router, and no group or user
1846 was specified on a transport, the group from the router was ignored.
1848 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1849 for visible compile-time settings that can be used to change these
1850 numbers, for those that want even more. Backwards compatibility with old
1851 spool files has been maintained. However, going back to a previous Exim
1852 release will lost any variables that are in spool files.
1854 PH/07 Two small changes when running in the test harness: increase delay when
1855 passing a TCP/IP connection to a new process, in case the original
1856 process has to generate a bounce, and remove special handling of
1857 127.0.0.2 (sic), which is no longer necessary.
1859 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1860 be the same on different OS.
1862 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1865 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1866 whether --show-vars was specified or not
1868 JJ/02 exipick: Added support for new ACL variable spool format introduced
1871 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1872 syntactically invalid From: or Reply-to: line, and a filter used this to
1873 generate an autoreply, and therefore failed to obtain an address for the
1874 autoreply, Exim could try to deliver to a non-existent relative file
1875 name, causing unrelated and misleading errors. What now happens is that
1876 it logs this as a hard delivery error, but does not attempt to create a
1879 PH/11 The exinext utility has a -C option for testing purposes, but although
1880 the given file was scanned by exinext itself; it wasn't being passed on
1881 when Exim was called.
1883 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1884 an end-of-file indication when reading a command response.
1886 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1887 compiled. In many other places in Exim, IPv6 addresses are always
1888 recognized, so I have changed this. It also means that IPv4 domain
1889 literals of the form [IPV4:n.n.n.n] are now always recognized.
1891 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1892 used if the router is not running as root, for example, when verifying at
1893 ACL time, or when using -bh. The debugging output from this situation was
1894 non-existent - all you got was a failure to exec. I have made two
1897 (a) Failures to set uid/gid, the current directory, or a process leader
1898 in a subprocess such as that created by queryprogram now generate
1899 suitable debugging ouput when -d is set.
1901 (b) The queryprogram router detects when it is not running as root,
1902 outputs suitable debugging information if -d is set, and then runs
1903 the subprocess without attempting to change uid/gid.
1905 PH/15 Minor change to Makefile for building test_host (undocumented testing
1908 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1909 additional section of a DNS packet that returns MX or SRV records.
1910 Instead, it always explicitly searches for A/AAAA records. This avoids
1911 major problems that occur when a DNS server includes only records of one
1912 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1913 fixed another bug: if SRV records were looked up and the corresponding
1914 address records were *not* found in the additional section, the port
1915 values from the SRV records were lost.
1917 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1918 using the correct key (the original address) when searching the retry
1919 rules in order to find which one to use for generating the retry hint.
1921 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1922 adding the default one. Similarly, if it contains a Reply-To: header, the
1923 errors_reply_to option, if set, is not used.
1925 PH/19 When calculating a retry time, Exim used to measure the "time since
1926 failure" by looking at the "first failed" field in the retry record. Now
1927 it does not use this if it is later than than the arrival time of the
1928 message. Instead it uses the arrival time. This makes for better
1929 behaviour in cases where some deliveries succeed, thus re-setting the
1930 "first failed" field. An example is a quota failure for a huge message
1931 when small messages continue to be delivered. Without this change, the
1932 "time since failure" will always be short, possible causing more frequent
1933 delivery attempts for the huge message than are intended.
1934 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1936 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1937 $1, $2, $3) because the numerical variables can be reset during some
1938 expansion items (e.g. "match"), thereby losing the authentication data.
1940 PH/21 Make -bV show the size of off_t variables so that the test suite can
1941 decide whether to run tests for quotas > 2G.
1943 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1944 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1945 filecount value is greater than 2G or if a quota value is greater than 2G
1946 on a system where the size of off_t is not greater than 4, a panic error
1949 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1950 never match. The debug and -bh output now contains an explicit error
1951 message indicating a malformed IPv4 address or mask.
1953 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1954 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1955 PH/23 above applies.
1957 PH/25 Do not write to syslog when running in the test harness. The only
1958 occasion when this arises is a failure to open the main or panic logs
1959 (for which there is an explicit test).
1961 PH/26 Added the /no_tell option to "control=freeze".
1963 PH/27 If a host name lookup failed very early in a connection, for example, if
1964 the IP address matched host_lookup and the reverse lookup yielded a name
1965 that did not have a forward lookup, an error message of the form "no IP
1966 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1967 could be logged. Now it outputs the IP address instead of "NULL".
1969 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1970 allows the sender and the authenticated sender to be set when
1971 submitting a message from within Exim. Since child_open_exim() is
1972 documented for local_scan(), the new function should be too.
1974 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1975 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1976 results in an empty string is now treated as unset.
1978 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1980 PH/31 Added sender_verify_fail logging option.
1982 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1983 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1984 tidied the source and removed it altogether.
1986 PH/33 When a queue run was abandoned because the load average was too high, a
1987 log line was always written; now it is written only if the queue_run log
1988 selector is set. In addition, the log line for abandonment now contains
1989 information about the queue run such as the pid. This is always present
1990 in "start" and "stop" lines but was omitted from the "abandon" line.
1992 PH/34 Omit spaces between a header name and the colon in the error message that
1993 is given when verify = headers_syntax fails (if there are lots of them,
1994 the message gets confusing).
1996 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1997 names, as there are now some PTR records that contain slashes. This check
1998 is only to protect against broken name servers that fall over on strange
1999 characters, so the fact that it applies to all lookups doesn't matter.
2001 PH/36 Now that the new test suite is complete, we can remove some of the
2002 special code in Exim that was needed for the old test suite. For example,
2003 sorting DNS records because real resolvers return them in an arbitrary
2004 order. The new test suite's fake resolver always returns records in the
2007 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2008 bounces) except when queue_only is set, to avoid logging races between
2009 the different processes.
2011 PH/38 Panic-die if .include specifies a non-absolute path.
2013 PH/39 A tweak to the "H" retry rule from its user.
2015 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2016 a label. They prevented compilation on older perls.
2018 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2019 a warning to be raised on newish perls.
2021 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2022 on queue. Changes to match documented behaviour of showing count of
2023 messages matching specified criteria.
2025 PH/40 Changed the default ident timeout from 30s to 5s.
2027 PH/41 Added support for the use of login_cap features, on those BSD systems
2028 that have them, for controlling the resources used by pipe deliveries.
2030 PH/42 The content-scanning code uses fopen() to create files in which to put
2031 message data. Previously it was not paying any attention to the mode of
2032 the files. Exim runs with umask(0) because the rest of the code creates
2033 files with open(), and sets the required mode explicitly. Thus, these
2034 files were ending up world-writeable. This was not a big issue, because,
2035 being within the spool directory, they were not world-accessible. I have
2036 created a function called modefopen, which takes an additional mode
2037 argument. It sets umask(777), creates the file, chmods it to the required
2038 mode, then resets the umask. All the relevant calls to fopen() in the
2039 content scanning code have been changed to use this function.
2041 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2042 to 24 hours. This avoids potential overflow problems when processing G
2043 and H retry rules. I suspect nobody ever tinkers with this value.
2045 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2047 PH/45 When the plaintext authenticator is running as a client, the server's
2048 challenges are checked to ensure they are valid base64 strings. By
2049 default, the authentication attempt is cancelled if an invalid string is
2050 received. Setting client_ignore_invalid_base64 true ignores these errors.
2051 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2052 they are received. Thus, the responses can be made to depend on the
2053 challenges. If an invalid string is ignored, an empty string is placed in
2056 PH/46 Messages that are created by the autoreply transport now contains a
2057 References: header, in accordance with RFCs 2822 and 3834.
2059 PH/47 Added authenticated_sender_force to the smtp transport.
2061 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2063 PH/49 Installed latest patch from the Sieve maintainer.
2065 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2066 was also set, the appendfile transport was unnecessarily scanning a
2067 directory of message files (e.g. for maildir delivery) to find the count
2068 of files (along with the size), even though it did not need this
2069 information. It now does the scan only if it needs to find either the
2070 size of the count of files.
2072 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2074 PH/52 Two bugs concerned with error handling when the smtp transport is
2077 (i) Exim was not creating retry information for temporary errors given
2078 for individual recipients after the DATA command when the smtp transport
2079 was used in LMTP mode. This meant that they could be retried too
2080 frequently, and not timed out correctly.
2082 (ii) Exim was setting the flag that allows error details to be returned
2083 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2084 recipients that were returned after the DATA command.
2086 PH/53 This is related to PH/52, but is more general: for any failing address,
2087 when detailed error information was permitted to be returned to the
2088 sender, but the error was temporary, then after the final timeout, only
2089 "retry timeout exceeded" was returned. Now it returns the full error as
2090 well as "retry timeout exceeded".
2092 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2093 do this, and (what is worse) MTAs that accept it.
2095 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2096 will now be deprecated.
2098 PH/56 New os.c-cygwin from the Cygwin maintainer.
2100 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2101 formats (previously only available in exim formats via -bpr, -bpru,
2102 and -bpra. Now also available in native and exiqgrep formats)
2104 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2105 with very large, slow to parse queues
2107 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2109 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2111 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2112 responses to authentication challenges, though it was showing the
2113 challenges; (ii) I've removed the CR characters from the debug output for
2116 PH/58 Allow for the insertion of a newline as well as a space when a string
2117 is turned into more than one encoded-word during RFC 2047 encoding. The
2118 Sieve code now uses this.
2120 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2121 data_4xx, lost_connection, tls_required.
2123 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2124 message was being sent as an SMTP response.
2126 PH/61 Add -l and -k options to exicyclog.
2128 PH/62 When verifying, if an address was redirected to one new address, so that
2129 verification continued, and the new address failed or deferred after
2130 having set something in $address_data, the value of $address_data was not
2131 passed back to the ACL. This was different to the case when no
2132 redirection occurred. The value is now passed back in both cases.
2134 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2135 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2136 use login_cap.h, so on its own it isn't the distinguishing feature. The
2137 new name refers directly to the setclassresources() function.
2139 PH/65 Added configuration files for NetBSD3.
2141 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2143 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2144 is preferred over IPv4.
2146 PH/68 The bounce_return_message and bounce_return_body options were not being
2147 honoured for bounces generated during the reception of non-SMTP messages.
2148 In particular, this applied to messages rejected by the ACL. This bug has
2149 been fixed. However, if bounce_return_message is true and bounce_return_
2150 body is false, the headers that are returned for a non-SMTP message
2151 include only those that have been read before the error was detected.
2152 (In the case of an ACL rejection, they have all been read.)
2154 PH/69 The HTML version of the specification is now built in a directory called
2155 spec_html instead of spec.html, because the latter looks like a path with
2156 a MIME-type, and this confuses some software.
2158 PH/70 Catch two compiler warnings in sieve.c.
2160 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2161 function verify_get_ident() calls ip_connect() to connect a socket, but
2162 if the "connect()" function timed out, ip_connect() used to close the
2163 socket. However, verify_get_ident() also closes the socket later, and in
2164 between Exim writes to the log, which may get opened at this point. When
2165 the socket was closed in ip_connect(), the log could get the same file
2166 descriptor number as the socket. This naturally causes chaos. The fix is
2167 not to close the socket in ip_connect(); the socket should be closed by
2168 the function that creates it. There was only one place in the code where
2169 this was missing, in the iplookup router, which I don't think anybody now
2170 uses, but I've fixed it anyway.
2172 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2173 well as to direct DNS lookups. Otherwise the handling of names in host
2174 lists is inconsistent and therefore confusing.
2180 PH/01 Two changes to the default runtime configuration:
2182 (1) Move the checks for relay_from_hosts and authenticated clients from
2183 after to before the (commented out) DNS black list checks.
2185 (2) Add control=submission to the relay_from_hosts and authenticated
2186 clients checks, on the grounds that messages accepted by these
2187 statements are most likely to be submissions.
2189 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2191 (1) Generate an error if the third argument for the ${prvs expansion is
2194 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2197 (3) Reset the variables that are obtained from the first argument of
2198 ${prvscheck and used in the second argument before leaving the code,
2199 because their memory is reclaimed, so using them afterwards may do
2202 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2203 one (it's much easier than Tom thought :-).
2205 (5) Because of (4), we can now allow for the use of $prvscheck_result
2206 inside the third argument.
2208 PH/03 For some reason, the default setting of PATH when running a command from
2209 a pipe transport was just "/usr/bin". I have changed it to
2212 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2213 anything to be listed in the output from -bV.
2215 PH/05 When a filter generated an autoreply, the entire To: header line was
2216 quoted in the delivery log line, like this:
2218 => >A.N.Other <ano@some.domain> <original@ddress> ...
2220 This has been changed so that it extracts the operative address. There
2221 may be more than one such address. If so, they are comma-separated, like
2224 => >ano@some.domain,ona@other.domain <original@ddress> ...
2226 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2227 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2228 not being looked up in the rDNS to get a host name, Exim was showing the
2229 IP address twice in Received: lines, even though the IP addresses were
2230 identical. For example:
2232 Received: from [1.2.3.4] (helo=[1.2.3.4])
2234 However, if the real host name was known, it was omitting the HELO data
2235 if it matched the actual IP address. This has been tidied up so that it
2236 doesn't show the same IP address twice.
2238 PH/07 When both +timestamp and +memory debugging was on, the value given by
2239 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2240 called by the debug printing, thereby overwriting the timestamp buffer.
2241 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2243 PH/08 When the original message was included in an autoreply transport, it
2244 always said "this is a copy of the message, including all the headers",
2245 even if body_only or headers_only was set. It now gives an appropriate
2248 PH/09 Applied a patch from the Sieve maintainer which:
2250 o fixes some comments
2251 o adds the (disabled) notify extension core
2252 o adds some debug output for the result of if/elsif tests
2253 o points to the current vacation draft in the documentation
2254 and documents the missing references header update
2258 o fixes a bug in processing the envelope test (when testing
2259 multiple envelope elements, the last element determinted the
2262 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2263 Electronic Mail") by including:
2265 Auto-submitted: auto-generated
2267 in the messages that it generates (bounce messages and others, such as
2268 warnings). In the case of bounce messages for non-SMTP mesages, there was
2269 also a typo: it was using "Auto_submitted" (underscore instead of
2270 hyphen). Since every message generated by Exim is necessarily in response
2271 to another message, thes have all been changed to:
2273 Auto-Submitted: auto-replied
2275 in accordance with these statements in the RFC:
2277 The auto-replied keyword:
2279 - SHOULD be used on messages sent in direct response to another
2280 message by an automatic process,
2282 - MUST NOT be used on manually-generated messages,
2284 - MAY be used on Delivery Status Notifications (DSNs) and Message
2285 Disposition Notifications (MDNs),
2287 - MUST NOT be used on messages generated by automatic or periodic
2288 processes, except for messages which are automatic responses to
2291 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2292 to the default Received: header definition.
2294 PH/12 Added log selector acl_warn_skipped (default on).
2296 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2297 variables because (a) they are in the wrong storage pool and (b) even if
2298 they were copied, it wouldn't work properly because of the caching.
2300 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2301 checking when decoding. Apparently there are clients that generate
2302 overlong encoded strings. Why am I not surprised?
2304 PH/15 If the first argument of "${if match_address" was not empty, but did not
2305 contain an "@" character, Exim crashed. Now it writes a panic log message
2306 and treats the condition as false.
2308 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2310 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2311 extension "envelope-auth". The code is finished and in agreement with
2312 other implementations, but there is no documentation so far and in fact,
2313 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2314 not changing the active code.
2316 Print executed "if" and "elsif" statements when debugging is used. This
2317 helps a great deal to understand what a filter does.
2319 Document more things not specified clearly in RFC3028. I had all this
2320 sorted out, when out of a sudden new issues came to my mind. Oops."
2322 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2325 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2326 canonical form (as documented). However, after a host name lookup from
2327 the IP address, check_host() was doing a simple string comparison with
2328 addresses acquired from the DNS when checking that the found name did
2329 have the original IP as one of its addresses. Since any found IPv6
2330 addresses are likely to be in abbreviated form, the comparison could
2331 fail. Luckily, there already exists a function for doing the comparison
2332 by converting both addresses to binary, so now that is used instead of
2333 the text comparison.
2335 PH/20 There was another similar case to PH/19, when a complete host name was
2336 given in a host list; looking up its IP address could give an abbreviated
2337 form, whereas the current host's name might or might not be abbreviated.
2338 The same fix has been applied.
2344 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2345 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2348 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2349 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2351 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2353 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2354 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2355 Also, for "Auto-Submitted", treat anything other than "no" as
2356 non-personal, in accordance with RFC 3834. (Previously it treated
2357 anything starting "auto-" as non-personal.)
2359 TF/01 The control=submission/name=... option had a problem with syntax
2360 errors if the name included a slash character. The /name= option
2361 now slurps the rest of the string, so it can include any characters
2362 but it must come last in the list of options (after /sender_retain
2365 PH/05 Some modifications to the interface to the fake nameserver for the new
2373 TK/01 Added the "success_on_redirect" address verification option. See
2374 NewStuff for rationale and an example.
2376 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2378 PH/02 Patch to exigrep to allow it to work on syslog lines.
2380 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2381 fread() to skip over the body file's header line, because in Cygwin the
2382 header line is locked and is inaccessible.
2384 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2385 co-exist for some time) to make it clear that it is the Exim ID that is
2386 referenced, not the Message-ID: header line.
2388 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2389 string_format() function, because snprintf() does not exist on all
2392 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2393 ${stat: expansion item.
2395 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2396 protocol synchronization error", to keep the pedants happy.
2398 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2399 well as for IRIX systems, when gcc is being used. See the host.c source
2402 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2404 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2407 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2408 transport and to the smtp transport in LMTP mode.
2410 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2412 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2413 run a filter in a subprocess. This could lead to confusion in subsequent
2414 lookups in the parent process. There should also be a search_tidyup() at
2415 the end of the subprocess.
2417 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2418 only if the host matched helo_try_verify_hosts, which caused the
2419 verification to occur when the EHLO/HELO command was issued. The ACL just
2420 tested the remembered result. Now, if a previous verification attempt has
2421 not happened, "verify = helo" does it there and then.
2423 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2425 TK/03 Fix log output including CR from clamd.
2427 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2428 exist provoked a memory error which could cause a segfault.
2430 PH/15 Installed PCRE 6.2
2432 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2434 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2435 of the problem. Specifically, suggested +O2 rather than +O1 for the
2438 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2440 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2441 record, the port was not correctly logged when the outgoing_port log
2442 selector was set (it logged the transort's default port).
2444 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2445 fallback_hosts, and "hosts" in the smtp transport.
2447 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2448 host errors such as "Connection refused".
2450 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2451 authentication with radiusclient 0.4.9:
2453 - Error returned from rc_read_config was caught wrongly
2454 - Username/password not passed on to radius server due to wrong length.
2456 The presumption is that some radiusclient API changes for 4.51/PH/17
2457 were not taken care of correctly. The code is still untested by me (my
2458 Linux distribution still has 0.3.2 of radiusclient), but it was
2459 contributed by a Radius user.
2461 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2462 expanding the "port" option of the smtp transport.
2464 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2465 while reading a MIME header. Thanks to Tom Hughes for a patch.
2467 PH/24 Include config.h inside local_scan.h so that configuration settings are
2470 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2471 that in an ACL for RCPT (for example), you can examine exactly what was
2474 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2475 commands, but it was not correctly comparing the address with the actual
2476 client host address. Thus, it would show the EHLO address in Received:
2477 header lines when this was not necessary.
2479 PH/27 Added the % operator to ${eval:}.
2481 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2482 it should be ignoring failures (because with -C, for example, it has lost
2483 privilege). It wasn't ignoring creation failures other than "already
2486 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2489 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2490 because an input line was too long, either on its own, or by virtue of
2491 too many continuations, the temporary file was not being removed, and the
2492 return code was incorrect.
2494 PH/31 Missing "BOOL" in function definition in filtertest.c.
2496 PH/32 Applied Sieve patches from the maintainer.
2498 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2500 PH/33 Added "verify = not_blind".
2502 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2503 Local/Makefile (with some defaults set). These are used in built scripts
2504 such as exicyclog, but they have never been used in the exim_install
2505 script (though there are many overriding facilities there). I have
2506 arranged that the exim_install script now takes note of these two
2509 PH/35 Installed configuration files for Dragonfly.
2511 PH/36 When a locally submitted message by a trusted user did not contain a
2512 From: header, and the sender address was obtained from -f or from an SMTP
2513 MAIL command, and the trusted user did not use -F to supply a sender
2514 name, $originator_name was incorrectly used when constructing a From:
2515 header. Furthermore, $originator_name was used for submission mode
2516 messages from external hosts without From: headers in a similar way,
2517 which is clearly wrong.
2519 PH/37 Added control=suppress_local_fixups.
2521 PH/38 When log_selector = +received_sender was set, and the addition of the
2522 sender made the log line's construction buffer exactly full, or one byte
2523 less than full, an overflow happened when the terminating "\n" was
2526 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2527 when the result of a list match is failure because a DNS lookup failed.
2529 PH/40 RM_COMMAND is now used in the building process.
2531 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2532 the "build-* directories that it finds.
2534 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2535 address in a domain literal was a prefix of an interface address.
2537 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2538 when verifying a sender address, unless rewrite_headers is false.
2540 PH/44 Wrote a long comment about why errors_to addresses are verified as
2541 recipients, not senders.
2543 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2544 the ratelimit ACL was added.
2546 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2548 PH/46 Added extra information about PostgreSQL errors to the error string.
2550 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2551 avoiding the need to install special zones in a real server. This is
2552 backwards compatible; if it can't find the fake resolver, it drops back.
2553 Thus, both old and new test suites can be run.
2555 TF/02 Added util/ratelimit.pl
2557 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2558 clock is set back in time.
2560 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2561 Candler <B.Candler@pobox.com>.
2563 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2564 OK for addresses that are the result of redirections.
2566 PH/48 A number of further additions for the benefit of the new test suite,
2567 including a fake gethostbyname() that interfaces to the fake DNS resolver
2570 TF/06 The fix for widen_domains has also been applied to qualify_single and
2571 search_parents which are the other dnslookup options that can cause
2574 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2577 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2579 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2580 service" if it encountered an MX record with an empty target hostname.
2581 The message is now "an MX or SRV record indicated no SMTP service".
2583 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2584 if the DNS of the sending site is misconfigured. This is quite a
2585 common situation. This change restores the behaviour of treating a
2586 helo verification defer as a failure.
2588 PH/51 If self=fail was set on a router, the bounce message did not include the
2589 actual error message.
2595 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2597 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2598 command itself ended in error, the underlying message about the transport
2599 filter timeout was being overwritten with the pipe command error. Now the
2600 underlying error message should be appended to the second error message.
2602 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2604 PH/02 Reduce the amount of output that "make" produces by default. Full output
2605 can still be requested.
2607 PH/03 The warning log line about a condition test deferring for a "warn" verb
2608 was being output only once per connection, rather than after each
2609 occurrence (because it was using the same function as for successful
2610 "warn" verbs). This seems wrong, so I have changed it.
2612 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2613 it should not have, which might have caused a crash in the right
2614 circumstances, but probably never did.
2616 PH/04 Installed a modified version of Tony Finch's patch to make submission
2617 mode fix the return path as well as the Sender: header line, and to
2618 add a /name= option so that you can make the user's friendly name appear
2621 TF/03 Added the control = fakedefer ACL modifier.
2623 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2624 Mark Lowes for thorough testing.
2626 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2628 TK/03 Merged latest SRS patch from Miles Wilton.
2630 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2631 with the definition in sysexits.h (which is #included earlier).
2632 Fortunately, Exim does not actually use EX_OK. The code used to try to
2633 preserve the sysexits.h value, by assumimg that macro definitions were
2634 scanned for macro replacements. I have been disabused of this notion,
2635 so now the code just undefines EX_OK before #including unistd.h.
2637 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2638 in the smtp transport. When a block could not be written in a single
2639 write() function, the timeout was being re-applied to each part-write.
2640 This seems wrong - if the receiver was accepting one byte at a time it
2641 would take for ever. The timeout is now adjusted when this happens. It
2642 doesn't have to be particularly precise.
2644 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2645 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2647 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2648 without a domain if the check to <postmaster@domain> fails.
2650 SC/01 Eximstats: added -xls and the ability to specify output files
2651 (patch written by Frank Heydlauf).
2653 SC/02 Eximstats: use FileHandles for outputing results.
2655 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2657 SC/04 Eximstats: fixed display of large numbers with -nvr option
2659 SC/05 Eximstats: fixed merging of reports with empty tables.
2661 SC/06 Eximstats: added the -include_original_destination flag
2663 SC/07 Eximstats: removed tabs and trailing whitespace.
2665 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2667 TK/06 MBOX spool code: Add real "From " MBOX separator line
2668 so the .eml file is really in mbox format (even though
2669 most programs do not really care). Patch from Alex Miller.
2671 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2672 The latter is generated from $received_to and is only set if the
2673 message has one envelope recipient. SA can use these headers,
2674 obviously out-of-the-box. Patch from Alex Miller.
2676 PH/08 The ${def test on a variable was returning false if the variable's
2677 value was "0", contrary to what the specification has always said!
2678 The result should be true unless the variable is empty.
2680 PH/09 The syntax error of a character other than { following "${if
2681 def:variable_name" (after optional whitespace) was not being diagnosed.
2682 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2683 accidental colon was present, for example, could give incorrect results.
2685 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2686 result is used (not including appendfile, where other changes are about
2689 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2690 This involved changing a lot of size variables from int to off_t. It
2691 should work with maildirs and everything.
2693 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2694 spamd dying while we are connected to it.
2696 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2699 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2700 function for BDB 4.3.
2702 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2704 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2705 the sources for PCRE have been re-arranged and more files are now
2708 PH/15 The code I had for printing potentially long long variables in PH/11
2709 above was not the best (it lost precision). The length of off_t variables
2710 is now inspected at build time, and an appropriate printing format (%ld
2711 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2712 to be "long long int" or "long int". This is needed for the internal
2713 formatting function string_vformat().
2715 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2716 the configuration file to be ":syslog", then the script "guesses" where
2717 the logs files are, rather than using the compiled in default. In our
2718 case the guess is not the same as the compiled default, so the script
2719 suddenly stopped working when I started to use syslog. The patch checks
2720 to see if log_file_path is "". If so, it attempts to read it from exim
2721 with no configuration file to get the compiled in version, before it
2722 falls back to the previous guessing code."
2724 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2725 implementing BATV in an Exim configuration. See NewStuff for the gory
2728 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2729 Makefile that are specific to HP-UX.
2731 PH/18 If the "use_postmaster" option was set for a recipient callout together
2732 with the "random" option, the postmaster address was used as the MAIL
2733 FROM address for the random test, but not for the subsequent recipient
2734 test. It is now used for both.
2736 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2737 patch removes a few documentation additions to RFC 3028, because the
2738 latest draft now contains them. It adds the new en;ascii-case comparator
2739 and a new error check for 8bit text in MIME parts. Comparator and
2740 require names are now matched exactly. I enabled the subaddress
2741 extension, but it is not well tested yet (read: it works for me)."
2743 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2744 rework some of the code of TK/09 above to avoid the hardwired use of
2745 "%lld" and "long long". Replaced the call to snprintf() with a call to
2748 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2749 records point to non-existent hosts", "retry timeout exceeded", and
2750 "retry time not reached for any host after a long failure period".
2752 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2753 experimental DomainKeys support:
2755 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2756 (2) On an error such as an illegally used "control", the wrong name for
2757 the control was given.
2759 These problems did NOT occur unless DomainKeys support was compiled.
2761 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2763 PH/24 Added ${if match_ip condition.
2765 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2766 databases so that it will be absolutely obvious if a crash occurs in the
2767 DB library. This is a regular occurrence (often caused by mis-matched
2770 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2771 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2772 picked up on a user's system that detects such things. There doesn't seem
2773 to be a gcc warning option for this - only an attribute that has to be
2774 put on the function's prototype. It seems that in Fedora Core 4 they have
2775 set this on a number of new functions. No doubt there will be more in due
2778 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2779 specify a transport. However, if an address that was verified by such a
2780 router was the subject of a callout, Exim crashed because it tried to
2781 read the rcpt_include_affixes from the non-existent transport. Now it
2782 just assumes that the setting of that option is false. This bug was
2783 introduced by 4.51/PH/31.
2785 PH/28 Changed -d+all to exclude +memory, because that information is very
2786 rarely of interest, but it makes the output a lot bigger. People tend to
2787 do -d+all out of habit.
2789 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2790 code in os-type was giving problems when libc.so lives in lib64, like on
2793 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2794 aren't the modern standard, and it seems that some systems' include files
2795 don't always have them. Exim was already checking for some of the newer
2796 ones like T_AAAA, and defining it itself. I've added checks for all the
2797 record types that Exim uses.
2799 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2800 not automatically generating a new one, as it is supposed to. This
2801 prevented TLS from working. If the file did exist, but contained invalid
2802 data, a new version was generated, as expected. It was only the case of a
2803 non-existent file that was broken.
2805 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2806 with a change in libdomainkeys > 0.64.
2808 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2809 from DNS. If the selector record carries the flag, it now has
2810 precedence over the domain-wide flag.
2812 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2814 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2815 the use of a port name that isn't defined in /etc/services) occurred, the
2816 message was deferred as in a normal delivery, and thus remained on the
2817 spool, instead of being failed because of the mua_wrapper setting. This
2818 is now fixed, and I tidied up some of the mua_wrapper messages at the
2821 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2822 the messages in a hash of arrays rather than using individual hashes.
2823 This is a bit cleaner and results in dramatic memory savings, albeit
2824 at a slight CPU cost.
2826 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2827 as requested by Marc Sherman.
2829 SC/10 Eximstats: added histograms for user specified patterns as requested
2832 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2834 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2835 fopen() in the content-scanning modules that did not already have it.
2841 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2842 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2844 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2846 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2848 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2849 to test Sieve filters that use "vacation".
2851 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2852 that changes the way the GnuTLS parameters are stored in the cache file.
2853 The new format can be generated externally. For backward compatibility,
2854 if the data in the cache doesn't make sense, Exim assumes it has read an
2855 old-format file, and it generates new data and writes a new file. This
2856 means that you can't go back to an older release without removing the
2859 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2860 work if there are any delivery delays because "one_time" forces the
2861 parent to be marked "delivered", so its unseen clone is never tried
2862 again. For this reason, Exim now forbids the simultaneous setting of
2865 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2866 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2867 entry if you want to know the details. The fix, however, neglected to
2868 consider the case when local delivery batching is involved. The test for
2869 "previously delivered" was not happening when checking to see if an
2870 address could be batched with a previous (undelivered) one; under
2871 certain circumstances this could lead to multiple deliveries to the same
2874 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2875 in its include files, and this causes problems building Exim.
2877 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2878 header_syntax) but Exim was just ignoring anything given after a slash.
2879 In particular, this caused confusion with an attempt to use "verify =
2880 reverse_host_lookup/defer_ok". An error is now given when options are
2881 supplied for verify items that do not have them. (Maybe reverse_host_
2882 lookup should have a defer_ok option, but that's a different point.)
2884 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2885 defined by RFC 821) to 2048, because there were problems with some AUTH
2886 commands, and RFC 1869 says the size should be increased for extended
2887 SMTP commands that take arguments.
2889 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2892 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2893 "unknown" error; now it says that the functionality isn't in the binary.
2895 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2896 an address' error message when a string expansion fails (syntax or
2897 whatever). Otherwise the password may appear in the log. Following change
2898 PH/42 below, there is no longer a chance of it appearing in a bounce
2901 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2903 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2904 of its addresses was ever tried. (Bugzilla bug #2.)
2906 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2907 the result incorrectly in the debug output. (It correctly added a newline
2908 to what was transported.)
2910 TF/01 Added $received_time.
2912 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2913 commented out examples of how to interface to a virus scanner and to
2914 SpamAssassin. Also added commented examples of av_scanner and
2915 spamd_address settings.
2917 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2918 and controls are allowed in which ACLs. There were a couple of minor
2919 errors. Some of the entries in the conditions table (which is a table of
2920 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2921 negation of where the condition IS allowed.
2923 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2925 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2926 header file does not have a version number, so I've had to invent a new
2927 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2928 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2929 radiusclient), but it was contributed by a Radius user.
2931 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2932 files or directories, for OpenSSL.
2934 PH/19 When an Exim process that is running as root has to create an Exim log
2935 file, it does so in a subprocess that runs as exim:exim so as to get the
2936 ownership right at creation (otherwise, other Exim processes might see
2937 the file with the wrong ownership). There was no test for failure of this
2938 fork() call, which would lead to the process getting stuck as it waited
2939 for a non-existent subprocess. Forks do occasionally fail when resources
2940 run out. I reviewed all the other calls to fork(); they all seem to check
2943 PH/20 When checking for unexpected SMTP input at connect time (before writing
2944 the banner), Exim was not dealing correctly with a non-positive return
2945 from the read() function. If the client had disconnected by this time,
2946 the result was a log entry for a synchronization error with an empty
2947 string after "input=" when read() returned zero. If read() returned -1
2948 (an event I could not check), uninitialized data bytes were printed.
2949 There were reports of junk text (parts of files, etc) appearing after
2952 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2954 PH/22 Added support for macro redefinition, and (re)definition in between
2955 driver and ACL definitions.
2957 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2958 forgetting to use the resulting value; it was using the unexpanded value.
2960 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2961 hadn't been configured. The fix is from Juergen Kreileder, who
2962 understands it better than I do:
2964 "Here's what I see happening with three configured cyrus_sasl
2965 authenticators configured (plain, login, cram-md5):
2967 On startup auth_cyrus_sasl_init() gets called for each of these.
2968 This means three calls to sasl_listmech() without a specified mech_list.
2969 => SASL tests which mechs of all available mechs actually work
2970 => three warnings about OTP not working
2971 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2973 With the patch, sasl_listmech() also gets called three times. But now
2974 SASL's mech_list option is set to the server_mech specified in the the
2975 authenticator. Or in other words, the answer from sasl_listmech()
2976 gets limited to just the mech you're testing for (which is different
2978 => the return list contains just 'plain' or 'login', 'cram-md5' or
2979 nothing depending on the value of ob->server_mech.
2981 I've just tested the patch: Authentication still works fine,
2982 unavailable mechs specified in the exim configuration are still
2983 caught, and the auth.log warnings about OTP are gone."
2985 PH/25 When debugging is enabled, the contents of the command line are added
2986 to the debugging output, even when log_selector=+arguments is not
2989 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2990 answer is "GNU", and only if the return is "GNU/something" is the answer
2993 PH/27 $acl_verify_message is now set immediately after the failure of a
2994 verification in an ACL, and so is available in subsequent modifiers. In
2995 particular, the message can be preserved by coding like this:
2997 warn !verify = sender
2998 set acl_m0 = $acl_verify_message
3000 Previously, $acl_verify_message was set only while expanding "message"
3001 and "log_message" when a very denied access.
3003 PH/28 Modified OS/os.c-Linux with
3005 -#ifndef OS_LOAD_AVERAGE
3006 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3008 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3009 nomenclature these days.)
3011 PH/29 Installed patch from the Sieve maintainer that adds the options
3012 sieve_useraddress and sieve_subaddress to the redirect router.
3014 PH/30 In these circumstances:
3015 . Two addresses routed to the same list of hosts;
3016 . First host does not offer TLS;
3017 . First host accepts first address;
3018 . First host gives temporary error to second address;
3019 . Second host offers TLS and a TLS session is established;
3020 . Second host accepts second address.
3021 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3022 and peerdn, if requested) that were in fact used only for the second
3025 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3026 attention to any local part prefix or suffix that was matched by the
3027 router that accepted the address. It now behaves in the same way as it
3028 does for delivery: the affixes are removed from the local part unless
3029 rcpt_include_affixes is set on the transport.
3031 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3032 timeout during the DATA phase of an incoming message.
3034 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3035 applied a patch sanctioned by the Sieve maintainer.
3037 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3038 the uid or gid is negative. A case of a negative gid caused this to be
3039 noticed. The fix allows for either to be negative.
3041 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3042 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3045 PH/36 The change PH/12 above was broken. Fixed it.
3047 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3048 the grounds that routing the same address twice would always produce the
3049 same answer. This might have been true once, but it is certainly no
3050 longer true now. Routing a child address may depend on the previous
3051 routing that produced that child. Some complicated redirection strategies
3052 went wrong when messages had multiple recipients, and made Exim's
3053 behaviour dependent on the order in which the addresses were given.
3055 I have moved the duplicate checking until after the routing is complete.
3056 Exim scans the addresses that are assigned to local and remote
3057 transports, and removes any duplicates. This means that more work will be
3058 done, as duplicates will always all be routed, but duplicates are
3059 presumably rare, so I don't expect this is of any significance.
3061 For deliveries to pipes, files, and autoreplies, the duplicate checking
3062 still happens during the routing process, since they are not going to be
3065 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3066 It corrects a timeout issue with spamd. This is Ian's comment: "The
3067 background is that sometimes spamd either never reads data from a
3068 connection it has accepted, or it never writes response data. The exiscan
3069 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3070 blindly assumes that writes won't block so it may never time out."
3072 PH/39 Allow G after quota size as well as K and M.
3074 PH/40 The value set for $authenticated_id in an authenticator may not contain
3075 binary zeroes or newlines because the value is written to log lines and
3076 to spool files. There was no check on this. Now the value is run through
3077 the string_printing() function so that such characters are converted to
3078 printable escape sequences.
3080 PH/41 $message_linecount is a new variable that contains the total number of
3081 lines in the message. Compare $body_linecount, which is the count for the
3084 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3085 bounce and delay warning messages, except in certain special cases, which
3088 (a) An SMTP error message from a remote host;
3089 (b) A message specified in a :fail: redirection;
3090 (c) A message specified in a "fail" command in a system filter;
3091 (d) A message specified in a FAIL return from the queryprogram router;
3092 (e) A message specified by the cannot_route_message router option.
3094 In these cases only, Exim does include the error details in bounce and
3095 warning messages. There are also a few cases where bland messages such
3096 as "unrouteable address" or "local delivery error" are given.
3098 PH/43 $value is now also set for the "else" part of a ${run expansion.
3100 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3101 being worked on, but at least Exim now implements the latest version to
3104 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3105 process to complete was treated as a delivery failure, a timeout while
3106 writing the message to the pipe was logged, but erroneously treated as a
3107 successful delivery. Such timeouts include transport filter timeouts. For
3108 consistency with the overall process timeout, these timeouts are now
3109 treated as errors, giving rise to delivery failures by default. However,
3110 there is now a new Boolean option for the pipe transport called
3111 timeout_defer, which, if set TRUE, converts the failures into defers for
3112 both kinds of timeout. A transport filter timeout is now identified in
3115 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3116 systems where "make" and "gmake" are different, calling "gmake" at top
3117 level broke things. I've arranged for the value of $(MAKE) to be passed
3118 from the Makefile to this script so that it can call the same version of
3122 A note about Exim versions 4.44 and 4.50
3123 ----------------------------------------
3125 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3126 changes of various kinds. As a consequence, a big documentation update was
3127 needed. This delayed the release for rather longer than seemed good, especially
3128 in the light of a couple of (minor) security issues. Therefore, the changes
3129 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3130 release. So 4.44 and 4.50 are in effect two different branches that both start
3133 I have left the 4.50 change log unchanged; it contains all the changes since
3134 4.43. The change log for 4.44 is below; many of its items are identical to
3135 those for 4.50. This seems to be the most sensible way to preserve the
3136 historical information.
3142 1. Minor wording change to the doc/README.SIEVE file.
3144 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3145 computation of the current number of files was incorrect.
3147 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3148 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3151 4. Give more explanation in the error message when the command for a transport
3152 filter fails to execute.
3154 5. There are several places where Exim runs a non-Exim command in a
3155 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3156 was being done only for the command run by the queryprogram router. It is
3157 now done for all such subprocesses. The other cases are: ${run, transport
3158 filters, and the commands run by the lmtp and pipe transports.
3160 6. Added CONFIGURE_GROUP build-time option.
3162 7. Some older OS have a limit of 256 on the maximum number of file
3163 descriptors. Exim was using setrlimit() to set 1000 as a large value
3164 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3165 systems. I've change it so that if it can't get 1000, it tries for 256.
3167 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3168 was an oversight, and furthermore, ever since the addition of extra
3169 controls (e.g. 4.43/32), the checks on when to allow different forms of
3170 "control" were broken. There should now be diagnostics for all cases when a
3171 control that does not make sense is encountered.
3173 9. Added the /retain_sender option to "control=submission".
3175 10. $recipients is now available in the predata ACL (oversight).
3177 11. Tidy the search cache before the fork to do a delivery from a message
3178 received from the command line. Otherwise the child will trigger a lookup
3179 failure and thereby defer the delivery if it tries to use (for example) a
3180 cached ldap connection that the parent has called unbind on.
3182 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3183 of $address_data from the recipient verification was clobbered by the
3184 sender verification.
3186 13. The value of address_data from a sender verification is now available in
3187 $sender_address_data in subsequent conditions in the ACL statement.
3189 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3191 15. Added a new option "connect=<time>" to callout options, to set a different
3194 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3195 was its contents. (It was OK if the option was not defined at all.)
3197 17. A "Completed" log line is now written for messages that are removed from
3198 the spool by the -Mrm option.
3200 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3201 information about exactly what failed.
3203 19. Added -dd to debug only the daemon process.
3205 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3206 handles timeouts, both on the server side and network timeouts. Renamed the
3207 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3209 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3210 It is now set to "smtps".
3212 22. $host_address is now set to the target address during the checking of
3213 ignore_target_hosts.
3215 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3216 being passed; this would have caused $sender_host_name to have been used if
3217 matching the list had actually called for a host name (not very likely,
3218 since this list is usually IP addresses). A host name is now passed as
3221 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3222 code that specifies a non-restarting handler (typically sigaction() in
3223 modern systems) in an attempt to fix a rare and obscure crash bug.
3225 25. Narrowed the window for a race in the daemon that could cause it to ignore
3226 SIGCHLD signals. This is not a major problem, because they are used only to
3227 wake it up if nothing else does.
3229 26. A malformed maildirsize file could cause Exim to calculate negative values
3230 for the mailbox size or file count. Odd effects could occur as a result.
3231 The maildirsize information is now recalculated if the size or filecount
3234 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3235 support for a long time. Removed HAVE_SYS_VFS_H.
3237 28. Installed the latest version of exipick from John Jetmore.
3239 29. In an address list, if the pattern was not a regular expression, an empty
3240 subject address (from a bounce message) matched only if the pattern was an
3241 empty string. Non-empty patterns were not even tested. This was the wrong
3242 because it is perfectly reasonable to use an empty address as part of a
3243 database query. An empty address is now tested by patterns that are
3244 lookups. However, all the other forms of pattern expect the subject to
3245 contain a local part and a domain, and therefore, for them, an empty
3246 address still always fails if the pattern is not itself empty.
3248 30. Exim went into a mad DNS loop when attempting to do a callout where the
3249 host was specified on an smtp transport, and looking it up yielded more
3250 than one IP address.
3252 31. Re-factored the code for checking spool and log partition space into a
3253 function that finds that data and another that does the check. The former
3254 is then used to implement four new variables: $spool_space, $log_space,
3255 $spool_inodes, and $log_inodes.
3257 32. The RFC2047 encoding function was originally intended for short strings
3258 such as real names; it was not keeping to the 75-character limit for
3259 encoded words that the RFC imposes. It now respects the limit, and
3260 generates multiple encoded words if necessary. To be on the safe side, I
3261 have increased the buffer size for the ${rfc2047: expansion operator from
3264 33. It is now permitted to omit both strings after an "if" condition; if the
3265 condition is true, the result is "true". As before, when the second string
3266 is omitted, a false condition yields an empty string. This makes it less
3267 cumbersome to write custom ACL and router conditions.
3269 34. Failure to deliver a bounce message always caused it to be frozen, even if
3270 there was an errors_to setting on the router. The errors_to setting is now
3273 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3274 canonical form (fully expanded) before being placed in
3275 $sender_host_address.
3277 36. The table in the code that translates DNS record types into text (T_A to
3278 "A" for instance) was missing entries for NS and CNAME. It is just possible
3279 that this could have caused confusion if both these types were looked up
3280 for the same domain, because the text type is used as part of Exim's
3281 per-process caching. But the chance of anyone hitting this buglet seems
3284 37. The dnsdb lookup has been extended in a number of ways.
3286 (1) There is a new type, "zns", which walks up the domain tree until it
3287 finds some nameserver records. It should be used with care.
3289 (2) There is a new type, "mxh", which is like "mx" except that it returns
3290 just the host names, not the priorities.
3292 (3) It is now possible to give a list of domains (or IP addresses) to be
3293 looked up. The behaviour when one of the lookups defers can be
3294 controlled by a keyword.
3296 (4) It is now possible to specify the separator character for use when
3297 multiple records are returned.
3299 38. The dnslists ACL condition has been extended: it is now possible to supply
3300 a list of IP addresses and/or domains to be looked up in a particular DNS
3303 39. Added log_selector=+queue_time_overall.
3305 40. When running the queue in the test harness, wait just a tad after forking a
3306 delivery process, to get repeatability of debugging output.
3308 41. Include certificate and key file names in error message when GnuTLS fails
3309 to set them up, because the GnuTLS error message doesn't include the name
3310 of the failing file when there is a problem reading it.
3312 42. Allow both -bf and -bF in the same test run.
3314 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3316 44. The "Exiscan patch" is now merged into the mainline Exim source.
3318 45. Sometimes the final signoff response after QUIT could fail to get
3319 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3320 < 0 before doing a fflush(). This bug looks as though it goes back to the
3321 introduction of TLS in release 3.20, but "sometimes" must have been rare
3322 because the tests only now provoked it.
3324 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3325 (this can affect the format of dates).
3327 47. exim_tidydb, when checking for the continued existence of a message for
3328 which it has found a message-specific retry record, was not finding
3329 messages that were in split spool directories. Consequently, it was
3330 deleting retry records that should have stayed in existence.
3332 48. Steve fixed some bugs in eximstats.
3334 49. The SPA authentication driver was not abandoning authentication and moving
3335 on to the next authenticator when an expansion was forced to fail,
3336 contradicting the general specification for all authenticators. Instead it
3337 was generating a temporary error. It now behaves as specified.
3339 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3340 (the order specifies the preference for clients). The order is now AES256,
3341 AES128, 3DES, ARCFOUR128.
3343 51. Small patch to Sieve code - explicitly set From: when generating an
3346 52. Exim crashed if a remote delivery caused a very long error message to be
3347 recorded - for instance if somebody sent an entire SpamAssassin report back
3348 as a large number of 550 error lines. This bug was coincidentally fixed by
3349 increasing the size of one of Exim's internal buffers (big_buffer) that
3350 happened as part of the Exiscan merge. However, to be on the safe side, I
3351 have made the code more robust (and fixed the comments that describe what
3354 53. Now that there can be additional text after "Completed" in log lines (if
3355 the queue_time_overall log selector is set), a one-byte patch to exigrep
3356 was needed to allow it to recognize "Completed" as not the last thing in
3359 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3360 patch that reportedly fixes this has been added. I am not expert enough to
3361 create a test for it. This is what the patch creator wrote:
3363 "I found a little strange behaviour of ldap code when working with
3364 Windows 2003 AD Domain, where users was placed in more than one
3365 Organization Units. When I tried to give exim partial DN, the exit code
3366 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3367 But simultaneously result of request was absolutely normal ldap result,
3368 so I produce this patch..."
3370 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3371 REFERENCE, so I have modified the code to exclude the patch when that macro
3374 55. Some experimental protocols are using DNS PTR records for new purposes. The
3375 keys for these records are domain names, not reversed IP addresses. The
3376 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3377 leaves it alone. Component reversal etc. now happens only for IP addresses.
3380 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3382 57. Double the size of the debug message buffer (to 2048) so that more of very
3383 long debug lines gets shown.
3385 58. The exicyclog utility now does better if the number of log files to keep
3386 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3388 59. Two changes related to the smtp_active_hostname option:
3390 (1) $smtp_active_hostname is now available as a variable.
3391 (2) The default for smtp_banner uses $smtp_active_hostname instead
3392 of $primary_hostname.
3394 60. The host_aton() function is supposed to be passed a string that is known
3395 to be a valid IP address. However, in the case of IPv6 addresses, it was
3396 not checking this. This is a hostage to fortune. Exim now panics and dies
3397 if the condition is not met. A case was found where this could be provoked
3398 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3399 components; fortuitously, this particular loophole had already been fixed
3400 by change 4.50/55 above.
3402 If there are any other similar loopholes, the new check in host_aton()
3403 itself should stop them being exploited. The report I received stated that
3404 data on the command line could provoke the exploit when Exim was running as
3405 exim, but did not say which command line option was involved. All I could
3406 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3407 running as the user.
3410 61. There was a buffer overflow vulnerability in the SPA authentication code
3411 (which came originally from the Samba project). I have added a test to the
3412 spa_base64_to_bits() function which I hope fixes it.
3415 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3416 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3418 63. The daemon start-up calls getloadavg() while still root for those OS that
3419 need the first call to be done as root, but it missed one case: when
3420 deliver_queue_load_max is set with deliver_drop_privilege. This is
3421 necessary for the benefit of the queue runner, because there is no re-exec
3422 when deliver_drop_privilege is set.
3424 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3425 This has been fixed.
3427 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3428 was in use, was not putting the data itself into the right store pool;
3429 consequently, it could be overwritten for a subsequent message in the same
3430 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3433 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3435 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3436 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3437 the function were treating the return as a boolean value, which happened to
3438 work because 0=false and not-0=true, but is not correct code.
3440 68. The host_aton() function was not handling scoped IPv6 addresses (those
3441 with, for example, "%eth0" on the end) correctly.
3443 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3444 negated items (that is, ~something) in unsigned ints. Some compilers
3445 apparently mutter when there is no cast.
3447 70. If an address verification called from an ACL failed, and did not produce a
3448 user-specific message (i.e. there was only a "system" message), nothing was
3449 put in $acl_verify_message. In this situation, it now puts the system
3452 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3453 session; change 4.31/43 added the unwanted input to the log line - except
3454 that it did not do this in the start of session case. It now does.
3456 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3457 This is wrong and can cause the other end to generate a synchronization
3458 error if it is another Exim or anything else that does the synchronization
3459 check. A QUIT command is no longer sent after a timeout.
3461 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3462 during host lookups.
3464 74. The defer_ok option of callout verification was not working if it was used
3465 when verifying addresses in header lines, that is, for this case:
3467 verify = header_sender/callout=defer_ok
3469 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3470 those file descriptors could be used for SMTP connections. If anything
3471 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3472 could be sent to the SMTP client, causing chaos. The daemon now opens
3473 stdin, stdout, and stderr to /dev/null when it puts itself into the
3476 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3477 log by default. The user can override this with suitable Perl magic.
3479 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3480 the log message when discard triggers, was not working for the DATA ACL or
3481 for the non-SMTP ACL.
3483 78. Error message wording change in sieve.c.
3485 79. If smtp_accept_max_per_host was set, the number of connections could be
3486 restricted to fewer than expected, because the daemon was trying to set up
3487 a new connection before checking whether the processes handling previous
3488 connections had finished. The check for completed processes is now done
3489 earlier. On busy systems, this bug wouldn't be noticed because something
3490 else would have woken the daemon, and it would have reaped the completed
3493 80. If a message was submitted locally by a user whose login name contained one
3494 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3495 It caused a spool format error. I have fixed the spool reading code. A
3496 related problem was that the "from" clause in the Received: line became
3497 illegal because of the space(s). It is now covered by ${quote_local_part.
3499 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3502 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3504 83. Patch from Sieve maintainer to fix unterminated string problem in
3505 "vacation" handling.
3507 84. Some minor changes to the Linux configuration files to help with other
3508 OS variants using glibc.
3510 85. One more patch for Sieve to update vacation handling to latest spec.
3513 ----------------------------------------------------
3514 See the note above about the 4.44 and 4.50 releases.
3515 ----------------------------------------------------
3521 1. Change 4.43/35 introduced a bug that caused file counts to be
3522 incorrectly computed when quota_filecount was set in an appendfile
3525 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3526 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3529 3. Give more explanation in the error message when the command for a transport
3530 filter fails to execute.
3532 4. There are several places where Exim runs a non-Exim command in a
3533 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3534 was being done only for the command run by the queryprogram router. It is
3535 now done for all such subprocesses. The other cases are: ${run, transport
3536 filters, and the commands run by the lmtp and pipe transports.
3538 5. Some older OS have a limit of 256 on the maximum number of file
3539 descriptors. Exim was using setrlimit() to set 1000 as a large value
3540 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3541 systems. I've change it so that if it can't get 1000, it tries for 256.
3543 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3544 was an oversight, and furthermore, ever since the addition of extra
3545 controls (e.g. 4.43/32), the checks on when to allow different forms of
3546 "control" were broken. There should now be diagnostics for all cases when a
3547 control that does not make sense is encountered.
3549 7. $recipients is now available in the predata ACL (oversight).
3551 8. Tidy the search cache before the fork to do a delivery from a message
3552 received from the command line. Otherwise the child will trigger a lookup
3553 failure and thereby defer the delivery if it tries to use (for example) a
3554 cached ldap connection that the parent has called unbind on.
3556 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3557 of $address_data from the recipient verification was clobbered by the
3558 sender verification.
3560 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3561 was its contents. (It was OK if the option was not defined at all.)
3563 11. A "Completed" log line is now written for messages that are removed from
3564 the spool by the -Mrm option.
3566 12. $host_address is now set to the target address during the checking of
3567 ignore_target_hosts.
3569 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3570 being passed; this would have caused $sender_host_name to have been used if
3571 matching the list had actually called for a host name (not very likely,
3572 since this list is usually IP addresses). A host name is now passed as
3575 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3576 code that specifies a non-restarting handler (typically sigaction() in
3577 modern systems) in an attempt to fix a rare and obscure crash bug.
3579 15. Narrowed the window for a race in the daemon that could cause it to ignore
3580 SIGCHLD signals. This is not a major problem, because they are used only to
3581 wake it up if nothing else does.
3583 16. A malformed maildirsize file could cause Exim to calculate negative values
3584 for the mailbox size or file count. Odd effects could occur as a result.
3585 The maildirsize information is now recalculated if the size or filecount
3588 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3589 support for a long time. Removed HAVE_SYS_VFS_H.
3591 18. Updated exipick to current release from John Jetmore.
3593 19. Allow an empty sender to be matched against a lookup in an address list.
3594 Previously the only cases considered were a regular expression, or an
3597 20. Exim went into a mad DNS lookup loop when doing a callout where the
3598 host was specified on the transport, if the DNS lookup yielded more than
3601 21. The RFC2047 encoding function was originally intended for short strings
3602 such as real names; it was not keeping to the 75-character limit for
3603 encoded words that the RFC imposes. It now respects the limit, and
3604 generates multiple encoded words if necessary. To be on the safe side, I
3605 have increased the buffer size for the ${rfc2047: expansion operator from
3608 22. Failure to deliver a bounce message always caused it to be frozen, even if
3609 there was an errors_to setting on the router. The errors_to setting is now
3612 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3613 canonical form (fully expanded) before being placed in
3614 $sender_host_address.
3616 24. Updated eximstats to version 1.33
3618 25. Include certificate and key file names in error message when GnuTLS fails
3619 to set them up, because the GnuTLS error message doesn't include the name
3620 of the failing file when there is a problem reading it.
3622 26. Expand error message when OpenSSL has problems setting up cert/key files.
3625 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3626 (this can affect the format of dates).
3628 28. exim_tidydb, when checking for the continued existence of a message for
3629 which it has found a message-specific retry record, was not finding
3630 messages that were in split spool directories. Consequently, it was
3631 deleting retry records that should have stayed in existence.
3633 29. eximstats updated to version 1.35
3634 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3635 1.35 - bugfix such that pie charts by volume are generated correctly
3637 30. The SPA authentication driver was not abandoning authentication and moving
3638 on to the next authenticator when an expansion was forced to fail,
3639 contradicting the general specification for all authenticators. Instead it
3640 was generating a temporary error. It now behaves as specified.
3642 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3643 (the order specifies the preference for clients). The order is now AES256,
3644 AES128, 3DES, ARCFOUR128.
3646 31. Small patch to Sieve code - explicitly set From: when generating an
3649 32. Exim crashed if a remote delivery caused a very long error message to be
3650 recorded - for instance if somebody sent an entire SpamAssassin report back
3651 as a large number of 550 error lines. This bug was coincidentally fixed by
3652 increasing the size of one of Exim's internal buffers (big_buffer) that
3653 happened as part of the Exiscan merge. However, to be on the safe side, I
3654 have made the code more robust (and fixed the comments that describe what
3657 33. Some experimental protocols are using DNS PTR records for new purposes. The
3658 keys for these records are domain names, not reversed IP addresses. The
3659 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3660 leaves it alone. Component reversal etc. now happens only for IP addresses.
3663 34. The host_aton() function is supposed to be passed a string that is known
3664 to be a valid IP address. However, in the case of IPv6 addresses, it was
3665 not checking this. This is a hostage to fortune. Exim now panics and dies
3666 if the condition is not met. A case was found where this could be provoked
3667 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3668 components; fortuitously, this particular loophole had already been fixed
3669 by change 4.50/55 or 4.44/33 above.
3671 If there are any other similar loopholes, the new check in host_aton()
3672 itself should stop them being exploited. The report I received stated that
3673 data on the command line could provoke the exploit when Exim was running as
3674 exim, but did not say which command line option was involved. All I could
3675 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3676 running as the user.
3679 35. There was a buffer overflow vulnerability in the SPA authentication code
3680 (which came originally from the Samba project). I have added a test to the
3681 spa_base64_to_bits() function which I hope fixes it.
3684 36. The daemon start-up calls getloadavg() while still root for those OS that
3685 need the first call to be done as root, but it missed one case: when
3686 deliver_queue_load_max is set with deliver_drop_privilege. This is
3687 necessary for the benefit of the queue runner, because there is no re-exec
3688 when deliver_drop_privilege is set.
3690 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3691 was in use, was not putting the data itself into the right store pool;
3692 consequently, it could be overwritten for a subsequent message in the same
3693 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3696 38. Sometimes the final signoff response after QUIT could fail to get
3697 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3698 < 0 before doing a fflush(). This bug looks as though it goes back to the
3699 introduction of TLS in release 3.20, but "sometimes" must have been rare
3700 because the tests only now provoked it.
3706 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3707 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3708 Now it returns an integer. A number of calls were still expecting a T/F
3709 return. Fortuitously, in all cases, the tests worked in OK situations,
3710 which is the norm. However, things would have gone wrong on any write
3711 failures on the smtp file descriptor. This function is used when sending
3712 messages over SMTP and also when doing verify callouts.
3714 2. When Exim is called to do synchronous delivery of a locally submitted
3715 message (the -odf or -odi options), it no longer closes stderr before doing
3718 3. Implemented the mua_wrapper option.
3720 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3722 5. Implemented the functions header_remove(), header_testname(),
3723 header_add_at_position(), and receive_remove_recipient(), and exported them
3726 6. If an ACL "warn" statement specified the addition of headers, Exim already
3727 inserted X-ACL-Warn: at the start if there was no header name. However, it
3728 was not making this test for the second and subsequent header lines if
3729 there were newlines in the string. This meant that an invalid header could
3730 be inserted if Exim was badly configured.
3732 7. Allow an ACL "warn" statement to add header lines at the start or after all
3733 the Received: headers, as well as at the end.
3735 8. Added the rcpt_4xx retry error code.
3737 9. Added postmaster_mailfrom=xxx to callout verification option.
3739 10. Added mailfrom=xxxx to the callout verification option, for verify=
3742 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3743 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3745 12. Inserted some casts to stop certain compilers warning when using pointer
3746 differences as field lengths or precisions in printf-type calls (mostly
3747 affecting debugging statements).
3749 13. Added optional readline() support for -be (dynamically loaded).
3751 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3752 same clock tick as a message's arrival, so that its received time was the
3753 same as the "first fail" time on the retry record, and that message
3754 remained on the queue past the ultimate address timeout, every queue runner
3755 would try a delivery (because it was past the ultimate address timeout) but
3756 after another failure, the ultimate address timeout, which should have then
3757 bounced the address, did not kick in. This was a "< instead of <=" error;
3758 in most cases the first failure would have been in the next clock tick
3759 after the received time, and all would be well.
3761 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3762 being recognized when the domain list was tested by the match_domain
3763 condition in an expansion string.
3765 16. Added the ${str2b64: operator.
3767 17. Exim was always calling setrlimit() to set a large limit for the number of
3768 processes, without checking whether the existing limit was already
3769 adequate. (It did check for the limit on file descriptors.) Furthermore,
3770 errors from getrlimit() and setrlimit() were being ignored. Now they are
3771 logged to the main and panic logs, but Exim does carry on, to try to do its
3772 job under whatever limits there are.
3774 18. Imported PCRE 5.0.
3776 19. Trivial typo in log message " temporarily refused connection" (the leading
3779 20. If the log selector return_path_on_delivery was set and an address was
3780 redirected to /dev/null, the delivery process crashed because it assumed
3781 that a return path would always be set for a "successful" delivery. In this
3782 case, the whole delivery is bypassed as an optimization, and therefore no
3785 21. Internal re-arrangement: the function for sending a challenge and reading
3786 a response while authentication was assuming a zero-terminated challenge
3787 string. It's now changed to take a pointer and a length, to allow for
3788 binary data in such strings.
3790 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3792 23. Exim was not respecting finduser_retries when seeking the login of the
3793 uid under which it was called; it was always trying 10 times. (The default
3794 setting of finduser_retries is zero.) Also, it was sleeping after the final
3795 failure, which is pointless.
3797 24. Implemented tls_on_connect_ports.
3799 25. Implemented acl_smtp_predata.
3801 26. If the domain in control=submission is set empty, Exim assumes that the
3802 authenticated id is a complete email address when it generates From: or
3803 Sender: header lines.
3805 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3806 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3807 chown and chgrp in /bin and hostname in /usr/bin.
3809 28. Exim was keeping the "process log" file open after each use, just as it
3810 does for the main log. This opens the possibility of it remaining open for
3811 long periods when the USR1 signal hits a daemon. Occasional processlog
3812 errors were reported, that could have been caused by this. Anyway, it seems
3813 much more sensible not to leave this file open at all, so that is what now
3816 29. The long-running daemon process does not normally write to the log once it
3817 has entered its main loop, and it closes the log before doing so. This is
3818 so that log files can straightforwardly be renamed and moved. However,
3819 there are a couple of unusual error situations where the daemon does write
3820 log entries, and I had neglected to close the log afterwards.
3822 30. The text of an SMTP error response that was received during a remote
3823 delivery was being truncated at 512 bytes. This is too short for some of
3824 the long messages that one sometimes sees. I've increased the limit to
3827 31. It is now possible to make retry rules that apply only when a message has a
3828 specific sender, in particular, an empty sender.
3830 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3831 it possible to be selective about when SMTP synchronization is enforced.
3833 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3835 32. Implemented hosts_connection_nolog.
3837 33. Added an ACL for QUIT.
3839 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3842 35. Added mailbox_size and mailbox_filecount to appendfile.
3844 36. Added control = no_multiline_responses to ACLs.
3846 37. There was a bug in the logic of the code that waits for the clock to tick
3847 in the case where the clock went backwards by a substantial amount such
3848 that the microsecond fraction of "now" was more than the microsecond
3849 fraction of "then" (but the whole seconds number was less).
3851 38. Added support for the libradius Radius client library this is found on
3852 FreeBSD (previously only the radiusclient library was supported).
3858 1. When certain lookups returned multiple values in the form name=value, the
3859 quoting of the values was not always being done properly. Specifically:
3860 (a) If the value started with a double quote, but contained no whitespace,
3862 (b) If the value contained whitespace other than a space character (i.e.
3863 tabs or newlines or carriage returns) it was not quoted.
3864 This fix has been applied to the mysql and pgsql lookups by writing a
3865 separate quoting function and calling it from the lookup code. The fix
3866 should probably also be applied to nisplus, ibase and oracle lookups, but
3867 since I cannot test any of those, I have not disturbed their existing code.
3869 2. A hit in the callout cache for a specific address caused a log line with no
3870 reason for rejecting RCPT. Now it says "Previous (cached) callout
3871 verification failure".
3873 3. There was an off-by-one bug in the queryprogram router. An over-long
3874 return line was truncated at 256 instead of 255 characters, thereby
3875 overflowing its buffer with the terminating zero. As well as fixing this, I
3876 have increased the buffer size to 1024 (and made a note to document this).
3878 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3879 when Exim is waiting for an SMTP response from a remote server, Exim
3880 restarts its select() call on the socket, thereby resetting its timeout.
3881 This is not a problem when such interrupts are rare. Somebody set up a cron
3882 job to run exiwhat every 2 minutes, which is less than the normal select()
3883 timeout (5 or 10 minutes). This meant that the select() timeout never
3884 kicked in because it was always reset. I have fixed this by comparing the
3885 time when an interrupt arrives with the time at the start of the first call
3886 to select(). If more time than the timeout has elapsed, the interrupt is
3887 treated as a timeout.
3889 5. Some internal re-factoring in preparation for the addition of Sieve
3890 extensions (by MH). In particular, the "personal" test is moved to a
3891 separate function, and given an option for scanning Cc: and Bcc: (which is
3892 not set for Exim filters).
3894 6. When Exim created an email address using the login of the caller as the
3895 local part (e.g. when creating a From: or Sender: header line), it was not
3896 quoting the local part when it contained special characters such as @.
3898 7. Installed new OpenBSD configuration files.
3900 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3901 try to make them clearer.
3903 9. Callout options, other than the timeout value, were being ignored when
3904 verifying sender addresses in header lines. For example, when using
3906 verify = header_sender/callout=no_cache
3908 the cache was (incorrectly) being used.
3910 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3911 only the Cygwin environment.
3913 11. When return_path_on_delivery was set as a log selector, if different remote
3914 addresses in the same message used different return paths and parallel
3915 remote delivery occurred, the wrong values would sometimes be logged.
3916 (Whenever a remote delivery process finished, the return path value from
3917 the most recently started remote delivery process was logged.)
3919 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3920 lines when AUTH and/or TLS are in use. This is the "received protocol"
3921 field. Exim used to use "asmtp" for authenticated SMTP, without any
3922 indication (in the protocol name) for TLS use. Now it follows the RFC and
3923 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3924 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3925 names appear in log lines as well as in Received: header lines.
3927 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3928 extensions, and comparison tests, and to fix some bugs.
3930 14. Changes to the "personal" filter test:
3932 (1) The test was buggy in that it was just doing the equivalent of
3933 "contains" tests on header lines. For example, if a user's address was
3934 anne@some.where, the "personal" test would incorrectly be true for
3936 To: susanne@some.where
3938 This test is now done by extracting each address from the header in turn,
3939 and checking the entire address. Other tests that are part of "personal"
3940 are now done using regular expressions (for example, to check local parts
3941 of addresses in From: header lines).
3943 (2) The list of non-personal local parts in From: addresses has been
3944 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3945 taken from the Sieve specification recommendations.
3947 (3) If the message contains any header line starting with "List-" it is
3948 treated as non-personal.
3950 (4) The test for "circular" in the Subject: header line has been removed
3951 because it now seems ill-conceived.
3953 15. Minor typos in src/EDITME comments corrected.
3955 16. Installed latest exipick from John Jetmore.
3957 17. If headers_add on a router specified a text string that was too long for
3958 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3959 of string_sprintf() is now avoided.
3961 18. $message_body_size was not set (it was always zero) when running the DATA
3962 ACL and the local_scan() function.
3964 19. For the "mail" command in an Exim filter, no default was being set for
3965 the once_repeat time, causing a random time value to be used if "once" was
3966 specified. (If the value happened to be <= 0, no repeat happened.) The
3967 default is now 0s, meaning "never repeat". The "vacation" command was OK
3968 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3969 (I found it when inspecting the code).
3971 20. There is now an overall timeout for performing a callout verification. It
3972 defaults to 4 times the callout timeout, which applies to individual SMTP
3973 commands during the callout. The overall timeout applies when there is more
3974 than one host that can be tried. The timeout is checked before trying the
3975 next host. This prevents very long delays if there are a large number of
3976 hosts and all are timing out (e.g. when the network connections are timing
3977 out). The value of the overall timeout can be changed by specifying an
3978 additional sub-option for "callout", called "maxwait". For example:
3980 verify = sender/callout=5s,maxwait=20s
3982 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3983 to the end before writing, but this should make it even safer).
3985 22. Exim was forgetting that it had advertised PIPELINING for the second and
3986 subsequent messages on an SMTP connection. It was also not resetting its
3987 memory on STARTTLS and an internal HELO.
3989 23. When Exim logs an SMTP synchronization error within a session, it now
3990 records whether PIPELINING has been advertised or not.
3992 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3993 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3994 rather than long int.
3996 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3998 26. Added the never_mail option to autoreply.
4004 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4005 crash if the getsockname() call failed; this can happen if a connection is
4006 closed very soon after it is established. The problem was simply in the
4007 order in which certain operations were done, causing Exim to try to write
4008 to the SMTP stream before it had set up the file descriptor. The bug has
4009 been fixed by making things happen in the correct order.
4015 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4016 before the connection was closed, thus losing the rejection response.
4018 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4019 some early Solaris releases, but causes trouble in current releases where
4020 socklen_t is defined.
4022 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4025 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4028 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4029 that releasing the top of it at the end releases what was used for sub-
4030 expansions (unless the block got too big). However, discard this block if
4031 the first thing is a variable or header, so that we can use its block when
4032 it is dynamic (useful for very large $message_headers, for example).
4034 6. Lookups now cache *every* query, not just the most recent. A new, separate
4035 store pool is used for this. It can be recovered when all lookup caches are
4036 flushed. Lookups now release memory at the end of their result strings.
4037 This has involved some general refactoring of the lookup sources.
4039 7. Some code has been added to the store_xxx() functions to reduce the amount
4040 of flapping under certain conditions.
4042 8. log_incoming_interface used to affect only the <= reception log lines. Now
4043 it causes the local interface and port to be added to several more SMTP log
4044 lines, for example "SMTP connection from", and rejection lines.
4046 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4048 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4050 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4051 could be overwritten at the end of the current message (or the start of a
4052 new message if it was set in a HELO ACL). The value is now preserved for
4053 the duration of the SMTP connection.
4055 12. If a transport had a headers_rewrite setting, and a matching header line
4056 contained an unqualified address, that address was qualified, even if it
4057 did not match any rewriting rules. The underlying bug was that the values
4058 of the flags that permit the existence of unqualified sender and recipient
4059 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4060 non-local messages, and by -bnq for local messages) were not being
4061 preserved with the message after it was received.
4063 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4064 "next input=" as part of the text comprising the host identity instead of
4065 the correct text. The code was using the same buffer for two different
4066 strings. However, depending on which order the printing function evaluated
4067 its arguments, the bug did not always show up. Under Linux, for example, my
4068 test suite worked just fine.
4070 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4071 doesn't work with some older versions of Perl. It has been changed to "my",
4072 which in any case is probably the better facility to use.
4074 15. A really picky compiler found some instances of statements for creating
4075 error messages that either had too many or two few arguments for the format
4078 16. The size of the buffer for calls to the DNS resolver has been increased
4079 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4080 for addresses that have a lot of PTR records. This alleviates a problem; it
4081 does not fully solve it.
4083 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4084 buffer now truncates the list and logs the incident, which is the same
4085 action as happens when Exim is looking up a host name and its aliases.
4086 Previously in this situation something unpredictable would happen;
4087 sometimes it was "internal error: store_reset failed".
4089 18. If a server dropped the connection unexpectedly when an Exim client was
4090 using GnuTLS and trying to read a response, the client delivery process
4091 crashed while trying to generate an error log message.
4093 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4094 string, for example:
4096 warn message = H1: something\nH2: something
4098 the text was added as a single header line from Exim's point of view
4099 though it ended up OK in the delivered message. However, searching for the
4100 second and subsequent header lines using $h_h2: did not work. This has been
4101 fixed. Similarly, if a system filter added multiple headers in this way,
4102 the routers could not see them.
4104 20. Expanded the error message when iplsearch is called with an invalid key to
4105 suggest using net-iplsearch in a host list.
4107 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4108 ACLs are no longer actually imposed (and a message to that effect is
4111 22. If a "gecos" field in a passwd entry contained escaped characters, in
4112 particular, if it contained a \" sequence, Exim got it wrong when building
4113 a From: or a Sender: header from that name. A second bug also caused
4114 incorrect handling when an unquoted " was present following a character
4115 that needed quoting.
4117 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4118 was not being matched caselessly.
4120 24. Arranged for all hyphens in the exim.8 source to be escaped with
4123 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4124 properly. Recipient callout cache records were still being keyed to include
4125 the sender, even when use_sender was set false. This led to far more
4126 callouts that were necessary. The sender is no longer included in the key
4127 when use_sender is false.
4129 26. Added "control = submission" modifier to ACLs.
4131 27. Added the ${base62d: operator to decode base 62 numbers.
4133 28. dnsdb lookups can now access SRV records.
4135 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4136 the configuration file.
4138 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4139 (-v) mode. This makes the output for a verbose queue run more intelligible.
4141 31. Added a use_postmaster feature to recipient callouts.
4143 32. Added the $body_zerocount variable, containing the number of binary zero
4144 bytes in the message body.
4146 33. The time of last modification of the "new" subdirectory is now used as the
4147 "mailbox time last read" when there is a quota error for a maildir
4150 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4152 35. Added +ignore_unknown as a special item in host lists.
4154 36. Code for decoding IPv6 addresses in host lists is now included, even if
4155 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4156 address was recognized as an IP address, but was then not correctly decoded
4157 into binary, causing unexpected and incorrect effects when compared with
4164 1. Very minor rewording of debugging text in manualroute to say "list of
4165 hosts" instead of "hostlist".
4167 2. If verify=header_syntax was set, and a header line with an unqualified
4168 address (no domain) and a large number of spaces between the end of the
4169 name and the colon was received, the reception process suffered a buffer
4170 overflow, and (when I tested it) crashed. This was caused by some obsolete
4171 code that should have been removed. The fix is to remove it!
4173 3. When running in the test harness, delay a bit after writing a bounce
4174 message to get a bit more predictability in the log output.
4176 4. Added a call to search_tidyup() just before forking a reception process. In
4177 theory, someone could use a lookup in the expansion of smtp_accept_max_
4178 per_host which, without the tidyup, could leave open a database connection.
4180 5. Added the variables $recipient_data and $sender_data which get set from a
4181 lookup success in an ACL "recipients" or "senders" condition, or a router
4182 "senders" option, similar to $domain_data and $local_part_data.
4184 6. Moved the writing of debug_print from before to after the "senders" test
4187 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4188 problems for message scanning, either using a data ACL, or using
4189 local_scan() because the Received: header was not generated till after they
4190 were called (in order to set the time as the time of reception completion).
4191 I have revised the way this works. The header is now generated after the
4192 body is received, but before the ACL or local_scan() are called. After they
4193 are run, the timestamp in the header is updated.
4199 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4200 before starting a queue runner without re-exec. This happened only when
4201 deliver_drop_privilege was set or when the Exim user was set to root. The
4202 effect of the bug was that timeouts during subsequent deliveries caused
4203 crashes instead of being properly handled. The handler is now left at its
4204 default (and expected) setting.
4206 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4207 message, again when deliver_drop_privilege is set or Exim is run as root.
4208 The bug described in (1) was not present in this case, but the tidying up
4209 of the other signals was missing. I have made the two cases consistent.
4211 3. The ignore_target_hosts setting on a manualroute router was being ignored
4212 for hosts that were looked up using the /MX notation.
4214 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4217 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4218 operated on the sender address. After changing the $sender_address to <>
4219 for the sender address verify, Exim was re-instated it as the original
4220 (before rewriting) address, but remembering that it had rewritten it, so it
4221 wasn't rewriting it again. This bug also had the effect of breaking the
4222 sender address verification caching when the sender address was rewritten.
4224 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4225 This has been changed so that if the ip literal address matches
4226 ignore_target_hosts, the router declines.
4228 7. Added expansion conditions match_domain, match_address, and match_local_
4229 part (NOT match_host).
4231 8. The placeholder for the Received: header didn't have a length field set.
4233 9. Added code to Exim itself and to exim_lock to test for a specific race
4234 condition that could lead to file corruption when using MBX delivery. The
4235 issue is with the lockfile that is created in /tmp. If this file is removed
4236 after a process has opened it but before that process has acquired a lock,
4237 there is the potential for a second process to recreate the file and also
4238 acquire a lock. This could lead to two Exim processes writing to the file
4239 at the same time. The added code performs the same test as UW imapd; it
4240 checks after acquiring the lock that its file descriptor still refers to
4241 the same named file.
4243 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4244 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4247 11. Added the smtp_active_hostname option. If used, this will typically be made
4248 to depend on the incoming interface address. Because $interface_address is
4249 not set up until the daemon has forked a reception process, error responses
4250 that can happen earlier (such as "too many connections") no longer contain
4253 12. If an expansion in a condition on a "warn" statement fails because a lookup
4254 defers, the "warn" statement is abandoned, and the next ACL statement is
4255 processed. Previously this caused the whole ACL to be aborted.
4257 13. Added the iplsearch lookup type.
4259 14. Added ident_timeout as a log selector.
4261 15. Added tls_certificate_verified as a log selector.
4263 16. Added a global option tls_require_ciphers (compare the smtp transport
4264 option of the same name). This controls incoming TLS connections.
4266 17. I finally figured out how to make tls_require_ciphers do a similar thing
4267 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4268 before starting the TLS session.
4270 18. Tabs are now shown as \t in -bP output.
4272 19. If the log selector return_path_on_delivery was set, Exim crashed when
4273 bouncing a message because it had too many Received: header lines.
4275 20. If two routers both had headers_remove settings, and the first one included
4276 a superfluous trailing colon, the final name in the first list and the
4277 first name in the second list were incorrectly joined into one item (with a
4278 colon in the middle).
4284 1. Added -C and -D options to the exinext utility, mainly to make it easier
4285 to include in the automated testing, but these could be helpful when
4286 multiple configurations are in use.
4288 2. The exinext utility was not formatting the output nicely when there was
4289 an alternate port involved in the retry record key, nor when there was a
4290 message id as well (for retries that were specific to a specific message
4291 and a specific host). It was also confused by IPv6 addresses, because of
4292 the additional colons they contain. I have fixed the IPv4 problem, and
4293 patched it up to do a reasonable job for IPv6.
4295 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4296 delivery, the log line now contains "pipelined" if PIPELINING was used.
4298 4. An SMTP transport process used to panic and die if the bind() call to set
4299 an explicit outgoing interface failed. This has been changed; it is now
4300 treated in the same way as a connect() failure.
4302 5. A reference to $sender_host_name in the part of a conditional expansion
4303 that was being skipped was still causing a DNS lookup. This no longer
4306 6. The def: expansion condition was not recognizing references to header lines
4307 that used bh_ and bheader_.
4309 7. Added the _cache feature to named lists.
4311 8. The code for checking quota_filecount in the appendfile transport was
4312 allowing one more file than it should have been.
4314 9. For compatibility with Sendmail, the command line option
4322 and sets the incoming protocol and host name (for trusted callers). The
4323 host name and its colon can be omitted when only the protocol is to be set.
4324 Note the Exim already has two private options, -pd and -ps, that refer to
4325 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4326 "s", but I don't think that's a major issue.
4328 10. A number of refactoring changes to the code, none of which should affect
4331 (a) The number of logging options was getting close to filling up the
4332 32-bit word that was used as a bit map. I have split them into two classes:
4333 those that are passed in the argument to log_write(), and those that are
4334 only ever tested independently outside of that function. These are now in
4335 separate 32-bit words, so there is plenty of room for expansion again.
4336 There is no change in the user interface or the logging behaviour.
4338 (b) When building, for example, log lines, the code previously used a
4339 macro that called string_cat() twice, in order to add two strings. This is
4340 not really sufficiently general. Furthermore, there was one instance where
4341 it was actually wrong because one of the argument was used twice, and in
4342 one call a function was used. (As it happened, calling the function twice
4343 did not affect the overall behaviour.) The macro has been replaced by a
4344 function that can join an arbitrary number of extra strings onto a growing
4347 (c) The code for expansion conditions now uses a table and a binary chop
4348 instead of a serial search (which was left over from when there were very
4349 few conditions). Also, it now recognizes conditions like "pam" even when
4350 the relevant support is not compiled in: a suitably worded error message is
4351 given if an attempt is made to use such a condition.
4353 11. Added ${time_interval:xxxxx}.
4355 12. A bug was causing one of the ddress fields not to be passed back correctly
4356 from remote delivery subprocesses. The field in question was not being
4357 subsequently used, so this caused to problems in practice.
4359 13. Added new log selectors queue_time and deliver_time.
4361 14. Might have fixed a bug in maildirsizefile handling that threw up
4362 "unexpected character" debug warnings, and recalculated the data
4363 unnecessarily. In any case, I expanded the warning message to give more
4366 15. Added the message "Restricted characters in address" to the statements in
4367 the default ACL that block characters like @ and % in local parts.
4369 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4370 Three changes have been made:
4372 (a) There was a serious bug; a negative response to MAIL caused the whole
4373 recipient domain to be cached as invalid, thereby blocking all messages
4374 to all local parts at the same domain, from all senders. This bug has
4375 been fixed. The domain is no longer cached after a negative response to
4376 MAIL if the sender used is not empty.
4378 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4381 (c) A new callout option, "use_sender" has been added for people who want
4382 the modified behaviour.
4388 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4391 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4392 indeed breaks things for older releases.
4394 3. Added additional logging to the case where there is a problem reading data
4395 from a filter that is running in a subprocess using a pipe, in order to
4396 try to track down a specific problem.
4398 4. Testing facility fudge: when running in the test harness and attempting
4399 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4400 getting "No route to host". Convert this to a timeout.
4402 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4405 6. Some OS don't have socklen_t but use size_t instead. This affects the
4406 fifth argument of getsockopt() amongst other things. This is now
4407 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4408 can be set for individual OS. I have set it for SunOS5, OSF1, and
4409 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4410 some earlier ones do not.
4412 7. Change 4.30/15 was not doing the test caselessly.
4414 8. The standard form for an IPv6 address literal was being rejected by address
4415 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4416 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4417 this, as well as the form without the "IPv6" on the front (but only when
4418 address literals are enabled, of course).
4420 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4422 10. Exim crashed if a message with an empty sender address specified by -f
4423 encountered a router with an errors_to setting. This could be provoked only
4424 by a command such as
4428 where an empty string was supplied; "<>" did not hit this bug.
4430 11. Installed PCRE release 4.5.
4432 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4433 remained set. It is now erased.
4435 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4436 times from message ids (which are base 36 rather than the normal 62).
4438 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4439 were being counted as actual protocol errors, and logged if the log
4440 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4441 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4442 and DATA following a set of rejected RCPTs do not count as protocol errors.
4443 In other words, Exim assumes they were pipelined, though this may not
4444 actually be the case. Of course, in all cases the client gets an
4445 appropriate error code.
4447 15. If a lookup fails in an ACL condition, a message about the failure may
4448 be available; it is used if testing the ACL cannot continue, because most
4449 such messages specify what the cause of the deferral is. However, some
4450 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4451 that caused an old message to be retained and used if a later statement
4452 caused a defer, replacing the real cause of the deferral.
4454 16. If an IP address had so many PTR records that the DNS lookup buffer
4455 was not large enough to hold them, Exim could crash while trying to process
4456 the truncated data. It now detects and logs this case.
4458 17. Further to 4.21/58, another change has been made: if (and only if) the
4459 first line of a message (the first header line) ends with CRLF, a bare LF
4460 in a subsequent header line has a space inserted after it, so as not to
4461 terminate the header.
4463 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4464 unnecessarily, used atoi() instead of strtol(), and didn't check the
4465 termination when getting file sizes from file names by regex.
4467 19. Completely re-implemented the support for maildirsize files, in the light
4468 of a number of problems with the previous contributed implementation
4469 (4.30/29). In particular:
4471 . If the quota is zero, the maildirsize file is maintained, but no quota is
4474 . If the maildir directory does not exist, it is created before any attempt
4475 to write a maildirsize file.
4477 . The quota value in the file is just a cache; if the quota is changed in
4478 the transport, the new value overrides.
4480 . A regular expression is available for excluding directories from the
4483 20. The autoreply transport checks the characters in options that define the
4484 message's headers; it allows continued headers, but it was checking with
4485 isspace() after an embedded newline instead of explicitly looking for a
4488 21. If all the "regular" hosts to which an address was routed had passed their
4489 expiry times, and had not reached their retry times, the address was
4490 bounced, even if fallback hosts were defined. Now Exim should go on to try
4493 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4494 equivalent code in the SMTP transport. Some hosts send humungous responses
4495 to HELO/EHLO, more than 1024 it seems.
4497 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4498 gives compiler warnings in some environments. I've now done it "properly",
4501 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4502 (because of problems with the built-in one) was declared to return uschar *
4503 instead of char *, causing compiler failure.
4505 25. Fixed a file descriptor leak when processing alias/forward files.
4507 26. Fixed a minor format string issue in dbfn.c.
4509 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4511 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4512 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4515 29. When a sender address is verified, it is cached, to save repeating the test
4516 when there is more than one recipient in a message. However, when the
4517 verification involves a callout, it is possible for different callout
4518 options to be set for different recipients. It is too complicated to keep
4519 track of this in the cache, so now Exim always runs a verification when a
4520 callout is required, relying on the callout cache for the optimization.
4521 The overhead is duplication of the address routing, but this should not be
4524 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4525 to be verified with callout=postmaster, and the main callout worked but the
4526 postmaster check failed, the verification correctly failed. However, if a
4527 subsequent RCPT command asked for sender verification *without* the
4528 postmaster check, incorrect caching caused this verification also to fail,
4531 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4532 it was not caching the DNS options (qualify_single, search_parents) that
4533 were used when the lookup failed. A subsequent lookup with different
4534 options therefore always gave the same answer, though there were cases
4535 where it should not have. (Example: a "domains = !$mx_any" option on a
4536 dnslookup router: the "domains" option is always processed without any
4537 widening, but the router might have qualify_single set.) Now Exim uses the
4538 cached value only when the same options are set.
4540 32. Added John Jetmore's "exipick" utility to the distribution.
4542 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4543 than a timeout (e.g. a certificate is required, and is not provided), an
4544 Exim server now closes the connection immediately. Previously it waited for
4545 the client to close - but if the client is SSL, it seems that they each
4546 wait for each other, leading to a delay before one of them times out.
4548 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4549 maintained 0.8.x compatibility because I don't think many are using it, and
4550 it is clearly obsolete.
4552 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4555 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4556 client certificate was expired. A simple patch fixes this, though I don't
4557 understand the full logic of why the verify callback is called multiple
4560 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4561 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4562 which causes problems with some clients (such as the Certicom SSL Plus
4563 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4564 disables the coutermeasure allowing Eudora to connect."
4566 38. Exim was not checking that a write() to a log file succeeded. This could
4567 lead to Bad Things if a log got too big, in particular if it hit a file
4568 size limit. Exim now panics and dies if it cannot write to a log file, just
4569 as it does if it cannot open a log file.
4571 39. Modified OS/Makefile-Linux so that it now contains
4573 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4575 The two -D definitions ensure that Exim is compiled with large file
4576 support, which makes it possible to handle log files that are bigger than
4579 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4580 instance) a domain was checked against a named list that involved a lookup,
4581 causing $domain_data to be set, then another domain was checked against the
4582 same list, then the first domain was re-checked, the value of $domain_data
4583 after the final check could be wrong. In particular, if the second check
4584 failed, it could be set empty. This bug probably also applied to
4587 41. The strip_trailing_dot option was not being applied to the address given
4588 with the -f command-line option.
4590 42. The code for reading a message's header from the spool was incrementing
4591 $received_count, but never initializing it. This meant that the value was
4592 incorrect (doubled) while delivering a message in the same process in which
4593 it was received. In the most common configuration of Exim, this never
4594 happens - a fresh exec is done - but it can happen when
4595 deliver_drop_privilege is set.
4597 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4598 it now includes up to 150 characters of the unexpected data in the log
4601 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4602 and building data strings. The size of both of these buffers was 10 000
4603 bytes - far larger than anybody would *ever* want, thought I. Needless to
4604 say, somebody hit the limit. I have increased the maximum line length to
4605 20 000 and the maximum data length of concatenated lines to 100 000. I have
4606 also fixed two bugs, because there was no checking on these buffers. Tsk,
4607 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4608 buffer is too small.
4610 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4611 lsearch lookups. Now it does.
4613 46. When parsing a route_list item in a manualroute router, a fixed-length
4614 buffer was used for the list of hosts. I made this 1024 bytes long,
4615 thinking that nobody would ever have a list of hosts that long. Wrong.
4616 Somebody had a whole pile of complicated expansion conditions, and the
4617 string was silently truncated, leading to an expansion error. It turns out
4618 that it is easier to change to an unlimited length (owing to other changes
4619 that have happened since this code was originally written) than to build
4620 structure for giving a limitation error. The length of the item that
4621 expands into the list of hosts is now unlimited.
4623 47. The lsearch lookup could not handle data where the length of text line was
4624 more than 4095 characters. Such lines were truncated, leading to shortened
4625 data being returned. It should now handle lines of any length.
4627 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4628 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4631 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4632 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4633 no longer happens in comment lines. A list of replacements is now placed
4634 at the head of all of the source files, except those whose only change is
4635 to replace PERL_COMMAND in the very first #! line.
4637 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4638 messages on the queue, with a bottom-up merge sort, using code contributed
4639 by Michael Haardt. This should make operations like -bp somewhat faster on
4640 large queues. It won't affect queue runners, except when queue_run_in_order
4643 51. Installed eximstats 1.31 in the distribution.
4645 52. Added support for SRV lookups to the dnslookup router.
4647 53. If an ACL referred to $message_body or $message_body_end, the value was not
4648 reset for any messages that followed in the same SMTP session.
4650 54. The store-handling optimization for building very long strings was not
4651 differentiating between the different store pools. I don't think this
4652 actually made any difference in practice, but I've tidied it.
4654 55. While running the routers to verify a sender address, $sender_address
4655 was still set to the sender address. This is wrong, because when routing to
4656 send a bounce to the sender, it would be empty. Therefore, I have changed
4657 it so that, while verifying a sender address, $sender_address is set to <>.
4658 (There is no change to what happens when verifying a recipient address.)
4660 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4661 target A or AAAA records (if not already returned) without resetting the
4662 qualify_single or search_parents options of the DNS resolver. These are
4663 inappropriate in this case because the targets of MX and SRV records must
4664 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4665 target that, when qualified, matched something in the local domain. These
4666 two options are now turned off when doing these lookups.
4668 57. It seems that at least some releases of Reiserfs (which does not have the
4669 concept of a fixed number of inodes) returns zero and not -1 for the
4670 number of available inodes. This interacted badly with check_spool_inodes,
4671 which assumed that -1 was the "no such thing" setting. What I have done is
4672 to check that the total number of inodes is greater than zero before doing
4673 the test of how many are available.
4675 58. When a "warn" ACL statement has a log_message modifier, the message is
4676 remembered, and not repeated. This is to avoid a lot of repetition when a
4677 message has many recipients that cause the same warning to be written.
4678 Howewer, Exim was preserving the list of already written lines for an
4679 entire SMTP session, which doesn't seem right. The memory is now reset if a
4680 new message is started.
4682 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4683 debugging output unless log_rewrite was also set.
4685 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4686 of (int)(handle) when we know that handle contains (void *)(-1).
4688 61. The Exim daemon panic-logs an error return when it closes the incoming
4689 connection. However "connection reset by peer" seems to be common, and
4690 isn't really an error worthy of noting specially, so that particular error
4693 62. When Exim is trying to find all the local interfaces, it used to panic and
4694 die if the ioctl to get the interface flags failed. However, it seems that
4695 on at least one OS (Solaris 9) it is possible to have an interface that is
4696 included in the list of interfaces, but for which you get a failure error
4697 for this call. This happens when the interface is not "plumbed" into a
4698 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4699 failure of the "get flags" call assumes that the interface is down.
4701 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4702 makes life easier for people who are doing arithmetic on fields extracted
4703 from dates, where you often get leading zeros that should not be
4704 interpreted as octal.
4706 64. Added qualify_domain to the redirect router, to override the global
4709 65. If a pathologically long header line contained very many addresses (the
4710 report of this problem mentioned 10 000) and each of them was rewritten,
4711 Exim could use up a very large amount of memory. (It kept on making new
4712 copies of the header line as it rewrote, and never released the old ones.)
4713 At the expense of a bit more processing, the header rewriting function has
4714 been changed so that it no longer eats memory in this way.
4716 66. The generation of the Received: header has been moved from the time that a
4717 message starts to be received, to the time that it finishes. The timestamp
4718 in the Received: header should now be very close to that of the <= log
4719 line. There are two side-effects of this change:
4721 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4722 logged header lines no longer include the local Received: line, because
4723 it has not yet been created. The same applies to a copy of the message
4724 that is returned to a non-SMTP sender when a message is rejected.
4726 (b) When a filter file is tested using -bf, no additional Received: header
4727 is added to the test message. After some thought, I decided that this
4730 This change does not affect the value of $received_for. It is still set
4731 after address rewriting, but before local_scan() is called.
4733 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4735 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4736 gave an unhelpful panic error message, and a defer error. I have managed to
4737 change this behaviour so that it now rejects any supplied certificate,
4738 which seems right, as the list of acceptable certificates is empty.
4740 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4741 gave an unhelpful defer error. I have not managed to make this reject any
4742 supplied certificates, but the error message it gives is "no certificate
4743 supplied", which is not helpful.
4745 70. exigrep's output now also includes lines that are not associated with any
4746 message, but which match the given pattern. Implemented by a patch from
4747 Martin Sluka, which also tidied up the Perl a bit.
4749 71. Recipient callout verification, like sender verification, was using <> in
4750 the MAIL FROM command. This isn't really the right thing, since the actual
4751 sender may affect whether the remote host accepts the recipient or not. I
4752 have changed it to use the actual sender in the callout; this means that
4753 the cache record is now keyed on a recipient/sender pair, not just the
4754 recipient address. There doesn't seem to be a real danger of callout loops,
4755 since a callout by the remote host to check the sender would use <>.
4756 [SEE ABOVE: changed after hitting problems.]
4758 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4759 temporary errors. However, in the case of such a code being given after
4760 the end of a data transmission (i.e. after ".") Exim was failing to write
4761 a retry record for the message. (Yes, there was some broken host that was
4762 actually sending 8xx at this point.)
4764 73. An unknown lookup type in a host list could cause Exim to panic-die when
4765 the list was checked. (An example that provoked this was putting <; in the
4766 middle of a list instead of at the start.) If this happened during a DATA
4767 ACL check, a -D file could be left lying around. This kind of configuration
4768 error no longer causes Exim to die; instead it causes a defer errror. The
4769 incident is still logged to the main and panic logs.
4771 74. Buglet left over from Exim 3 conversion. The message "too many messages
4772 in one connection" was written to the rejectlog but not the mainlog, except
4773 when address rewriting (yes!) was being logged.
4775 75. Added write_rejectlog option.
4777 76. When a system filter was run not as root (that is, when system_filter_user
4778 was set), the values of the $n variables were not being returned to the
4779 main process; thus, they were not subsequently available in the $sn
4782 77. Added +return_path_on_delivery log selector.
4784 78. A connection timeout was being treated differently from recipients deferred
4785 when testing hosts_max_try with a message that was older than the host's
4786 retry timeout. (The host should not be counted, thus allowing all hosts to
4787 be tried at least once before bouncing.) This may have been the cause of an
4788 occasionally reported bug whereby a message would remain on the queue
4789 longer than the retry timeout, but would be bounced if a delivery was
4790 forced. I say "may" because I never totally pinned down the problem;
4791 setting up timeout/retry tests is difficult. See also the next item.
4793 79. The ultimate address timeout was not being applied to errors that involved
4794 a combination of host plus message (for example, a timeout on a MAIL
4795 command). When an address resolved to a number of possible hosts, and they
4796 were not all tried for each delivery (e.g. because of hosts_max_try), a
4797 message could remain on the queue longer than the retry timeout.
4799 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4802 81. Fixed an obscure SMTP outgoing bug which required at least the following
4803 conditions: (a) there was another message waiting for the same server;
4804 (b) the server returned 5xx to all RCPT commands in the first message so
4805 that the message was not completed; (c) the server dropped the connection
4806 or gave a negative response to the RSET that Exim sends to abort the
4807 transaction. The observed case was a dropped connection after DATA that had
4808 been sent in pipelining mode. That is, the server had advertised PIPELINING
4809 but was not implementing it correctly. The effect of the bug was incorrect
4810 behaviour, such as trying another host, and this could lead to a crash.
4816 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4817 and daemon.c were passed as pointers to ints; they should have been
4818 pointers to socklen_t variables (which are typically unsigned ints).
4820 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4823 3. Fixed a really odd bug that affected only the testing scheme; patching a
4824 certain fixed string in the binary changed the value of another string that
4825 happened to be identical to the end of the original first string.
4827 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4828 name", it returns that address as the IP address. On some operating
4829 systems (e.g. Solaris), it also passes back the IP address string as the
4830 "host name". However, on others (e.g. Linux), it passes back an empty
4831 string. Exim wasn't checking for this, and was changing the host name to an
4832 empty string, assuming it had been canonicized.
4834 5. Although rare, it is permitted to have more than one PTR record for a given
4835 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4836 all the names associated with an address, because they do in Solaris.
4837 However, it seems that they do not in Linux for data that comes from the
4838 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4839 I found this out when I moved to a new Linux workstation and tried to run
4840 the Exim test suite.
4842 To get round this problem I have changed the code so that it now does its
4843 own call to the DNS to look up PTR records when searching for a host name.
4844 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4845 addresses that are only in /etc/hosts are still found.
4847 This behaviour is, however, controlled by an option called host_lookup_
4848 order, which defaults to "bydns:byaddr". If people want to use the other
4849 order, or indeed, just use one or the other means of lookup, they can
4850 specify it in this variable.
4852 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4853 some operating systems, this comes back from gethostbyaddr() as an empty
4854 string, and this is what Exim used to test for. However, it seems that in
4855 other systems, "." is yielded. Exim now tests for this case too.
4857 7. The values of check_spool_space and check_log_space are now held internally
4858 as a number of kilobytes instead of an absolute number of bytes. If a
4859 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4860 kilobyte. This means that much larger values can be stored.
4862 8. Exim monitor: an attempt to get the action menu when not actually pointing
4863 at a message produces an empty menu entitled "No message selected". This
4864 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4865 no entries in it ("Shell widget menu has zero width and/or height"). So I
4866 have added a single, blank menu entry in this case.
4868 9. Added ${quote_local_part.
4870 10. MIME decoding is now applied to the contents of Subject: header lines when
4873 11. Now that a reference to $sender_host_address automatically causes a reverse
4874 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4875 host lookup before query-style lookups in lists that might use this
4876 variable. This has therefore been abolished, and the "net-" prefix is no
4877 longer necessary for query-style lookups.
4879 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4880 have been a typo for LFLAGS, so it has been changed.
4882 13. The install script calls Exim with "-C /dev/null" in order to find the
4883 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4884 to be output. Howeve, since Exim outputs its version number before the
4885 error, it didn't break the script. It just looked ugly. I fixed this by
4886 always allowing "-C /dev/null" if the caller is root.
4888 14. Ignore overlarge ACL variable number when reading spool file - insurance
4889 against a later release with more variables having written the file.
4891 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4892 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4893 this, as well as the form without the "IPv6" on the front.
4895 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4896 OS/Makefile-Darwin file.
4898 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4899 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4901 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4902 variable. However, some OS use ints for inodes. I've added cast to long int
4903 to get rid of the compiler warning.
4905 19. I had forgotten to lock out "/../" in configuration file names when
4906 ALT_CONFIG_PREFIX was set.
4908 20. Routers used for verification do not need to specify transports. However,
4909 if such a router generated a host list, and callout was configured, Exim
4910 crashed, because it could not find a port number from the (non-existent)
4911 transport. It now assumes port 25 in this circumstance.
4913 21. Added the -t option to exigrep.
4915 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4916 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4917 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4918 them is set, LOOKUP_LSEARCH is forced.
4920 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4922 24. Added sender and host information to the "rejected by local_scan()" log
4923 line; previously there was no indication of these.
4925 25. Added .include_if_exists.
4927 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4928 Linux. It turns out that not all file systems support this. Apparently some
4929 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4930 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4931 not supported on the file descriptor, is now ignored when Exim is trying to
4932 sync a directory. This applies only to Linux.
4934 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4936 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4937 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4940 29. Added support for maildirsize files from supplied patch (modified a bit).
4942 30. The use of :fail: followed by an empty string could lead Exim to respond to
4943 sender verification failures with (e.g.):
4945 550 Verification failed for <xxx>
4946 550 Sender verify failed
4948 where the first response line was missing the '-' that indicates it is not
4949 the final line of the response.
4951 31. The loop for finding the name of the user that called Exim had a hardwired
4952 limit of 10; it now uses the value of finduser_retries, which is used for
4953 all other user lookups.
4955 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4958 33. Exim was neglecting to zero errno before one call of strtol() when
4959 expanding a string and expecting an integer value. On some systems this
4960 resulted in spurious "integer overflow" errors. Also, it was casting the
4961 result into an int without checking.
4963 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4964 did not work. The code looks as if it has *never* worked, though it appears
4965 to have been documented since at least releast 1.62. I have made it work.
4967 35. The "timeout_DNS" error in retry rules, also documented since at least
4968 1.62, also never worked. As it isn't clear exactly what this means, and
4969 clearly it isn't a major issue, I have abolished the feature by treating it
4970 as "timeout", and writing a warning to the main and panic logs.
4972 36. The display of retry rules for -brt wasn't always showing the error code
4975 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4976 timeout_connect_A, timeout_connect_MX.
4978 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4979 to the empty sender.
4981 39. The daemon was not analysing the content of -oX till after it had closed
4982 stderr and disconnected from the controlling terminal. This meant that any
4983 syntax errors were only noted on the panic log, and the return code from
4984 the command was 0. By re-arranging the code a little, I've made the
4985 decoding happen first, so such errors now appear on stderr, and the return
4986 code is 1. However, the actual setting up of the sockets still happens in
4987 the disconnected process, so errors there are still only recorded on the
4990 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4991 connections (as happens on some IP stacks) was logged at start up time as
4992 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4993 it from "IPv6 and IPv4", which means that two separate sockets are being
4996 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4997 says whether AF_INET or AF_INET6 was passed as an argument.
4999 42. Exiwhat output was messed up when time zones were included in log
5002 43. Exiwhat now gives more information about the daemon's listening ports,
5003 and whether -tls-on-connect was used.
5005 44. The "port" option of the smtp transport is now expanded.
5007 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5008 silently ignored. Now an error message is written to the main and panic
5011 46. There's a new ACL modifier called "logwrite" which writes to a log file
5012 as soon as it is encountered.
5014 47. Added $local_user_uid and $local_user_gid at routing time.
5016 48. Exim crashed when trying to verify a sender address that was being
5019 49. Exim was recognizing only a space character after ".include". It now also
5020 recognizes a tab character.
5022 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5023 extracting the relevant information from the specification. The man page no
5024 longer contains scrambled data for the -d option, and I've added a section
5025 at the front about calling Exim under different names.
5027 51. Added "extra_headers" argument to the "mail" command in filter files.
5029 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5032 53. Installed eximstats 1.29.
5034 54. Added transport_filter_timeout as a generic transport option.
5036 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5037 Cc: header lines. This was required by RFC 822, but it not required by RFC
5040 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5041 incoming messages that did not have them. Now it does so only if the
5042 message originates locally, that is, if there is no associated remote host
5043 address. When Resent- header lines are present, this applies to the Resent-
5044 lines rather than the non-Resent- lines.
5046 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5047 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5049 58. Messages for configuration errors now include the name of the main
5050 configuration file - useful now that there may be more than one file in a
5051 list (.included file names were always shown).
5053 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5054 for those rare installations that do not start the daemon as root or run it
5055 setuid root. I've cut out the call to initgroups() if the daemon is not
5058 60. The Exim user and group can now be bound into the binary as text strings
5059 that are looked up at the start of Exim's processing.
5061 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5063 62. Added $mailstore_basename variable.
5065 63. Installed patch to sieve.c from Michael Haardt.
5067 64. When Exim failed to open the panic log after failing to open the main log,
5068 the original message it was trying to log was written to stderr and debug
5069 output, but if they were not available (the usual case in production), it
5070 was lost. Now it is written to syslog before the two lines that record the
5071 failures to open the logs.
5073 65. Users' Exim filters run in subprocesses under the user's uid. It is
5074 possible for a "deliver" command or an alias in a "personal" command to
5075 provoke an address rewrite. If logging of address rewriting is configured,
5076 this fails because the process is not running as root or exim. There may be
5077 a better way of dealing with this, but for the moment (because 4.30 needs
5078 to be released), I have disabled address rewrite logging when running a
5079 filter in a non-root, non-exim process.
5085 1. The buildconfig auxiliary program wasn't quoting the value set for
5086 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5087 not defined. This bug was masked in 4.22 by the effect that was fixed in
5090 2. Some messages that were rejected after a message id was allocated were
5091 shown as "incomplete" by exigrep. It no longer does this for messages that
5092 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5094 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5095 have allow_domain_literals set, the ID did not get logged in the <= line.
5096 Domain literals are now always recognized in Message-ID: header lines.
5098 4. The first argument for a ${extract expansion item is the key name or field
5099 number. Leading and trailing spaces in this item were not being ignored,
5100 causing some misleading effects.
5102 5. When deliver_drop_privilege was set, single queue runner processes started
5103 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5104 same command in the process it spins off) were not dropping privilege.
5106 6. When the daemon running as "exim" started a queue runner, it always
5107 re-executed Exim in the spun-off process. This is a waste of effort when
5108 deliver_drop_privilege is set. The new process now just calls the
5109 queue-runner function directly.
5115 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5118 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5119 ignored. Though the use of -oP was forcing the writing of a pid file, it
5120 was always written to the default place.
5122 3. If the message "no IP address found for host xxxx" is generated during
5123 incoming verification, it is now followed by identification of the incoming
5124 connection (so you can more easily find what provoked it).
5126 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5128 5. Added some features to "harden" Exim a bit more against certain attacks:
5130 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5131 be put in Local/Makefile. This is like the never_users runtime option,
5132 but it cannot be overridden. The default setting is "root".
5134 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5135 prefix string with which any file named in a -C command line option
5138 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5139 is retained for -C and -D only if the caller of Exim is root. Without
5140 it, the exim user may also use -C and -D and retain privilege.
5142 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5143 command line option is disabled.
5145 6. Macro names set by the -D option must start with an upper case letter, just
5146 like macro names defined in the configuration file.
5148 7. Added "dereference=" facility to LDAP.
5150 8. Two instances of the typo "uknown" in the source files are fixed.
5152 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5153 the Configure-Makefile script screwed up while processing it.
5155 10. Incorporated PCRE 4.4.
5157 11. The SMTP synchronization check was not operating right at the start of an
5158 SMTP session. For example, it could not catch a HELO sent before the client
5159 waited for the greeting. There is now a check for outstanding input at the
5160 point when the greeting is written. Because of the duplex, asynchronous
5161 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5162 way, but not yet received, when the check is performed.
5164 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5165 on TCP/IP sockets, because this apparently causes some broken clients to
5168 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5169 unchanged) from the Cygwin maintainer.
5171 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5172 was supported instead of testing for mbx. Effectively a typo.
5174 15. The spa authenticator server code was not checking that the input it
5175 received was valid base64.
5177 16. The debug output line for the "set" modifier in ACLs was not showing the
5178 name of the variable that was being set.
5180 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5182 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5183 lookup on its own. Something else was needed to trigger the lookup. For
5184 example, a match in host_lookup or the need for a host name in a host list.
5185 Now, if $sender_host_name is referenced and the host name has not yet been
5186 looked up, a lookup is performed. If the lookup fails, the variable remains
5187 empty, and $host_lookup_failed is set to "1".
5189 19. Added "eqi" as a case-independent comparison operator.
5191 20. The saslauthd authentication condition could segfault if neither service
5192 nor realm was specified.
5194 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5195 error message that was logged was misleading, and incoming SMTP
5196 connections were dropped. The message is now more accurate, and temporary
5197 errors are given to SMTP connections.
5199 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5200 (except RSET) with a 421 error, until QUIT is received. However, it was
5201 failing to send a response to QUIT.
5203 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5204 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5205 ACL is now run after the helo_try_verify_hosts code.
5207 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5208 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5209 case-independent, so other case variants are also recognized). Apparently
5210 some systems use these upper case variants.
5212 25. If more than two messages were waiting for the same host, and a transport
5213 filter was specified for the transport, Exim sent two messages over the
5214 same TCP/IP connection, and then failed with "socket operation on non-
5215 socket" when it tried to send the third.
5217 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5219 27. The extern definition of crypt16() in expand.c was not being excluded when
5220 the OS had its own crypt16() function.
5222 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5223 as a preferred synonym for return_size_limit, both as an option and as an
5226 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5228 30. Changed the default configuration ACL to relax the local part checking rule
5229 for addresses that are not in any local domains. For these addresses,
5230 slashes and pipe symbols are allowed within local parts, but the sequence
5231 /../ is explicitly forbidden.
5233 31. SPA server authentication was not clearing the challenge buffer before
5236 32. log_message in a "warn" ACL statement was writing to the reject log as
5237 well as to the main log, which contradicts the documentation and doesn't
5238 seem right (because no rejection is happening). So I have stopped it.
5240 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5241 However, I am unable to do any testing of this.
5243 34. Fixed an infelicity in the appendfile transport. When checking directories
5244 for a mailbox, to see if any needed to be created, it was accidentally
5245 using path names with one or more superfluous leading slashes; tracing
5246 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5248 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5249 contents are added to the log line that is written for every discarded
5250 recipient. (Previously a log_message setting was ignored.)
5252 36. The ${quote: operator now quotes the string if it is empty.
5254 37. The install script runs exim in order to find its version number. If for
5255 some reason other than non-existence or emptiness, which it checks, it
5256 could not run './exim', it was installing it with an empty version number,
5257 i.e. as "exim-". This error state is now caught, and the installation is
5260 38. An argument was missing from the function that creates an error message
5261 when Exim fails to connect to the socket for saslauthd authentication.
5262 This could cause Exim to crash, or give a corrupted message.
5264 39. Added isip, isip4, and isip6 to ${if conditions.
5266 40. The ACL variables $acl_xx are now saved with the message, and can be
5267 accessed later in routers, transports, and filters.
5269 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5270 strings in the file are not string-expanded.
5272 42. If a MAIL command specified a SIZE value that was too large to fit into an
5273 int variable, the check against message_size_limit failed. Such values are
5274 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5275 day this will have to be increased, but I don't think I want to be around
5276 when emails are that large.
5283 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5284 iconv() is not standard in FreeBSD.
5286 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5287 IPv6 enabled. The observed symptom was a segmentation fault on return from
5288 the function os_common_find_running_interfaces() in src/os.c.
5290 3. In the check_special_case() function in daemon.c I had used "errno" as an
5291 argument name, which causes warnings on some systems. This was basically a
5292 typo, since it was named "eno" in the comments!
5294 4. The code that waits for the clock to tick (at a resolution of some fraction
5295 of a second) so as to ensure message-id uniqueness was always waiting for
5296 at least one whole tick, when it could have waited for less. [This is
5297 almost certainly not relevant at current processor speeds, where it is
5298 unlikely to ever wait at all. But we try to future-proof.]
5300 5. The function that sleeps for a time interval that includes fractions of a
5301 second contained a race. It did not block SIGALRM between setting the
5302 timer, and suspending (a couple of lines later). If the interval was short
5303 and the sigsuspend() was delayed until after it had expired, the suspension
5304 never ended. On busy systems this could lead to processes getting stuck for
5307 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5308 process, before it forks any delivery processes. The open lookup caching
5309 mechanism meant that the open file or database connection was passed into
5310 the delivery process. The problem was that delivery processes always tidy
5311 up cached lookup data. This could cause a problem for the next delivery
5312 process started by the queue runner, because the external queue runner
5313 process does not know about the closure. So the next delivery process
5314 still has data in the lookup cache. In the case of a file lookup, there was
5315 no problem because closing a file descriptor in a subprocess doesn't affect
5316 the parent. However, if the lookup was caching a connection to a database,
5317 the connection was closed, and the second delivery process was likely to
5318 see errors such as "PGSQL: query failed: server closed the connection
5319 unexpectedly". The problem has been fixed by closing all cached lookups
5320 in a queue runner before running a delivery process.
5322 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5323 seem to have the "const" qualifier which it has on other OS. I've
5326 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5327 *of the same type* (client or server) with the same public name that an
5328 error should be diagnosed.
5330 9. When Exim looked up a host name for an IP address, but failed to find the
5331 original IP address when looking up the host name (a safety check), it
5332 output the message "<ip address> does not match any IP for NULL", which was
5333 confusing, to say the least. The bug was that the host name should have
5334 appeared instead of "NULL".
5336 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5337 user that is built into the binary, and the -C or -D options is used, root
5338 privilege is dropped before the configuration file is read. In addition,
5339 logging is switched to stderr instead of the normal log files. If the
5340 configuration then re-defines the Exim user, the unprivileged environment
5341 is probably not what is expected, so Exim logs a panic warning message (but
5344 However, if deliver_drop_privilege is set, the unprivileged state may well
5345 be exactly what is intended, so the warning has been cut out in that case,
5346 and Exim is allowed to try to write to its normal log files.
5352 1. smtp_return_error_details was not giving details for temporary sender
5353 or receiver verification errors.
5355 2. Diagnose a configuration error if two authenticators have the same public
5358 3. Exim used not to create the message log file for a message until the first
5359 delivery attempt. This could be confusing when incoming messages were held
5360 for policy or load reasons. The message log file is now created at the time
5361 the message is received, and an initial "Received" line is written to it.
5363 4. The automatically generated man page for command line options had a minor
5364 bug that caused no ill effects; however, a more serious problem was that
5365 the procedure for building the man page automatically didn't always
5366 operate. Consequently, release 4.20 contains an out-of-date version. This
5367 shouldn't happen again.
5369 5. When building Exim with embedded Perl support, the script that builds the
5370 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5371 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5373 6. The freeze_tell option was not being used for messages that were frozen on
5374 arrival, either by an ACL or by local_scan().
5376 7. Added the smtp_incomplete_transaction log selector.
5378 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5379 was accepting AUTH without a new EHLO.
5381 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5382 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5385 10. Logging of TCP/IP connections (when configured) now happens in the main
5386 daemon process instead of the child process, so that the TCP/IP connection
5387 count is more accurate (but it can never be perfect).
5389 11. The use of "drop" in a nested ACL was not being handled correctly in the
5390 outer ACL. Now, if condition failure induced by the nested "drop" causes
5391 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5392 or "require"), the connection is dropped.
5394 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5395 that yield "discard" can now be used with an "accept" or a "discard" verb,
5396 but an error is generated for any others (because I can't see a useful way
5397 to define what should happen).
5399 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5400 are now processed in the same way as lines in the Exim configuration file.
5401 In particular, continuation lines are supported.
5403 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5405 15. Added -ti meaning -t -i.
5407 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5408 domains, and warn by logging if others are found.
5410 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5411 returned from GIFCONF when Exim is trying to find the list of interfaces on
5412 a host. The code in os.c has been modified to copy each ifreq to an aligned
5413 structure in all cases.
5415 Also, in some cases, the returned ifreq's were being copied to a 'struct
5416 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5417 means the last couple of bytes of an IPv6 address could be chopped if the
5418 ifreq contained only a normal sockaddr (14 bytes storage).
5420 18. Named domain lists were not supported in the hosts_treat_as_local option.
5421 An entry such as +xxxx was not recognized, and was treated as a literal
5424 19. Ensure that header lines added by a DATA ACL are included in the reject log
5425 if the ACL subsequently rejects the message.
5427 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5428 MD5 (which is deprecated).
5430 21. When testing a filter file using -bf, Exim was writing a message when it
5431 took the sender from a "From " line in the message, but it was not doing so
5432 when it took $return_path from a Return-Path: header line. It now does.
5434 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5435 with a valid header line field name (a series of printing characters
5436 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5438 23. Changed "disc" in the source to "disk" to conform to the documentation and
5439 the book and for uniformity.
5441 24. Ignore Sendmail's -Ooption=value command line item.
5443 25. When execve() failed while trying to run a command in a pipe transport,
5444 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5445 could be confused with a return value of 69 from the command itself. This
5446 has been changed to 127, the value the shell returns if it is asked to run
5447 a non-existent command. The wording for the related log line suggests a
5448 non-existent command as the problem.
5450 26. If received_header_text expands to an empty string, do not add a Received:
5451 header line to the message. (Well, it adds a token one on the spool, but
5452 marks it "old" so that it doesn't get used or transmitted.)
5454 27. Installed eximstats 1.28 (addition of -nt option).
5456 28. There was no check for failure on the call to getsockname() in the daemon
5457 code. This can fail if there is a shortage of resources on the system, with
5458 ENOMEM, for example. A temporary error is now given on failure.
5460 29. Contrary to the C standard, it seems that in some environments, the
5461 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5462 program. Exim now does this explicitly; it affects the formatting of
5463 timestamps using strftime().
5465 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5466 complaints. I've now initialized all the variables, to avoid this.
5468 32. Header lines added by a system filter were not being "seen" during
5469 transport-time rewrites.
5471 33. The info_callback() function passed to OpenSSL is set up with type void
5472 (*)(SSL *, int, int), as described somewhere. However, when calling the
5473 function (actually a macro) that sets it up, the type void(*)() is
5474 expected. I've put in a cast to prevent warnings from picky compilers.
5476 34. If a DNS black list lookup found a CNAME record, but there were no A
5477 records associated with the domain it pointed at, Exim crashed.
5479 35. If a DNS black list lookup returned more than one A record, Exim ignored
5480 all but the first. It now scans all returned addresses if a particular IP
5481 value is being sought. In this situation, the contents of the
5482 $dnslist_value variable are a list of all the addresses, separated by a
5485 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5486 to accept a host name and all its aliases if the forward lookup for any of
5487 them yielded the IP address of the incoming connection. Now it accepts only
5488 those names whose forward lookup yields the correct IP address. Any other
5489 names are discarded. This closes a loophole whereby a rogue DNS
5490 administrator could create reverse DNS records to break through a
5491 wildcarded host restriction in an ACL.
5493 37. If a user filter or a system filter that ran in a subprocess used any of
5494 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5495 the wrong values were passed to the pipe command ($thisaddress had the
5496 value of $0, $0 had the value of $1, etc). This bug was introduced by
5497 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5499 38. Improved the line breaking for long SMTP error messages from ACLs.
5500 Previously, if there was no break point between 40 and 75 characters, Exim
5501 left the rest of the message alone. Two changes have been made: (a) I've
5502 reduced the minimum length to 35 characters; (b) if it can't find a break
5503 point between 35 and 75 characters, it looks ahead and uses the first one
5504 that it finds. This may give the occasional overlong line, but at least the
5505 remaining text gets split now.
5507 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5508 file descriptors might be low, and that setting 1000 would always raise it.
5509 It turns out that in some environments, the limit is already over 1000 and
5510 that lowering it causes trouble. So now Exim takes care not to decrease it.
5512 40. When delivering a message, the value of $return_path is set to $sender_
5513 address at the start of routing (routers may change the value). By an
5514 oversight, this default was not being set up when an address was tested by
5515 -bt or -bv, which affected the outcome if any router or filter referred to
5518 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5519 log only when "message" or "log_message" are set. However, if one of the
5520 conditions was an address verification, or a call to a nested ACL, the
5521 messages generated by the underlying test were being passed through. This
5522 no longer happens. The underlying message is available in $acl_verify_
5523 message for both "message" and "log_message" expansions, so it can be
5524 passed through if needed.
5526 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5527 new expansion $bh_ to give the encoded byte string without charset
5528 translation. Translation happens only if iconv() is available; HAVE_ICONV
5529 indicates this at build time. HEADERS_CHARSET gives the charset to
5530 translate to; headers_charset can change it in the configuration, and
5531 "headers charset" can change it in an individual filter file.
5533 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5534 that creates RFC 2047 encoded "words" labels them as that charset instead
5535 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5536 expansion operator; (ii) when Exim creates a From: line for a local
5537 message; (iii) when a header line is rewritten to include a "phrase" part.
5539 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5540 buggy, causing it to skip the first lines of messages whose message ID
5541 ended in 'D'. This would not have bitten before Exim release 4.14, because
5542 message IDs were unlikely to end in 'D' before then. The effect was to have
5543 incorrect size information for certain domains.
5545 45. #include "config.h" was missing at the start of the crypt16.c module. This
5546 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5549 46. If there was a timeout during a "random" callout check, Exim treated it as
5550 a failure of the random address, and carried on sending RSET and the real
5551 address. If the delay was just some slowness somewhere, the response to the
5552 original RCPT would be taken as a response to RSET and so on, causing
5553 mayhem of various kinds.
5555 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5556 when I implemented it. It didn't allow for the fact that some option values
5557 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5558 the right test for positive values.
5560 48. Domain names in DNS records are case-independent. Exim always looks them up
5561 in lower case. Some resolvers return domain names in exactly the case they
5562 appear in the zone file, that is, they may contain uppercase letters. Not
5563 all resolvers do this - some return always lower case. Exim was treating a
5564 change of case by a resolver as a change of domain, similar to a widening
5565 of a domain abbreviation. This triggered its re-routing code and so it was
5566 trying to route what was effectively the same domain again. This normally
5567 caused routing to fail (because the router wouldn't handle the domain
5568 twice). Now Exim checks for this case specially, and just changes the
5569 casing of the domain that it ultimately uses when it transmits the message
5572 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5575 50. If a filter generated a file delivery with a non-absolute name (possible if
5576 no home directory exists for the router), the forbid_file option was not
5579 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5580 the existing equality matching.
5582 52. Exim was using ints instead of ino_t variables in some places where it was
5583 dealing with inode numbers.
5585 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5586 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5587 called TMPDIR, and if it finds it is different, it changes its value.
5589 54. The smtp_printf() function is now made available to local_scan() so
5590 additional output lines can be written before returning. There is also an
5591 smtp_fflush() function to enable the detection of a dropped connection.
5592 The variables smtp_input and smtp_batched_input are exported to
5595 55. Changed the default runtime configuration: the message "Unknown user"
5596 has been removed from the ACL, and instead placed on the localuser router,
5597 using the cannot_route_message feature. This means that any verification
5598 failures that generate their own messages won't get overridden. Similarly,
5599 the "Unrouteable address" message that was in the ACL for unverifiable
5600 relay addresses has also been removed.
5602 56. Added hosts_avoid_esmtp to the smtp transport.
5604 57. The exicyclog script was not checking for the esoteric option
5605 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5606 will work only if exicyclog is run under the appropriate euid.
5608 58. Following a discussion on the list, the rules by which Exim recognises line
5609 endings on incoming messages have been changed. The -dropcr and drop_cr
5610 options are now no-ops, retained only for backwards compatibility. The
5611 following line terminators are recognized: LF CRLF CR. However, special
5612 processing applies to CR:
5614 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5615 nor a local message in the state where . is a terminator.
5617 (ii) If a bare CR is encountered in a header line, an extra space is added
5618 after the line terminator so as not to end the header. The reasoning
5619 behind this is that bare CRs in header lines are most likely either
5620 to be mistakes, or people trying to play silly games.
5622 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5623 was being incorrectly given as 18 bytes larger than it should have been.
5624 This is a VOB (very old bug).
5626 60. This may never have affected anything current, but just in case it has:
5627 When the local host is found other than at the start of a list of hosts,
5628 the local host, those with the same MX, and any that follow, are discarded.
5629 When the list in question was part of a longer list of hosts, the following
5630 hosts (not currently being processed) were also being discarded. This no
5631 longer happens. I'm not sure if this situation could ever has previously
5634 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5637 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5638 header. This is something that is recommended in a new Internet Draft, and
5639 is something that is documented as being done by Sendmail. There are two
5640 possible values. For messages generated by the autoreply transport, Exim
5643 Auto-Submitted: auto-replied
5645 whereas for all other generated messages (e.g. bounces) it adds
5647 Auto-Submitted: auto-generated
5649 63. The "personal" condition in filters now includes a test for the
5650 Auto-Submitted: header. If it contains the string "auto-" the message it
5651 not considered personal.
5653 64. Added rcpt_include_affixes as a generic transport option.
5655 65. Added queue_only_override (default true).
5657 66. Added the syslog_duplication option.
5659 67. If what should have been the first header line of a message consisted of
5660 a space followed by a colon, Exim was mis-interpreting it as a header line.
5661 It isn't of course - it is syntactically invalid and should therefore be
5662 treated as the start of the message body. The misbehaviour could have
5663 caused a number of strange effects, including loss of data in subsequent
5664 header lines, and spool format errors.
5666 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5667 client host had authenticated. This control can now be exercised by an ACL
5668 for more flexibility.
5670 69. By default, callouts do not happen when testing with -bh. There is now a
5671 variant, -bhc, which does actually run the callout code, including
5672 consulting and updating the callout cache.
5674 70. Added support for saslauthd authentication, courtesy of Alexander
5677 71. If statvfs() failed on the spool or log directories while checking their
5678 size for availability, Exim confusingly gave the error "space shortage".
5679 Furthermore, in debugging mode it crashed with a floating point exception.
5680 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5681 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5682 serious problem, Exim now writes to the main and panic logs when this
5683 happens, with details of the failure. It then refuses to accept the
5684 incoming message, giving the message "spool directory problem" or "log
5685 directory problem" with a 421 code for SMTP messages.
5687 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5688 0, 1, and 2 exist, because some OS complain for execs without them (see
5689 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5690 descriptors. However, the code omitted to check that the open succeeded,
5691 causing mysterious errors if for some reason the permissions on /dev/null
5692 got screwed. Now Exim writes a message to the main and panic logs, and
5693 bombs out if it can't open /dev/null.
5695 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5696 interact so that it is all more flexible. It is supposed to remain
5697 backwards compatible. Also added extra_local_interfaces.
5699 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5700 to bomb out with an assertion failure - to the client this appears as a
5701 connection drop. This problem occurs in the part of the code that was taken
5702 from the Samba project. Fortunately, the assertion is in a very simple
5703 function, so I have fixed this by reproducing the function inline in the
5704 one place where it is called, and arranging for authentication to fail
5705 instead of killing the process with assert().
5707 75. The SPA client code was not working when the server requested OEM rather
5708 than Unicode encoding.
5710 76. Added code to make require_files with a specific uid setting more usable in
5711 the case where statting the file as root fails - usually a non-root-mounted
5712 NFS file system. When this happens and the failure is EACCES, Exim now
5713 forks a subprocess and does the per-uid checking as the relevant uid.
5715 77. Added process_log_path.
5717 78. If log_file_path was not explicitly set, a setting of check_log_space or
5718 check_log_inodes was ignored.
5720 79. If a space check for the spool or log partitions fails, the incident is now
5721 logged. Of course, in the latter case the data may get lost...
5723 80. Added the %p formatting code to string_format() so that it can be used to
5724 print addresses in debug_print(). Adjusted all the address printing in the
5725 debugging in store.c to use %p rather than %d.
5727 81. There was a concern that a line of code in smtp_in.c could overflow a
5728 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5729 initially expressed, the concern was not well-founded, because trailing
5730 spaces are removed early. However, if the trailing spaces were followed by
5731 a NULL, they did not get removed, so the overflow was possible. Two fixes
5734 (a) I re-wrote the offending code in a cleaner fashion.
5735 (b) If an incoming SMTP command contains a NULL character, it is rejected
5738 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5739 runs initgroups(), so that if the Exim user is in any additional groups,
5740 they will be used during message reception.
5746 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob