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.
81 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
82 Pulled from Bugzilla 817 by Wolfgang Breyha.
84 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
91 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
93 This, or similar/improved, will also be change PP/11 of 4.82.
99 PP/01 Handle short writes when writing local log-files.
100 In practice, only affects FreeBSD (8 onwards).
101 Bugzilla 1053, with thanks to Dmitry Isaikin.
103 NM/01 Bugzilla 949 - Documentation tweak
105 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
108 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
110 PP/02 Implemented gsasl authenticator.
112 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
114 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
115 `pkg-config foo` for cflags/libs.
117 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
118 with rest of GSASL and with heimdal_gssapi.
120 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
121 `pkg-config foo` for cflags/libs for the TLS implementation.
123 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
124 properties get this fed in as external SSF. A number of robustness
125 and debugging improvements to the cyrus_sasl authenticator.
127 PP/08 cyrus_sasl server now expands the server_realm option.
129 PP/09 Bugzilla 1214 - Log authentication information in reject log.
130 Patch by Jeremy Harris.
132 PP/10 Added dbmjz lookup type.
134 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
136 PP/12 MAIL args handles TAB as well as SP, for better interop with
137 non-compliant senders.
138 Analysis and variant patch by Todd Lyons.
140 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
141 Bug report from Lars Müller <lars@samba.org> (via SUSE),
142 Patch from Dirk Mueller <dmueller@suse.com>
144 PP/13 tls_peerdn now print-escaped for spool files.
145 Observed some $tls_peerdn in wild which contained \n, which resulted
146 in spool file corruption.
148 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
149 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
150 or write after TLS renegotiation, which otherwise led to messages
153 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
154 as a tracking header (ie: a signed header comes before the signature).
155 Patch from Wolfgang Breyha.
157 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
158 comma-sep list; embedded commas doubled.
160 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
162 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
164 Report and patch from Dmitry Banschikov.
166 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
167 Removed SSL_clear() after SSL_new() which led to protocol negotiation
168 failures. We appear to now support TLS1.1+ with Exim.
170 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
171 lets Exim select keys and certificates based upon TLS SNI from client.
172 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
173 before an outbound SMTP session. New log_selector, +tls_sni.
175 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
176 NULL dereference. Report and patch from Alun Jones.
178 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
179 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
180 Not seeing resolver debug output on NetBSD, but suspect this is a
181 resolver implementation change.
183 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
184 Left warnings. Added "eximon gdb" invocation mode.
186 PP/21 Defaulting "accept_8bitmime" to true, not false.
188 PP/22 Added -bw for inetd wait mode support.
190 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
191 locate the relevant includes and libraries. Made this the default.
193 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
194 Bugzilla 1246, report and most of solution from Tomasz Kusy.
196 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
197 This may cause build issues on older platforms.
199 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
200 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
201 gnutls_require_mac & gnutls_require_protocols (no longer supported).
202 Added SNI support via GnuTLS too.
203 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
205 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
207 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
208 Applied second patch from Janne, implementing suggestion to default
209 multiple-strings-in-record handling to match SPF spec.
211 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
213 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
214 read-only, out of scope).
215 Patch from Wolfgang Breyha, report from Stuart Northfield.
217 PP/29 Fix three issues highlighted by clang analyser static analysis.
218 Only crash-plausible issue would require the Cambridge-specific
219 iplookup router and a misconfiguration.
220 Report from Marcin Mirosław.
222 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
224 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
225 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
226 As part of this, removing so much warning spew let me fix some minor
227 real issues in debug logging.
229 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
230 assignment on my part. Fixed.
232 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
233 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
234 Janne Snabb (who went above and beyond: thank you).
236 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
237 string otherwise requires a connection and a bunch more work and it's
238 relatively easy to get wrong. Should also expose TLS library linkage
241 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
242 64-bit ${eval} (JH/03).
244 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
245 GNU libc to support some of the 64-bit stuff, should not lead to
246 conflicts. Defined before os.h is pulled in, so if a given platform
247 needs to override this, it can.
249 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
250 protection layer was required, which is not implemented.
251 Bugzilla 1254, patch from Wolfgang Breyha.
253 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
254 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
255 tls_dhparam take prime identifiers. Also unbreak combination of
256 OpenSSL+DH_params+TLSSNI.
258 PP/39 Disable SSLv2 by default in OpenSSL support.
264 PP/01 Solaris build fix for Oracle's LDAP libraries.
265 Bugzilla 1109, patch from Stephen Usher.
267 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
269 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
272 TF/02 Fix a couple more cases where we did not log the error message
273 when unlink() failed. See also change 4.74-TF/03.
275 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
276 lock up or crash if it happened to be inside a call to libc when it
277 got a SIGUSR1 from exiwhat.
279 The SIGUSR1 handler appends the current process status to the process
280 log which is later printed by exiwhat. It used to use the general
281 purpose logging code to do this, but several functions it calls are
282 not safe for signals.
284 The new output code in the SIGUSR1 handler is specific to the process
285 log, and simple enough that it's easy to inspect for signal safety.
286 Removing some special cases also simplifies the general logging code.
287 Removing the spurious timestamps from the process log simplifies
290 TF/04 Improved ratelimit ACL condition.
292 The /noupdate option has been deprecated in favour of /readonly which
293 has clearer semantics. The /leaky, /strict, and /readonly update modes
294 are mutually exclusive. The update mode is no longer included in the
295 database key; it just determines when the database is updated. (This
296 means that when you upgrde Exim will forget old rate measurements.)
298 Exim now checks that the per_* options are used with an update mode that
299 makes sense for the current ACL. For example, when Exim is processing a
300 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
301 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
302 must specify per_mail/readonly. If you omit the update mode it defaults to
303 /leaky where that makes sense (as before) or /readonly where required.
305 The /noupdate option is now undocumented but still supported for
306 backwards compatibility. It is equivalent to /readonly except that in
307 ACLs where /readonly is required you may specify /leaky/noupdate or
308 /strict/noupdate which are treated the same as /readonly.
310 A useful new feature is the /count= option. This is a generalization
311 of the per_byte option, so that you can measure the throughput of other
312 aggregate values. For example, the per_byte option is now equivalent
313 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
315 The per_rcpt option has been generalized using the /count= mechanism
316 (though it's more complicated than the per_byte equivalence). When it is
317 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
318 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
319 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
320 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
321 Note that using per_rcpt with a non-readonly update mode in more than
322 one ACL will cause the recipients to be double-counted. (The per_mail
323 and per_byte options don't have this problem.)
325 The handling of very low rates has changed slightly. If the computed rate
326 is less than the event's count (usually one) then this event is the first
327 after a long gap. In this case the rate is set to the same as this event's
328 count, so that the first message of a spam run is counted properly.
330 The major new feature is a mechanism for counting the rate of unique
331 events. The new per_addr option counts the number of different
332 recipients that someone has sent messages to in the last time period. It
333 behaves like per_rcpt if all the recipient addresses are different, but
334 duplicate recipient addresses do not increase the measured rate. Like
335 the /count= option this is a general mechanism, so the per_addr option
336 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
337 example, measure the rate that a client uses different sender addresses
338 with the options per_mail/unique=$sender_address. There are further
339 details in the main documentation.
341 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
343 TF/06 Removed a few PCRE remnants.
345 TF/07 Automatically extract Exim's version number from tags in the git
346 repository when doing development or release builds.
348 PP/02 Raise smtp_cmd_buffer_size to 16kB.
349 Bugzilla 879. Patch from Paul Fisher.
351 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
352 Heavily based on revision 40f9a89a from Simon Arlott's tree.
355 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
357 PP/05 Variable $av_failed, true if the AV scanner deferred.
358 Bugzilla 1078. Patch from John Horne.
360 PP/06 Stop make process more reliably on build failure.
361 Bugzilla 1087. Patch from Heiko Schlittermann.
363 PP/07 Make maildir_use_size_file an _expandable_ boolean.
364 Bugzilla 1089. Patch from Heiko Schlittermann.
366 PP/08 Handle ${run} returning more data than OS pipe buffer size.
367 Bugzilla 1131. Patch from Holger Weiß.
369 PP/09 Handle IPv6 addresses with SPF.
370 Bugzilla 860. Patch from Wolfgang Breyha.
372 PP/10 GnuTLS: support TLS 1.2 & 1.1.
374 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
377 PP/11 match_* no longer expand right-hand-side by default.
378 New compile-time build option, EXPAND_LISTMATCH_RHS.
379 New expansion conditions, "inlist", "inlisti".
381 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
383 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
385 PP/14 fix log_write() format string regression from TF/03.
386 Bugzilla 1152. Patch from Dmitry Isaikin.
392 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
394 PP/02 Harmonised TLS library version reporting; only show if debugging.
395 Layout now matches that introduced for other libraries in 4.74 PP/03.
397 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
399 PP/04 New "dns_use_edns0" global option.
401 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
404 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
405 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
407 TK/01 Updated PolarSSL code to 0.14.2.
408 Bugzilla 1097. Patch from Andreas Metzler.
410 PP/07 Catch divide-by-zero in ${eval:...}.
413 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
416 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
417 format-string attack -- SECURITY: remote arbitrary code execution.
419 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
420 time unintentionally subject to list matching rules, letting the header
421 cause arbitrary Exim lookups (of items which can occur in lists, *not*
422 arbitrary string expansion). This allowed for information disclosure.
424 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
425 INT_MIN/-1 -- value coerced to INT_MAX.
431 NM/01 Workround for PCRE version dependency in version reporting
434 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
435 This fixes portability to compilers other than gcc, notably
436 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
438 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
439 makefiles for portability to HP-UX and POSIX correctness.
441 PP/01 Permit LOOKUP_foo enabling on the make command-line.
442 Also via indented variable definition in the Makefile.
443 (Debugging by Oliver Heesakkers).
445 PP/02 Restore caching of spamd results with expanded spamd_address.
446 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
448 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
449 Improves build reliability. Fix from: Frank Elsner
451 NM/02 Fix wide character breakage in the rfc2047 coding
452 Fixes bug 1064. Patch from Andrey N. Oktyabrski
454 NM/03 Allow underscore in dnslist lookups
455 Fixes bug 1026. Patch from Graeme Fowler
457 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
458 Code patches from Adam Ciarcinski of NetBSD.
460 NM/04 Fixed exiqgrep to cope with mailq missing size issue
463 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
464 is logged, to avoid truncation. Patch from John Horne.
466 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
467 Patch from Jakob Hirsch.
469 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
470 SQL string expansion failure details.
471 Patch from Andrey Oktyabrski.
473 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
474 Patch from Simon Arlott.
476 PP/09 New lookups functionality failed to compile on old gcc which rejects
477 extern declarations in function scope.
478 Patch from Oliver Fleischmann
480 PP/10 Use sig_atomic_t for flags set from signal handlers.
481 Check getgroups() return and improve debugging.
482 Fixed developed for diagnosis in bug 927 (which turned out to be
485 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
486 Patch from Mark Zealey.
488 PP/12 Bugzilla 1056: Improved spamd server selection.
489 Patch from Mark Zealey.
491 PP/13 Bugzilla 1086: Deal with maildir quota file races.
492 Based on patch from Heiko Schlittermann.
494 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
495 Patch from Uwe Doering, sign-off by Michael Haardt.
497 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
498 variable declaration deep within a block. Bug and patch from
501 PP/15 lookups-Makefile IRIX compatibilty coercion.
503 PP/16 Make DISABLE_DKIM build knob functional.
505 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
506 Patch by Simon Arlott
508 TF/03 Fix valgrind.h portability to C89 compilers that do not support
509 variable argument macros. Our copy now differs from upstream.
515 TF/01 Failure to get a lock on a hints database can have serious
516 consequences so log it to the panic log.
518 TF/02 Log LMTP confirmation messages in the same way as SMTP,
519 controlled using the smtp_confirmation log selector.
521 TF/03 Include the error message when we fail to unlink a spool file.
523 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
524 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
525 for maintaining out-of-tree patches for some time.
527 PP/01 Bugzilla 139: Documentation and portability issues.
528 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
529 Handle per-OS dynamic-module compilation flags.
531 PP/02 Let /dev/null have normal permissions.
532 The 4.73 fixes were a little too stringent and complained about the
533 permissions on /dev/null. Exempt it from some checks.
534 Reported by Andreas M. Kirchwitz.
536 PP/03 Report version information for many libraries, including
537 Exim version information for dynamically loaded libraries. Created
538 version.h, now support a version extension string for distributors
539 who patch heavily. Dynamic module ABI change.
541 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
542 privilege escalation vulnerability whereby the Exim run-time user
543 can cause root to append content of the attacker's choosing to
546 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
549 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
550 If dropping privileges for untrusted macros, we disabled normal logging
551 on the basis that it would fail; for the Exim run-time user, this is not
552 the case, and it resulted in successful deliveries going unlogged.
553 Fixed. Reported by Andreas Metzler.
559 PP/01 Date: & Message-Id: revert to normally being appended to a message,
560 only prepend for the Resent-* case. Fixes regression introduced in
561 Exim 4.70 by NM/22 for Bugzilla 607.
563 PP/02 Include check_rfc2047_length in configure.default because we're seeing
564 increasing numbers of administrators be bitten by this.
566 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
568 PP/03 Bugzilla 994: added openssl_options main configuration option.
570 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
572 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
574 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
576 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
577 without a peer certificate, leading to a segfault because of an
578 assumption that peers always have certificates. Be a little more
579 paranoid. Problem reported by Martin Tscholak.
581 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
582 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
583 NB: ClamAV planning to remove STREAM in "middle of 2010".
584 CL also introduces -bmalware, various -d+acl logging additions and
585 more caution in buffer sizes.
587 PP/09 Implemented reverse_ip expansion operator.
589 PP/10 Bugzilla 937: provide a "debug" ACL control.
591 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
593 PP/12 Bugzilla 973: Implement --version.
595 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
597 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
599 PP/15 Bugzilla 816: support multiple condition rules on Routers.
601 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
602 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
603 ignore trailing whitespace.
605 JJ/02 prevent non-panic DKIM error from being sent to paniclog
607 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
610 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
611 Notification from Dr Andrew Aitchison.
613 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
614 ExtendedDetectionInfo response format.
615 Notification from John Horne.
617 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
620 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
621 XSL and documented dependency on system catalogs, with examples of how
624 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
627 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
628 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
629 configuration file which is writeable by the Exim user or group.
631 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
632 of configuration files to cover files specified with the -C option if
633 they are going to be used with root privileges, not just the default
636 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
637 option (effectively making it always true).
639 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
640 files to be used while preserving root privileges.
642 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
643 that rogue child processes cannot use them.
645 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
646 run-time user, instead of root.
648 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
649 Exim run-time user without dropping privileges.
651 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
652 result string, instead of calling string_vformat() twice with the same
655 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
656 for other users. Others should always drop root privileges if they use
657 -C on the command line, even for a whitelisted configure file.
659 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
661 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
667 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
668 $data_path, and $header_path variables; fixed documentation bugs and
671 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
672 exipick to access non-standard spools, including the "frozen" queue
675 NM/01 Bugzilla 965: Support mysql stored procedures.
676 Patch from Alain Williams
678 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
680 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
681 Patch from Andreas Metzler
683 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
684 Patch from Kirill Miazine
686 NM/05 Bugzilla 671: Added umask to procmail example.
688 JJ/03 installed exipick 20100323.0, fixing doc bug
690 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
691 directory. Notification and patch from Dan Rosenberg.
693 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
695 TK/02 Improve log output when DKIM signing operation fails.
697 MH/01 Treat the transport option dkim_domain as a colon separated
698 list, not as a single string, and sign the message with each element,
699 omitting multiple occurences of the same signer.
701 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
702 Bugzilla 985, 986. Patch by Simon Arlott
704 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
705 Patch by Simon Arlott
707 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
708 MBX locking. Notification from Dan Rosenberg.
714 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
716 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
718 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
720 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
722 NM/04 Bugzilla 915: Flag broken perl installation during build.
728 TK/01 Added patch by Johannes Berg that expands the main option
729 "spamd_address" if it starts with a dollar sign.
731 TK/02 Write list of recipients to X-Envelope-Sender header when building
732 the mbox-format spool file for content scanning (suggested by Jakob
735 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
736 (http://www.dcc-servers.net/) support via dccifd. Activated by
737 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
739 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
740 by Mark Daniel Reidel <mr@df.eu>.
742 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
743 When building exim an external PCRE library is now needed -
744 PCRE is a system library on the majority of modern systems.
745 See entry on PCRE_LIBS in EDITME file.
747 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
748 conversation. Added nologin parameter to request.
749 Patch contributed by Kirill Miazine.
751 TF/01 Do not log submission mode rewrites if they do not change the address.
753 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
755 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
756 log files in place. Contributed by Roberto Lima.
758 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
760 TF/03 Bugzilla 615: When checking the local_parts router precondition
761 after a local_part_suffix or local_part_prefix option, Exim now
762 does not use the address's named list lookup cache, since this
763 contains cached lookups for the whole local part.
765 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
766 Robert Millan. Documentation is in experimental-spec.txt.
768 TF/04 Bugzilla 668: Fix parallel build (make -j).
770 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
772 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
773 Patch provided by Jan Srzednicki.
775 TF/05 Leading white space used to be stripped from $spam_report which
776 wrecked the formatting. Now it is preserved.
778 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
779 that they are available at delivery time.
781 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
783 TF/08 TLS error reporting now respects the incoming_interface and
784 incoming_port log selectors.
786 TF/09 Produce a more useful error message if an SMTP transport's hosts
787 setting expands to an empty string.
789 NM/06 Bugzilla 744: EXPN did not work under TLS.
790 Patch provided by Phil Pennock.
792 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
793 Patch provided by Richard Godbee.
795 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
796 acl_smtp_notquit, added index entry.
798 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
799 Patch provided by Eugene Bujak.
801 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
802 accept(). Patch provided by Maxim Dounin.
804 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
805 Patch provided by Phil Pennock.
807 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
809 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
810 Patch provided by Brad "anomie" Jorsch.
812 NM/14 Bugzilla 622: Added timeout setting to transport filter.
813 Patch provided by Dean Brooks.
815 TK/05 Add native DKIM support (does not depend on external libraries).
817 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
818 Patch provided by Graeme Fowler.
820 NM/16 Bugzilla 851: Documentation example syntax fix.
822 NM/17 Changed NOTICE file to remove references to embedded PCRE.
824 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
827 NM/19 Bugzilla 745: TLS version reporting.
828 Patch provided by Phil Pennock.
830 NM/20 Bugzilla 167: bool: condition support.
831 Patch provided by Phil Pennock.
833 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
834 clients. Patch provided by Phil Pennock.
836 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
837 Patch provided by Brad "anomie" Jorsch.
839 NM/23 Bugzilla 687: Fix misparses in eximstats.
840 Patch provided by Heiko Schlittermann.
842 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
843 Patch provided by Heiko Schlittermann.
845 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
846 plus update to original patch.
848 NM/26 Bugzilla 799: Documentation correction for ratelimit.
850 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
851 Patch provided by David Brownlee.
853 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
855 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
857 NM/30 Bugzilla 888: TLS documentation bugfixes.
859 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
861 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
862 Unlike the original bugzilla I have changed all shell scripts in src tree.
864 NM/33 Bugzilla 898: Transport filter timeout fix.
865 Patch by Todd Rinaldo.
867 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
868 Patch by Serge Demonchaux.
870 NM/35 Bugzilla 39: Base64 decode bug fixes.
871 Patch by Jakob Hirsch.
873 NM/36 Bugzilla 909: Correct connect() call in dcc code.
875 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
877 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
879 NM/39 Bugzilla 911: Fixed MakeLinks build script.
885 TK/01 Add preliminary DKIM support. Currently requires a forked version of
886 ALT-N's libdkim that I have put here:
887 http://duncanthrax.net/exim-experimental/
889 Note to Michael Haardt: I had to rename some vars in sieve.c. They
890 were called 'true' and it seems that C99 defines that as a reserved
891 keyword to be used with 'bool' variable types. That means you could
892 not include C99-style headers which use bools without triggering
893 build errors in sieve.c.
895 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
896 as mailq or other aliases. Changed the --help handling significantly
897 to do whats expected. exim_usage() emits usage/help information.
899 SC/01 Added the -bylocaldomain option to eximstats.
901 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
903 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
905 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
911 PH/01 Another patch from the Sieve maintainer.
913 PH/02 When an IPv6 address is converted to a string for single-key lookup
914 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
915 dots are used instead of colons so that keys in lsearch files need not
916 contain colons. This was done some time before quoting was made available
917 in lsearch files. However, iplsearch files do require colons in IPv6 keys
918 (notated using the quote facility) so as to distinguish them from IPv4
919 keys. This meant that lookups for IP addresses in host lists did not work
920 for iplsearch lookups.
922 This has been fixed by arranging for IPv6 addresses to be expressed with
923 colons if the lookup type is iplsearch. This is not incompatible, because
924 previously such lookups could never work.
926 The situation is now rather anomolous, since one *can* have colons in
927 ordinary lsearch keys. However, making the change in all cases is
928 incompatible and would probably break a number of configurations.
930 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
933 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
934 conversion specification without a maximum field width, thereby enabling
935 a rogue spamd server to cause a buffer overflow. While nobody in their
936 right mind would setup Exim to query an untrusted spamd server, an
937 attacker that gains access to a server running spamd could potentially
938 exploit this vulnerability to run arbitrary code as the Exim user.
940 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
941 $primary_hostname instead of what libspf2 thinks the hosts name is.
943 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
944 a directory entry by the name of the lookup key. Previously, if a
945 symlink pointed to a non-existing file or a file in a directory that
946 Exim lacked permissions to read, a lookup for a key matching that
947 symlink would fail. Now it is enough that a matching directory entry
948 exists, symlink or not. (Bugzilla 503.)
950 PH/03 The body_linecount and body_zerocount variables are now exported in the
953 PH/04 Added the $dnslist_matched variable.
955 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
956 This means they are set thereafter only if the connection becomes
959 PH/06 Added the client_condition to authenticators so that some can be skipped
960 by clients under certain conditions.
962 PH/07 The error message for a badly-placed control=no_multiline_responses left
963 "_responses" off the end of the name.
965 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
967 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
968 (without spaces) instead of just copying the configuration text.
970 PH/10 Added the /noupdate option to the ratelimit ACL condition.
972 PH/11 Added $max_received_linelength.
974 PH/12 Added +ignore_defer and +include_defer to host lists.
976 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
977 way in which PCRE > 7.0 is built.
979 PH/14 Implemented queue_only_load_latch.
981 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
982 MAIL command. The effect was to mangle the value on 64-bit systems.
984 PH/16 Another patch from the Sieve maintainer.
986 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
988 PH/18 If a system quota error occurred while trying to create the file for
989 a maildir delivery, the message "Mailbox is full" was not appended to the
990 bounce if the delivery eventually timed out. Change 4.67/27 below applied
991 only to a quota excession during the actual writing of the file.
993 PH/19 It seems that peer DN values may contain newlines (and other non-printing
994 characters?) which causes problems in log lines. The DN values are now
995 passed through string_printing() before being added to log lines.
997 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
998 and InterBase are left for another time.)
1000 PH/21 Added message_body_newlines option.
1002 PH/22 Guard against possible overflow in moan_check_errorcopy().
1004 PH/23 POSIX allows open() to be a macro; guard against that.
1006 PH/24 If the recipient of an error message contained an @ in the local part
1007 (suitably quoted, of course), incorrect values were put in $domain and
1008 $local_part during the evaluation of errors_copy.
1014 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1015 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1018 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1019 issue a MAIL command.
1021 PH/02 In an ACL statement such as
1023 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1025 if a client was not listed at all, or was listed with a value other than
1026 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1027 the condition was not true (as it should be), so access was not denied.
1028 The bug was that the ! inversion was incorrectly passed on to the second
1029 item. This has been fixed.
1031 PH/03 Added additional dnslists conditions == and =& which are different from
1032 = and & when the dns lookup returns more than one IP address.
1034 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1035 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1037 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1038 FSYNC, which compiles an option called disable_fsync that allows for
1039 bypassing fsync(). The documentation is heavily laced with warnings.
1041 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1043 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1044 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1045 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1046 including adding "make clean"; (3) Added -fPIC when compiling the test
1047 dynamically loaded module, to get rid of a warning.
1049 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1050 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1051 = 0s. The bug is otherwise harmless.
1053 PH/07 There was a bug in the dovecot authenticator such that the value of
1054 $auth1 could be overwritten, and so not correctly preserved, after a
1055 successful authentication. This usually meant that the value preserved by
1056 the server_setid option was incorrect.
1058 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1060 PH/09 Installed PCRE release 7.0.
1062 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1063 run for batched SMTP input. It is now run at the start of every message
1064 in the batch. While fixing this I discovered that the process information
1065 (output by running exiwhat) was not always getting set for -bs and -bS
1066 input. This is fixed, and it now also says "batched" for BSMTP.
1068 PH/11 Added control=no_pipelining.
1070 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1071 patch, slightly modified), and move the expansion of helo_data till after
1072 the connection is made in the smtp transport (so it can use these
1075 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1077 PH/14 Added log_selector = +pid.
1079 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1081 PH/16 Add ${if forany and ${if forall.
1083 PH/17 Added dsn_from option to vary the From: line in DSNs.
1085 PH/18 Flush SMTP output before performing a callout, unless control =
1086 no_callout_flush is set.
1088 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1089 was true (the default) a successful delivery failed to delete the retry
1090 item, thus causing premature timeout of the address. The bug is now
1093 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1095 PH/21 Long custom messages for fakedefer and fakereject are now split up
1096 into multiline reponses in the same way that messages for "deny" and
1097 other ACL rejections are.
1099 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1100 with slight modification.
1102 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1103 draft, changing the syntax and factoring some duplicate code".
1105 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1106 for deliveries of the second and subsequent messages over the same SMTP
1109 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1110 ${reduce, with only minor "tidies".
1112 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1114 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1115 expansion side effects.
1117 PH/27 When a message times out after an over-quota error from an Exim-imposed
1118 quota, the bounce message says "mailbox is full". This message was not
1119 being given when it was a system quota that was exceeded. It now should
1122 MH/03 Made $recipients available in local_scan(). local_scan() already has
1123 better access to the recipient list through recipients_list[], but
1124 $recipients can be useful in postmaster-provided expansion strings.
1126 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1127 in the case of a MAIL command with additional options following the
1128 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1129 were accidentally chopped off.
1131 PH/29 SMTP synchronization checks are implemented when a command is read -
1132 there is a check that no more input is waiting when there shouldn't be
1133 any. However, for some commands, a delay in an ACL can mean that it is
1134 some time before the response is written. In this time, more input might
1135 arrive, invalidly. So now there are extra checks after an ACL has run for
1136 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1137 pipelining has not been advertised.
1139 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1141 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1142 returned in the SMTP response when smtp_return_error_details was false.
1143 This has been fixed.
1145 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1146 instead of the C I/O that was originally supplied, because problems were
1147 reported on Solaris.
1149 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1150 Exim which did not show up earlier: it was assuming that a call to
1151 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1152 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1153 was a macro that became an assignment, so it seemed to work. This has
1154 changed to a proper function call with a void return, hence the compile
1155 error. Exim's code has been fixed.
1157 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1160 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1162 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1164 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1165 because it turns out to be too big or there is a timeout) there is no
1166 "Completed" line in the log. When some messages of this type were
1167 selected by exigrep, they were listed as "not completed". Others were
1168 picked up by some special patterns. I have improved the selection
1169 criteria to be more general.
1171 PH/38 The host_find_failed option in the manualroute router can now be set
1172 to "ignore", to completely ignore a host whose IP address cannot be
1173 found. If all hosts are ignored, the behaviour is controlled by the new
1174 host_all_ignored option.
1176 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1177 homing or because of multiple MX records with /mx) generated more than
1178 one IP address, and the following item turned out to be the local host,
1179 all the secondary addresses of the first item were incorrectly removed
1180 from the list, along with the local host and any following hosts (which
1181 is what is supposed to happen).
1183 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1184 whoever called Exim into the -H file. In the case of the daemon it was
1185 behaving confusingly. When first started, it used values for whoever
1186 started the daemon, but after a SIGHUP it used the Exim user (because it
1187 calls itself on a restart). I have changed the code so that it now always
1190 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1191 message are rejected with the same error (e.g. no authentication or bad
1192 sender address), and a DATA command is nevertheless sent (as can happen
1193 with PIPELINING or a stupid MUA), the error message that was given to the
1194 RCPT commands is included in the rejection of the DATA command. This is
1195 intended to be helpful for MUAs that show only the final error to their
1198 PH/42 Another patch from the Sieve maintainer.
1200 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1201 Eximstats - Fixed some broken HTML links and added missing column headers
1203 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1204 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1206 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1208 PH/43 Yet another patch from the Sieve maintainer.
1210 PH/44 I found a way to check for a TCP/IP connection going away before sending
1211 the response to the final '.' that terminates a message, but only in the
1212 case where the client has not sent further data following the '.'
1213 (unfortunately, this is allowed). However, in many cases there won't be
1214 any further data because there won't be any more messages to send. A call
1215 to select() can be used: if it shows that the input is "ready", there is
1216 either input waiting, or the socket has been closed. An attempt to read
1217 the next input character can distinguish the two cases. Previously, Exim
1218 would have sent an OK response which the client would never have see.
1219 This could lead to message repetition. This fix should cure that, at
1220 least in a lot of common cases.
1222 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1223 advertised in response to EHLO.
1229 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1230 fixed by 4.65/MH/01 (is this a record?) are fixed:
1232 (i) An empty string was always treated as zero by the numeric comparison
1233 operators. This behaviour has been restored.
1235 (ii) It is documented that the numeric comparison operators always treat
1236 their arguments as decimal numbers. This was broken in that numbers
1237 starting with 0 were being interpreted as octal.
1239 While fixing these problems I realized that there was another issue that
1240 hadn't been noticed. Values of message_size_limit (both the global option
1241 and the transport option) were treated as octal if they started with 0.
1242 The documentation was vague. These values are now always treated as
1243 decimal, and I will make that clear in the documentation.
1249 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1250 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1253 MH/01 Don't check that the operands of numeric comparison operators are
1254 integers when their expansion is in "skipping" mode (fixes bug
1255 introduced by 4.64-PH/07).
1257 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1258 child addresses, Exim now panics and dies. Previously, because the count
1259 is held in a short int, deliveries were likely to be lost. As such a
1260 large number of recipients for a single message is ridiculous
1261 (performance will be very, very poor), I have chosen to impose a limit
1262 rather than extend the field.
1268 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1269 leftover -K file (the existence of which was triggered by #402).
1270 While we were at it, introduced process PID as part of the -K
1271 filename. This should rule out race conditions when creating
1274 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1275 processing considerably. Previous code took too long for large mails,
1276 triggering a timeout which in turn triggers #401.
1278 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1279 in the DK code in transports.c. sendfile() is not really portable,
1280 hence the _LINUX specificness.
1282 TF/01 In the add_headers option to the mail command in an Exim filter,
1283 there was a bug that Exim would claim a syntax error in any
1284 header after the first one which had an odd number of characters
1287 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1288 callout verification, Exim cached a "reject" for the entire domain. This
1289 is correct for most verifications, but it is not correct for a recipient
1290 verification with use_sender or use_postmaster set, because in that case
1291 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1292 case of MAIL FROM:<> rejection from other early rejections (e.g.
1293 rejection of HELO). When verifying a recipient using a non-null MAIL
1294 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1295 Whatever the result of the callout, the value of the domain cache is
1296 left unchanged (for any other kind of callout, getting as far as trying
1297 RCPT means that the domain itself is ok).
1299 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1302 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1303 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1304 session, the daemon ignores ECONNECT errors and logs others; it now
1305 ignores EPIPE as well.
1307 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1308 (quoted-printable decoding).
1310 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1311 later the small subsequent patch to fix an introduced bug.
1313 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1315 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1317 PH/08 An error is now given if message_size_limit is specified negative.
1319 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1320 to be given (somewhat) arbitrary names.
1322 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1325 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1326 miscellaneous code fixes
1328 PH/10 Added the log_reject_target ACL modifier to specify where to log
1331 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1332 hostname. This is wrong, because it relates to the incoming message (and
1333 probably the interface on which it is arriving) and not to the outgoing
1334 callout (which could be using a different interface). This has been
1335 changed to use the value of the helo_data option from the smtp transport
1336 instead - this is what is used when a message is actually being sent. If
1337 there is no remote transport (possible with a router that sets up host
1338 addresses), $smtp_active_hostname is used.
1340 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1341 tweaks were necessary in order to get it to work (see also 21 below):
1342 (a) The code assumed that strncpy() returns a negative number on buffer
1343 overflow, which isn't the case. Replaced with Exim's string_format()
1345 (b) There were several signed/unsigned issues. I just did the minimum
1346 hacking in of casts. There is scope for a larger refactoring.
1347 (c) The code used strcasecmp() which is not a standard C function.
1348 Replaced with Exim's strcmpic() function.
1349 (d) The code set only $1; it now sets $auth1 as well.
1350 (e) A simple test gave the error "authentication client didn't specify
1351 service in request". It would seem that Dovecot has changed its
1352 interface. Fortunately there's a specification; I followed it and
1353 changed what the client sends and it appears to be working now.
1355 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1358 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1359 address A is aliased to B and C, where B exists and C does not. Without
1360 -v the output is "A verified" because verification stops after a
1361 successful redirection if more than one address is generated. However,
1362 with -v the child addresses are also verified. Exim was outputting "A
1363 failed to verify" and then showing the successful verification for C,
1364 with its parentage. It now outputs "B failed to verify", showing B's
1365 parentage before showing the successful verification of C.
1367 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1368 look up a TXT record in a specific list after matching in a combined
1371 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1372 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1373 they consult the DNS. I had assumed they would set it the way they
1374 wanted; and indeed my experiments on Linux seem to show that in some
1375 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1376 To be on the safe side, however, I have now made the interface to
1377 host_find_byname() similar to host_find_bydns(), with an argument
1378 containing the DNS resolver options. The host_find_byname() function now
1379 sets these options at its start, just as host_find_bydns() does. The smtp
1380 transport options dns_qualify_single and dns_search_parents are passed to
1381 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1382 of host_find_byname() use the default settings of RES_DEFNAMES
1383 (qualify_single) but not RES_DNSRCH (search_parents).
1385 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1386 spool_read_header() do less string testing, by means of a preliminary
1387 switch on the second character of optional "-foo" lines. (This is
1388 overdue, caused by the large number of possibilities that now exist.
1389 Originally there were few.) While I was there, I also converted the
1390 str(n)cmp tests so they don't re-test the leading "-" and the first
1391 character, in the hope this might squeeze out yet more improvement.
1393 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1394 flag allowing group syntax was set by the header_syntax check but not
1395 turned off, possible causing trouble later; (2) The flag was not being
1396 set at all for the header_verify test, causing "group"-style headers to
1397 be rejected. I have now set it in this case, and also caused header_
1398 verify to ignore an empty address taken from a group. While doing this, I
1399 came across some other cases where the code for allowing group syntax
1400 while scanning a header line wasn't quite right (mostly, not resetting
1401 the flag correctly in the right place). These bugs could have caused
1402 trouble for malformed header lines. I hope it is now all correct.
1404 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1405 with the "reply" argument non-NULL. The code, however (which originally
1406 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1407 but it didn't always do it. This confused somebody who was copying the
1408 code for some other use. I have removed all the tests.
1410 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1411 feature that was used to support insecure browsers during the U.S. crypto
1412 embargo. It requires special client support, and Exim is probably the
1413 only MTA that supported it -- and would never use it because real RSA is
1414 always available. This code has been removed, because it had the bad
1415 effect of slowing Exim down by computing (never used) parameters for the
1416 RSA_EXPORT functionality.
1418 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1419 authenticator to fail if there's a tab character in the incoming data
1420 (there should never be unless someone is messing about, as it's supposed
1421 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1422 if the connection is using TLS or if the remote IP is the same as the
1423 local IP, and the "valid-client-cert option" if a client certificate has
1426 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1427 authenticators. This can be used for authorization after authentication
1428 succeeds. (In the case of plaintext, it servers for both authentication
1431 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1432 if any retry times were supplied.
1434 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1435 connection, where there is no client IP address to check. In this
1436 situation, the verify now always succeeds.
1438 PH/25 Applied John Jetmore's -Mset patch.
1440 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1442 PH/27 In a string expansion for a processed (not raw) header when multiple
1443 headers of the same name were present, leading whitespace was being
1444 removed from all of them, but trailing whitespace was being removed only
1445 from the last one. Now trailing whitespace is removed from each header
1446 before concatenation. Completely empty headers in a concatenation (as
1447 before) are ignored.
1449 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1450 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1452 PH/29 [Removed. This was a change that I later backed out, and forgot to
1453 correct the ChangeLog entry (that I had efficiently created) before
1454 committing the later change.]
1456 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1457 address errors (4xx response to RCPT) over the same connection as other
1458 messages routed to the same hosts. Such deliveries are always "forced",
1459 so retry times are not inspected. This resulted in far too many retries
1460 for the affected addresses. The effect occurred only when there were more
1461 hosts than the hosts_max_try setting in the smtp transport when it had
1462 the 4xx errors. Those hosts that it had tried were not added to the list
1463 of hosts for which the message was waiting, so if all were tried, there
1464 was no problem. Two fixes have been applied:
1466 (i) If there are any address or message errors in an SMTP delivery, none
1467 of the hosts (tried or untried) are now added to the list of hosts
1468 for which the message is waiting, so the message should not be a
1469 candidate for sending over the same connection that was used for a
1470 successful delivery of some other message. This seems entirely
1471 reasonable: after all the message is NOT "waiting for some host".
1472 This is so "obvious" that I'm not sure why it wasn't done
1473 previously. Hope I haven't missed anything, but it can't do any
1474 harm, as the worst effect is to miss an optimization.
1476 (ii) If, despite (i), such a delivery is accidentally attempted, the
1477 routing retry time is respected, so at least it doesn't keep
1478 hammering the server.
1480 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1481 in ${readsocket because some servers need this prod.
1483 PH/32 Added some extra debug output when updating a wait-xxx database.
1485 PH/33 The hint "could be header name not terminated by colon", which has been
1486 given for certain expansion errors for a long time, was not being given
1487 for the ${if def:h_colon_omitted{... case.
1489 PH/34 The spec says: "With one important exception, whenever a domain list is
1490 being scanned, $domain contains the subject domain." There was at least
1491 one case where this was not true.
1493 PH/35 The error "getsockname() failed: connection reset by peer" was being
1494 written to the panic log as well as the main log, but it isn't really
1495 panic-worthy as it just means the connection died rather early on. I have
1496 removed the panic log writing for the ECONNRESET error when getsockname()
1499 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1500 runs only) independently of the message's sender address. This meant
1501 that, if the 4xx error was in fact related to the sender, a different
1502 message to the same recipient with a different sender could confuse
1503 things. In particualar, this can happen when sending to a greylisting
1504 server, but other circumstances could also provoke similar problems.
1505 I have changed the default so that the retry time for these errors is now
1506 based a combination of the sender and recipient addresses. This change
1507 can be overridden by setting address_retry_include_sender=false in the
1510 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1511 remote server are returned as part of bounce messages. This was not
1512 happening for LMTP over a pipe (the lmtp transport), but now it is the
1513 same for both kinds of LMTP.
1515 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1516 in header lines that were in fact CNAMEs. This is no longer the case.
1518 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1519 and queue runs started by the daemon processed all messages. This has
1520 been fixed so that -R and -S can now usefully be given with -q<time>.
1522 PH/40 Import PCRE release 6.7 (fixes some bugs).
1524 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1526 PH/42 Give an error if -q is specified more than once.
1528 PH/43 Renamed the variables $interface_address and $interface_port as
1529 $received_ip_address and $received_port, to make it clear that these
1530 values apply to message reception, and not to the outgoing interface when
1531 a message is delivered. (The old names remain recognized, of course.)
1533 PH/44 There was no timeout on the connect() call when using a Unix domain
1534 socket in the ${readsocket expansion. There now is.
1536 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1537 be meaningful with "accept".
1539 SC/01 Eximstats V1.43
1540 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1542 SC/02 Eximstats V1.44
1543 Use a glob alias rather than an array ref in the generated
1544 parser. This improves both readability and performance.
1546 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1547 Collect SpamAssassin and rejection statistics.
1548 Don't display local sender or destination tables unless
1549 there is data to show.
1550 Added average volumes into the top table text output.
1552 SC/04 Eximstats V1.46
1553 Collect data on the number of addresses (recipients)
1554 as well as the number of messages.
1556 SC/05 Eximstats V1.47
1557 Added 'Message too big' to the list of mail rejection
1558 reasons (thanks to Marco Gaiarin).
1560 SC/06 Eximstats V1.48
1561 Mainlog lines which have GMT offsets and are too short to
1562 have a flag are now skipped.
1564 SC/07 Eximstats V1.49 (Alain Williams)
1565 Added the -emptyok flag.
1567 SC/08 Eximstats V1.50
1568 Fixes for obtaining the IP address from reject messages.
1570 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1571 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1572 whitesspace changes from 4.64-PH/27
1574 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1577 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1578 are found, allow negative numbers in numeric criteria)
1580 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1582 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1583 to match changes made in 4.64-PH/43
1585 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1587 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1588 "domain" is actually an IP address, and force a failure. This locks out
1589 those revolvers/nameservers that support "A-for-A" lookups, in
1590 contravention of the specifications.
1592 PH/48 When a host name was looked up from an IP address, and the subsequent
1593 forward lookup of the name timed out, the host name was left in
1594 $sender_host_name, contrary to the specification.
1596 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1597 restricted to single-key lookups, Exim was not diagnosing an error if
1598 * or *@ was used with a query-style lookup.
1600 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1602 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1603 long ago, but noone interested enough thought of it. Let's just say that
1604 the "1.1" means that there are some new functions that weren't there at
1605 some point in the past.
1607 PH/51 Error processing for expansion failure of helo_data from an smtp
1608 transport during callout processing was broken.
1610 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1611 tested/used via the -bh/-bhc/-bs options.
1613 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1614 bug, fixed in subsequent PCRE releases).
1616 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1617 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1619 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1625 SC/01 Use a glob alias rather than an array ref in eximstats generated
1626 parser. This improves both readability and performance.
1628 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1629 Don't display local sender or destination tables in eximstats unless
1630 there is data to show.
1631 Added average volumes into the eximstats top table text output.
1633 SC/03 Collect data on the number of addresses (recipients) as well
1634 as the number of messages in eximstats.
1636 TF/01 Correct an error in the documentation for the redirect router. Exim
1637 does (usually) call initgroups() when daemonizing.
1639 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1640 with consistent privilege compared to when running as a daemon.
1642 TF/03 Note in the spec that $authenticated_id is not set for local
1643 submissions from trusted users.
1645 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1646 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1648 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1649 by adding some example configuration directives to the default
1650 configuration file. A little bit of work is required to uncomment the
1651 directives and define how usernames and passwords are checked, but
1652 there is now a framework to start from.
1654 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1655 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1656 without this. I don't know how relevant this is to other LDAP libraries.
1658 PH/02 Add the verb name to the "unknown ACL verb" error.
1660 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1662 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1664 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1665 directory not expanded when it should be if an expanded home directory
1666 was set for the address (which is overridden by the transport).
1668 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1671 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1672 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1673 because it is too late at that time, and has no effect.
1675 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1676 security issue with \' (bugzilla #107). I could not use the
1677 PQescapeStringConn() function, because it needs a PGconn value as one of
1680 PH/08 When testing addresses using -bt, indicate those final addresses that
1681 are duplicates that would not cause an additional delivery. At least one
1682 person was confused, thinking that -bt output corresponded to deliveries.
1683 (Suppressing duplicates isn't a good idea as you lose the information
1684 about possibly different redirections that led to the duplicates.)
1686 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1687 systems where poll() doesn't work, in particular OS X.
1689 PH/10 Added more information to debugging output for retry time not reached.
1691 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1692 operations in malware.c.
1694 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1697 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1698 syslog_duplication set false, log lines that would normally be written
1699 both the the main log and to the reject log were not written to syslog at
1702 PH/14 In the default configuration, change the use of "message" in ACL warn
1703 statements to "add_header".
1705 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1706 not followed by a command (e.g. "seen endif").
1708 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1709 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1712 PH/17 Added extra conditions to the default value of delay_warning_condition
1716 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1717 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1718 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1721 The Auto-Submitted: and various List- headers are standardised, whereas I
1722 don't think Precedence: ever was.
1724 PH/18 Refactored debugging code in route_finduser() to show more information,
1725 in particular, the error code if getpwnam() issues one.
1727 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1728 This is apparently needed in addition to the PH/07 change above to avoid
1729 any possible encoding problems.
1731 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1732 but not after initializing Perl.
1734 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1735 output them only if debugging. By default they are written stderr,
1736 apparently, which is not desirable.
1738 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1741 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1744 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1746 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1747 authenticated or an ident call has been made. Suppress the default
1748 values for $authenticated_id and $authenticated_sender (but permit -oMai
1749 and -oMas) when testing with -bh.
1751 PH/24 Re-jigged the order of the tests in the default configuration so that the
1752 tests for valid domains and recipients precede the DNS black list and CSA
1753 tests, on the grounds that those ones are more expensive.
1755 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1756 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1757 This bug exists in every version of Exim that I still have, right back to
1760 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1761 However, an attempt to turn on case-sensitivity in a regex key by
1762 including (?-i) didn't work because the subject string was already
1763 lowercased, and the effects were non-intuitive. It turns out that a
1764 one-line patch can be used to allow (?-i) to work as expected.
1770 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1771 other effects) broke the use of negated acl sub-conditions.
1773 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1776 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1777 "Deny" causes Exim to reject the incoming connection with a 554 error.
1778 Unfortunately, if there is a major crisis, such as a disk failure,
1779 tcp-wrappers gives "deny", whereas what one would like would be some
1780 kind of temporary error. A kludge has been added to help with this.
1781 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1782 554 error is used if errno is still zero or contains ENOENT (which occurs
1783 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1786 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1788 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1789 errors. Otherwise a message that provokes a temporary error (when other
1790 messages do not) can cause a whole host to time out.
1792 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1793 addresses were routed directly to files or pipes from a redirect router.
1794 File deliveries just didn't batch; pipe deliveries might have suffered
1797 PH/06 A failure to get a lock for a hints database would erroneously always say
1798 "Failed to get write lock", even when it was really a read lock.
1800 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1801 of 0600. This has been changed to use the value of the lockfile_mode
1802 option (which defaults to 0600).
1804 PH/08 Applied small patch from the Sieve maintainer.
1806 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1807 folder from quota calculations, a direct delivery into this folder messed
1808 up the contents of the maildirsize file. This was because the regex was
1809 used only to exclude .Trash (or whatever) when the size of the mailbox
1810 was calculated. There was no check that a delivery was happening into an
1811 excluded directory. This bug has been fixed by ignoring all quota
1812 processing for deliveries into excluded directories.
1814 PH/10 Added the maildirfolder_create_regex option to appendfile.
1820 PH/01 The code for finding all the local interface addresses on a FreeBSD
1821 system running IPv6 was broken. This may well have applied to all BSD
1822 systems, as well as to others that have similar system calls. The broken
1823 code found IPv4 interfaces correctly, but gave incorrect values for the
1824 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1825 that it would not match correctly against @[] and not recognize the IPv6
1828 PH/02 The ipliteral router was not recognizing addresses of the form user@
1829 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1831 PH/03 Added disable_ipv6.
1833 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1834 decoded form, because it is most often used to construct To: headers
1835 lines in autoreplies, and the decoded form may well be syntactically
1836 invalid. However, $reply_address has leading white space removed, and all
1837 newlines turned into spaces so that the autoreply transport does not
1840 PH/05 If group was specified without a user on a router, and no group or user
1841 was specified on a transport, the group from the router was ignored.
1843 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1844 for visible compile-time settings that can be used to change these
1845 numbers, for those that want even more. Backwards compatibility with old
1846 spool files has been maintained. However, going back to a previous Exim
1847 release will lost any variables that are in spool files.
1849 PH/07 Two small changes when running in the test harness: increase delay when
1850 passing a TCP/IP connection to a new process, in case the original
1851 process has to generate a bounce, and remove special handling of
1852 127.0.0.2 (sic), which is no longer necessary.
1854 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1855 be the same on different OS.
1857 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1860 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1861 whether --show-vars was specified or not
1863 JJ/02 exipick: Added support for new ACL variable spool format introduced
1866 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1867 syntactically invalid From: or Reply-to: line, and a filter used this to
1868 generate an autoreply, and therefore failed to obtain an address for the
1869 autoreply, Exim could try to deliver to a non-existent relative file
1870 name, causing unrelated and misleading errors. What now happens is that
1871 it logs this as a hard delivery error, but does not attempt to create a
1874 PH/11 The exinext utility has a -C option for testing purposes, but although
1875 the given file was scanned by exinext itself; it wasn't being passed on
1876 when Exim was called.
1878 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1879 an end-of-file indication when reading a command response.
1881 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1882 compiled. In many other places in Exim, IPv6 addresses are always
1883 recognized, so I have changed this. It also means that IPv4 domain
1884 literals of the form [IPV4:n.n.n.n] are now always recognized.
1886 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1887 used if the router is not running as root, for example, when verifying at
1888 ACL time, or when using -bh. The debugging output from this situation was
1889 non-existent - all you got was a failure to exec. I have made two
1892 (a) Failures to set uid/gid, the current directory, or a process leader
1893 in a subprocess such as that created by queryprogram now generate
1894 suitable debugging ouput when -d is set.
1896 (b) The queryprogram router detects when it is not running as root,
1897 outputs suitable debugging information if -d is set, and then runs
1898 the subprocess without attempting to change uid/gid.
1900 PH/15 Minor change to Makefile for building test_host (undocumented testing
1903 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1904 additional section of a DNS packet that returns MX or SRV records.
1905 Instead, it always explicitly searches for A/AAAA records. This avoids
1906 major problems that occur when a DNS server includes only records of one
1907 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1908 fixed another bug: if SRV records were looked up and the corresponding
1909 address records were *not* found in the additional section, the port
1910 values from the SRV records were lost.
1912 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1913 using the correct key (the original address) when searching the retry
1914 rules in order to find which one to use for generating the retry hint.
1916 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1917 adding the default one. Similarly, if it contains a Reply-To: header, the
1918 errors_reply_to option, if set, is not used.
1920 PH/19 When calculating a retry time, Exim used to measure the "time since
1921 failure" by looking at the "first failed" field in the retry record. Now
1922 it does not use this if it is later than than the arrival time of the
1923 message. Instead it uses the arrival time. This makes for better
1924 behaviour in cases where some deliveries succeed, thus re-setting the
1925 "first failed" field. An example is a quota failure for a huge message
1926 when small messages continue to be delivered. Without this change, the
1927 "time since failure" will always be short, possible causing more frequent
1928 delivery attempts for the huge message than are intended.
1929 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1931 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1932 $1, $2, $3) because the numerical variables can be reset during some
1933 expansion items (e.g. "match"), thereby losing the authentication data.
1935 PH/21 Make -bV show the size of off_t variables so that the test suite can
1936 decide whether to run tests for quotas > 2G.
1938 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1939 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1940 filecount value is greater than 2G or if a quota value is greater than 2G
1941 on a system where the size of off_t is not greater than 4, a panic error
1944 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1945 never match. The debug and -bh output now contains an explicit error
1946 message indicating a malformed IPv4 address or mask.
1948 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1949 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1950 PH/23 above applies.
1952 PH/25 Do not write to syslog when running in the test harness. The only
1953 occasion when this arises is a failure to open the main or panic logs
1954 (for which there is an explicit test).
1956 PH/26 Added the /no_tell option to "control=freeze".
1958 PH/27 If a host name lookup failed very early in a connection, for example, if
1959 the IP address matched host_lookup and the reverse lookup yielded a name
1960 that did not have a forward lookup, an error message of the form "no IP
1961 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1962 could be logged. Now it outputs the IP address instead of "NULL".
1964 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1965 allows the sender and the authenticated sender to be set when
1966 submitting a message from within Exim. Since child_open_exim() is
1967 documented for local_scan(), the new function should be too.
1969 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1970 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1971 results in an empty string is now treated as unset.
1973 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1975 PH/31 Added sender_verify_fail logging option.
1977 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1978 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1979 tidied the source and removed it altogether.
1981 PH/33 When a queue run was abandoned because the load average was too high, a
1982 log line was always written; now it is written only if the queue_run log
1983 selector is set. In addition, the log line for abandonment now contains
1984 information about the queue run such as the pid. This is always present
1985 in "start" and "stop" lines but was omitted from the "abandon" line.
1987 PH/34 Omit spaces between a header name and the colon in the error message that
1988 is given when verify = headers_syntax fails (if there are lots of them,
1989 the message gets confusing).
1991 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1992 names, as there are now some PTR records that contain slashes. This check
1993 is only to protect against broken name servers that fall over on strange
1994 characters, so the fact that it applies to all lookups doesn't matter.
1996 PH/36 Now that the new test suite is complete, we can remove some of the
1997 special code in Exim that was needed for the old test suite. For example,
1998 sorting DNS records because real resolvers return them in an arbitrary
1999 order. The new test suite's fake resolver always returns records in the
2002 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2003 bounces) except when queue_only is set, to avoid logging races between
2004 the different processes.
2006 PH/38 Panic-die if .include specifies a non-absolute path.
2008 PH/39 A tweak to the "H" retry rule from its user.
2010 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2011 a label. They prevented compilation on older perls.
2013 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2014 a warning to be raised on newish perls.
2016 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2017 on queue. Changes to match documented behaviour of showing count of
2018 messages matching specified criteria.
2020 PH/40 Changed the default ident timeout from 30s to 5s.
2022 PH/41 Added support for the use of login_cap features, on those BSD systems
2023 that have them, for controlling the resources used by pipe deliveries.
2025 PH/42 The content-scanning code uses fopen() to create files in which to put
2026 message data. Previously it was not paying any attention to the mode of
2027 the files. Exim runs with umask(0) because the rest of the code creates
2028 files with open(), and sets the required mode explicitly. Thus, these
2029 files were ending up world-writeable. This was not a big issue, because,
2030 being within the spool directory, they were not world-accessible. I have
2031 created a function called modefopen, which takes an additional mode
2032 argument. It sets umask(777), creates the file, chmods it to the required
2033 mode, then resets the umask. All the relevant calls to fopen() in the
2034 content scanning code have been changed to use this function.
2036 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2037 to 24 hours. This avoids potential overflow problems when processing G
2038 and H retry rules. I suspect nobody ever tinkers with this value.
2040 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2042 PH/45 When the plaintext authenticator is running as a client, the server's
2043 challenges are checked to ensure they are valid base64 strings. By
2044 default, the authentication attempt is cancelled if an invalid string is
2045 received. Setting client_ignore_invalid_base64 true ignores these errors.
2046 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2047 they are received. Thus, the responses can be made to depend on the
2048 challenges. If an invalid string is ignored, an empty string is placed in
2051 PH/46 Messages that are created by the autoreply transport now contains a
2052 References: header, in accordance with RFCs 2822 and 3834.
2054 PH/47 Added authenticated_sender_force to the smtp transport.
2056 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2058 PH/49 Installed latest patch from the Sieve maintainer.
2060 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2061 was also set, the appendfile transport was unnecessarily scanning a
2062 directory of message files (e.g. for maildir delivery) to find the count
2063 of files (along with the size), even though it did not need this
2064 information. It now does the scan only if it needs to find either the
2065 size of the count of files.
2067 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2069 PH/52 Two bugs concerned with error handling when the smtp transport is
2072 (i) Exim was not creating retry information for temporary errors given
2073 for individual recipients after the DATA command when the smtp transport
2074 was used in LMTP mode. This meant that they could be retried too
2075 frequently, and not timed out correctly.
2077 (ii) Exim was setting the flag that allows error details to be returned
2078 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2079 recipients that were returned after the DATA command.
2081 PH/53 This is related to PH/52, but is more general: for any failing address,
2082 when detailed error information was permitted to be returned to the
2083 sender, but the error was temporary, then after the final timeout, only
2084 "retry timeout exceeded" was returned. Now it returns the full error as
2085 well as "retry timeout exceeded".
2087 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2088 do this, and (what is worse) MTAs that accept it.
2090 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2091 will now be deprecated.
2093 PH/56 New os.c-cygwin from the Cygwin maintainer.
2095 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2096 formats (previously only available in exim formats via -bpr, -bpru,
2097 and -bpra. Now also available in native and exiqgrep formats)
2099 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2100 with very large, slow to parse queues
2102 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2104 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2106 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2107 responses to authentication challenges, though it was showing the
2108 challenges; (ii) I've removed the CR characters from the debug output for
2111 PH/58 Allow for the insertion of a newline as well as a space when a string
2112 is turned into more than one encoded-word during RFC 2047 encoding. The
2113 Sieve code now uses this.
2115 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2116 data_4xx, lost_connection, tls_required.
2118 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2119 message was being sent as an SMTP response.
2121 PH/61 Add -l and -k options to exicyclog.
2123 PH/62 When verifying, if an address was redirected to one new address, so that
2124 verification continued, and the new address failed or deferred after
2125 having set something in $address_data, the value of $address_data was not
2126 passed back to the ACL. This was different to the case when no
2127 redirection occurred. The value is now passed back in both cases.
2129 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2130 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2131 use login_cap.h, so on its own it isn't the distinguishing feature. The
2132 new name refers directly to the setclassresources() function.
2134 PH/65 Added configuration files for NetBSD3.
2136 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2138 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2139 is preferred over IPv4.
2141 PH/68 The bounce_return_message and bounce_return_body options were not being
2142 honoured for bounces generated during the reception of non-SMTP messages.
2143 In particular, this applied to messages rejected by the ACL. This bug has
2144 been fixed. However, if bounce_return_message is true and bounce_return_
2145 body is false, the headers that are returned for a non-SMTP message
2146 include only those that have been read before the error was detected.
2147 (In the case of an ACL rejection, they have all been read.)
2149 PH/69 The HTML version of the specification is now built in a directory called
2150 spec_html instead of spec.html, because the latter looks like a path with
2151 a MIME-type, and this confuses some software.
2153 PH/70 Catch two compiler warnings in sieve.c.
2155 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2156 function verify_get_ident() calls ip_connect() to connect a socket, but
2157 if the "connect()" function timed out, ip_connect() used to close the
2158 socket. However, verify_get_ident() also closes the socket later, and in
2159 between Exim writes to the log, which may get opened at this point. When
2160 the socket was closed in ip_connect(), the log could get the same file
2161 descriptor number as the socket. This naturally causes chaos. The fix is
2162 not to close the socket in ip_connect(); the socket should be closed by
2163 the function that creates it. There was only one place in the code where
2164 this was missing, in the iplookup router, which I don't think anybody now
2165 uses, but I've fixed it anyway.
2167 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2168 well as to direct DNS lookups. Otherwise the handling of names in host
2169 lists is inconsistent and therefore confusing.
2175 PH/01 Two changes to the default runtime configuration:
2177 (1) Move the checks for relay_from_hosts and authenticated clients from
2178 after to before the (commented out) DNS black list checks.
2180 (2) Add control=submission to the relay_from_hosts and authenticated
2181 clients checks, on the grounds that messages accepted by these
2182 statements are most likely to be submissions.
2184 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2186 (1) Generate an error if the third argument for the ${prvs expansion is
2189 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2192 (3) Reset the variables that are obtained from the first argument of
2193 ${prvscheck and used in the second argument before leaving the code,
2194 because their memory is reclaimed, so using them afterwards may do
2197 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2198 one (it's much easier than Tom thought :-).
2200 (5) Because of (4), we can now allow for the use of $prvscheck_result
2201 inside the third argument.
2203 PH/03 For some reason, the default setting of PATH when running a command from
2204 a pipe transport was just "/usr/bin". I have changed it to
2207 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2208 anything to be listed in the output from -bV.
2210 PH/05 When a filter generated an autoreply, the entire To: header line was
2211 quoted in the delivery log line, like this:
2213 => >A.N.Other <ano@some.domain> <original@ddress> ...
2215 This has been changed so that it extracts the operative address. There
2216 may be more than one such address. If so, they are comma-separated, like
2219 => >ano@some.domain,ona@other.domain <original@ddress> ...
2221 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2222 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2223 not being looked up in the rDNS to get a host name, Exim was showing the
2224 IP address twice in Received: lines, even though the IP addresses were
2225 identical. For example:
2227 Received: from [1.2.3.4] (helo=[1.2.3.4])
2229 However, if the real host name was known, it was omitting the HELO data
2230 if it matched the actual IP address. This has been tidied up so that it
2231 doesn't show the same IP address twice.
2233 PH/07 When both +timestamp and +memory debugging was on, the value given by
2234 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2235 called by the debug printing, thereby overwriting the timestamp buffer.
2236 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2238 PH/08 When the original message was included in an autoreply transport, it
2239 always said "this is a copy of the message, including all the headers",
2240 even if body_only or headers_only was set. It now gives an appropriate
2243 PH/09 Applied a patch from the Sieve maintainer which:
2245 o fixes some comments
2246 o adds the (disabled) notify extension core
2247 o adds some debug output for the result of if/elsif tests
2248 o points to the current vacation draft in the documentation
2249 and documents the missing references header update
2253 o fixes a bug in processing the envelope test (when testing
2254 multiple envelope elements, the last element determinted the
2257 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2258 Electronic Mail") by including:
2260 Auto-submitted: auto-generated
2262 in the messages that it generates (bounce messages and others, such as
2263 warnings). In the case of bounce messages for non-SMTP mesages, there was
2264 also a typo: it was using "Auto_submitted" (underscore instead of
2265 hyphen). Since every message generated by Exim is necessarily in response
2266 to another message, thes have all been changed to:
2268 Auto-Submitted: auto-replied
2270 in accordance with these statements in the RFC:
2272 The auto-replied keyword:
2274 - SHOULD be used on messages sent in direct response to another
2275 message by an automatic process,
2277 - MUST NOT be used on manually-generated messages,
2279 - MAY be used on Delivery Status Notifications (DSNs) and Message
2280 Disposition Notifications (MDNs),
2282 - MUST NOT be used on messages generated by automatic or periodic
2283 processes, except for messages which are automatic responses to
2286 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2287 to the default Received: header definition.
2289 PH/12 Added log selector acl_warn_skipped (default on).
2291 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2292 variables because (a) they are in the wrong storage pool and (b) even if
2293 they were copied, it wouldn't work properly because of the caching.
2295 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2296 checking when decoding. Apparently there are clients that generate
2297 overlong encoded strings. Why am I not surprised?
2299 PH/15 If the first argument of "${if match_address" was not empty, but did not
2300 contain an "@" character, Exim crashed. Now it writes a panic log message
2301 and treats the condition as false.
2303 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2305 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2306 extension "envelope-auth". The code is finished and in agreement with
2307 other implementations, but there is no documentation so far and in fact,
2308 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2309 not changing the active code.
2311 Print executed "if" and "elsif" statements when debugging is used. This
2312 helps a great deal to understand what a filter does.
2314 Document more things not specified clearly in RFC3028. I had all this
2315 sorted out, when out of a sudden new issues came to my mind. Oops."
2317 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2320 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2321 canonical form (as documented). However, after a host name lookup from
2322 the IP address, check_host() was doing a simple string comparison with
2323 addresses acquired from the DNS when checking that the found name did
2324 have the original IP as one of its addresses. Since any found IPv6
2325 addresses are likely to be in abbreviated form, the comparison could
2326 fail. Luckily, there already exists a function for doing the comparison
2327 by converting both addresses to binary, so now that is used instead of
2328 the text comparison.
2330 PH/20 There was another similar case to PH/19, when a complete host name was
2331 given in a host list; looking up its IP address could give an abbreviated
2332 form, whereas the current host's name might or might not be abbreviated.
2333 The same fix has been applied.
2339 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2340 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2343 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2344 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2346 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2348 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2349 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2350 Also, for "Auto-Submitted", treat anything other than "no" as
2351 non-personal, in accordance with RFC 3834. (Previously it treated
2352 anything starting "auto-" as non-personal.)
2354 TF/01 The control=submission/name=... option had a problem with syntax
2355 errors if the name included a slash character. The /name= option
2356 now slurps the rest of the string, so it can include any characters
2357 but it must come last in the list of options (after /sender_retain
2360 PH/05 Some modifications to the interface to the fake nameserver for the new
2368 TK/01 Added the "success_on_redirect" address verification option. See
2369 NewStuff for rationale and an example.
2371 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2373 PH/02 Patch to exigrep to allow it to work on syslog lines.
2375 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2376 fread() to skip over the body file's header line, because in Cygwin the
2377 header line is locked and is inaccessible.
2379 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2380 co-exist for some time) to make it clear that it is the Exim ID that is
2381 referenced, not the Message-ID: header line.
2383 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2384 string_format() function, because snprintf() does not exist on all
2387 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2388 ${stat: expansion item.
2390 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2391 protocol synchronization error", to keep the pedants happy.
2393 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2394 well as for IRIX systems, when gcc is being used. See the host.c source
2397 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2399 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2402 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2403 transport and to the smtp transport in LMTP mode.
2405 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2407 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2408 run a filter in a subprocess. This could lead to confusion in subsequent
2409 lookups in the parent process. There should also be a search_tidyup() at
2410 the end of the subprocess.
2412 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2413 only if the host matched helo_try_verify_hosts, which caused the
2414 verification to occur when the EHLO/HELO command was issued. The ACL just
2415 tested the remembered result. Now, if a previous verification attempt has
2416 not happened, "verify = helo" does it there and then.
2418 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2420 TK/03 Fix log output including CR from clamd.
2422 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2423 exist provoked a memory error which could cause a segfault.
2425 PH/15 Installed PCRE 6.2
2427 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2429 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2430 of the problem. Specifically, suggested +O2 rather than +O1 for the
2433 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2435 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2436 record, the port was not correctly logged when the outgoing_port log
2437 selector was set (it logged the transort's default port).
2439 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2440 fallback_hosts, and "hosts" in the smtp transport.
2442 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2443 host errors such as "Connection refused".
2445 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2446 authentication with radiusclient 0.4.9:
2448 - Error returned from rc_read_config was caught wrongly
2449 - Username/password not passed on to radius server due to wrong length.
2451 The presumption is that some radiusclient API changes for 4.51/PH/17
2452 were not taken care of correctly. The code is still untested by me (my
2453 Linux distribution still has 0.3.2 of radiusclient), but it was
2454 contributed by a Radius user.
2456 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2457 expanding the "port" option of the smtp transport.
2459 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2460 while reading a MIME header. Thanks to Tom Hughes for a patch.
2462 PH/24 Include config.h inside local_scan.h so that configuration settings are
2465 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2466 that in an ACL for RCPT (for example), you can examine exactly what was
2469 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2470 commands, but it was not correctly comparing the address with the actual
2471 client host address. Thus, it would show the EHLO address in Received:
2472 header lines when this was not necessary.
2474 PH/27 Added the % operator to ${eval:}.
2476 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2477 it should be ignoring failures (because with -C, for example, it has lost
2478 privilege). It wasn't ignoring creation failures other than "already
2481 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2484 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2485 because an input line was too long, either on its own, or by virtue of
2486 too many continuations, the temporary file was not being removed, and the
2487 return code was incorrect.
2489 PH/31 Missing "BOOL" in function definition in filtertest.c.
2491 PH/32 Applied Sieve patches from the maintainer.
2493 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2495 PH/33 Added "verify = not_blind".
2497 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2498 Local/Makefile (with some defaults set). These are used in built scripts
2499 such as exicyclog, but they have never been used in the exim_install
2500 script (though there are many overriding facilities there). I have
2501 arranged that the exim_install script now takes note of these two
2504 PH/35 Installed configuration files for Dragonfly.
2506 PH/36 When a locally submitted message by a trusted user did not contain a
2507 From: header, and the sender address was obtained from -f or from an SMTP
2508 MAIL command, and the trusted user did not use -F to supply a sender
2509 name, $originator_name was incorrectly used when constructing a From:
2510 header. Furthermore, $originator_name was used for submission mode
2511 messages from external hosts without From: headers in a similar way,
2512 which is clearly wrong.
2514 PH/37 Added control=suppress_local_fixups.
2516 PH/38 When log_selector = +received_sender was set, and the addition of the
2517 sender made the log line's construction buffer exactly full, or one byte
2518 less than full, an overflow happened when the terminating "\n" was
2521 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2522 when the result of a list match is failure because a DNS lookup failed.
2524 PH/40 RM_COMMAND is now used in the building process.
2526 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2527 the "build-* directories that it finds.
2529 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2530 address in a domain literal was a prefix of an interface address.
2532 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2533 when verifying a sender address, unless rewrite_headers is false.
2535 PH/44 Wrote a long comment about why errors_to addresses are verified as
2536 recipients, not senders.
2538 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2539 the ratelimit ACL was added.
2541 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2543 PH/46 Added extra information about PostgreSQL errors to the error string.
2545 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2546 avoiding the need to install special zones in a real server. This is
2547 backwards compatible; if it can't find the fake resolver, it drops back.
2548 Thus, both old and new test suites can be run.
2550 TF/02 Added util/ratelimit.pl
2552 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2553 clock is set back in time.
2555 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2556 Candler <B.Candler@pobox.com>.
2558 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2559 OK for addresses that are the result of redirections.
2561 PH/48 A number of further additions for the benefit of the new test suite,
2562 including a fake gethostbyname() that interfaces to the fake DNS resolver
2565 TF/06 The fix for widen_domains has also been applied to qualify_single and
2566 search_parents which are the other dnslookup options that can cause
2569 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2572 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2574 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2575 service" if it encountered an MX record with an empty target hostname.
2576 The message is now "an MX or SRV record indicated no SMTP service".
2578 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2579 if the DNS of the sending site is misconfigured. This is quite a
2580 common situation. This change restores the behaviour of treating a
2581 helo verification defer as a failure.
2583 PH/51 If self=fail was set on a router, the bounce message did not include the
2584 actual error message.
2590 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2592 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2593 command itself ended in error, the underlying message about the transport
2594 filter timeout was being overwritten with the pipe command error. Now the
2595 underlying error message should be appended to the second error message.
2597 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2599 PH/02 Reduce the amount of output that "make" produces by default. Full output
2600 can still be requested.
2602 PH/03 The warning log line about a condition test deferring for a "warn" verb
2603 was being output only once per connection, rather than after each
2604 occurrence (because it was using the same function as for successful
2605 "warn" verbs). This seems wrong, so I have changed it.
2607 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2608 it should not have, which might have caused a crash in the right
2609 circumstances, but probably never did.
2611 PH/04 Installed a modified version of Tony Finch's patch to make submission
2612 mode fix the return path as well as the Sender: header line, and to
2613 add a /name= option so that you can make the user's friendly name appear
2616 TF/03 Added the control = fakedefer ACL modifier.
2618 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2619 Mark Lowes for thorough testing.
2621 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2623 TK/03 Merged latest SRS patch from Miles Wilton.
2625 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2626 with the definition in sysexits.h (which is #included earlier).
2627 Fortunately, Exim does not actually use EX_OK. The code used to try to
2628 preserve the sysexits.h value, by assumimg that macro definitions were
2629 scanned for macro replacements. I have been disabused of this notion,
2630 so now the code just undefines EX_OK before #including unistd.h.
2632 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2633 in the smtp transport. When a block could not be written in a single
2634 write() function, the timeout was being re-applied to each part-write.
2635 This seems wrong - if the receiver was accepting one byte at a time it
2636 would take for ever. The timeout is now adjusted when this happens. It
2637 doesn't have to be particularly precise.
2639 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2640 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2642 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2643 without a domain if the check to <postmaster@domain> fails.
2645 SC/01 Eximstats: added -xls and the ability to specify output files
2646 (patch written by Frank Heydlauf).
2648 SC/02 Eximstats: use FileHandles for outputing results.
2650 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2652 SC/04 Eximstats: fixed display of large numbers with -nvr option
2654 SC/05 Eximstats: fixed merging of reports with empty tables.
2656 SC/06 Eximstats: added the -include_original_destination flag
2658 SC/07 Eximstats: removed tabs and trailing whitespace.
2660 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2662 TK/06 MBOX spool code: Add real "From " MBOX separator line
2663 so the .eml file is really in mbox format (even though
2664 most programs do not really care). Patch from Alex Miller.
2666 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2667 The latter is generated from $received_to and is only set if the
2668 message has one envelope recipient. SA can use these headers,
2669 obviously out-of-the-box. Patch from Alex Miller.
2671 PH/08 The ${def test on a variable was returning false if the variable's
2672 value was "0", contrary to what the specification has always said!
2673 The result should be true unless the variable is empty.
2675 PH/09 The syntax error of a character other than { following "${if
2676 def:variable_name" (after optional whitespace) was not being diagnosed.
2677 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2678 accidental colon was present, for example, could give incorrect results.
2680 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2681 result is used (not including appendfile, where other changes are about
2684 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2685 This involved changing a lot of size variables from int to off_t. It
2686 should work with maildirs and everything.
2688 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2689 spamd dying while we are connected to it.
2691 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2694 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2695 function for BDB 4.3.
2697 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2699 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2700 the sources for PCRE have been re-arranged and more files are now
2703 PH/15 The code I had for printing potentially long long variables in PH/11
2704 above was not the best (it lost precision). The length of off_t variables
2705 is now inspected at build time, and an appropriate printing format (%ld
2706 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2707 to be "long long int" or "long int". This is needed for the internal
2708 formatting function string_vformat().
2710 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2711 the configuration file to be ":syslog", then the script "guesses" where
2712 the logs files are, rather than using the compiled in default. In our
2713 case the guess is not the same as the compiled default, so the script
2714 suddenly stopped working when I started to use syslog. The patch checks
2715 to see if log_file_path is "". If so, it attempts to read it from exim
2716 with no configuration file to get the compiled in version, before it
2717 falls back to the previous guessing code."
2719 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2720 implementing BATV in an Exim configuration. See NewStuff for the gory
2723 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2724 Makefile that are specific to HP-UX.
2726 PH/18 If the "use_postmaster" option was set for a recipient callout together
2727 with the "random" option, the postmaster address was used as the MAIL
2728 FROM address for the random test, but not for the subsequent recipient
2729 test. It is now used for both.
2731 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2732 patch removes a few documentation additions to RFC 3028, because the
2733 latest draft now contains them. It adds the new en;ascii-case comparator
2734 and a new error check for 8bit text in MIME parts. Comparator and
2735 require names are now matched exactly. I enabled the subaddress
2736 extension, but it is not well tested yet (read: it works for me)."
2738 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2739 rework some of the code of TK/09 above to avoid the hardwired use of
2740 "%lld" and "long long". Replaced the call to snprintf() with a call to
2743 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2744 records point to non-existent hosts", "retry timeout exceeded", and
2745 "retry time not reached for any host after a long failure period".
2747 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2748 experimental DomainKeys support:
2750 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2751 (2) On an error such as an illegally used "control", the wrong name for
2752 the control was given.
2754 These problems did NOT occur unless DomainKeys support was compiled.
2756 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2758 PH/24 Added ${if match_ip condition.
2760 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2761 databases so that it will be absolutely obvious if a crash occurs in the
2762 DB library. This is a regular occurrence (often caused by mis-matched
2765 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2766 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2767 picked up on a user's system that detects such things. There doesn't seem
2768 to be a gcc warning option for this - only an attribute that has to be
2769 put on the function's prototype. It seems that in Fedora Core 4 they have
2770 set this on a number of new functions. No doubt there will be more in due
2773 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2774 specify a transport. However, if an address that was verified by such a
2775 router was the subject of a callout, Exim crashed because it tried to
2776 read the rcpt_include_affixes from the non-existent transport. Now it
2777 just assumes that the setting of that option is false. This bug was
2778 introduced by 4.51/PH/31.
2780 PH/28 Changed -d+all to exclude +memory, because that information is very
2781 rarely of interest, but it makes the output a lot bigger. People tend to
2782 do -d+all out of habit.
2784 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2785 code in os-type was giving problems when libc.so lives in lib64, like on
2788 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2789 aren't the modern standard, and it seems that some systems' include files
2790 don't always have them. Exim was already checking for some of the newer
2791 ones like T_AAAA, and defining it itself. I've added checks for all the
2792 record types that Exim uses.
2794 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2795 not automatically generating a new one, as it is supposed to. This
2796 prevented TLS from working. If the file did exist, but contained invalid
2797 data, a new version was generated, as expected. It was only the case of a
2798 non-existent file that was broken.
2800 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2801 with a change in libdomainkeys > 0.64.
2803 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2804 from DNS. If the selector record carries the flag, it now has
2805 precedence over the domain-wide flag.
2807 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2809 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2810 the use of a port name that isn't defined in /etc/services) occurred, the
2811 message was deferred as in a normal delivery, and thus remained on the
2812 spool, instead of being failed because of the mua_wrapper setting. This
2813 is now fixed, and I tidied up some of the mua_wrapper messages at the
2816 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2817 the messages in a hash of arrays rather than using individual hashes.
2818 This is a bit cleaner and results in dramatic memory savings, albeit
2819 at a slight CPU cost.
2821 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2822 as requested by Marc Sherman.
2824 SC/10 Eximstats: added histograms for user specified patterns as requested
2827 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2829 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2830 fopen() in the content-scanning modules that did not already have it.
2836 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2837 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2839 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2841 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2843 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2844 to test Sieve filters that use "vacation".
2846 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2847 that changes the way the GnuTLS parameters are stored in the cache file.
2848 The new format can be generated externally. For backward compatibility,
2849 if the data in the cache doesn't make sense, Exim assumes it has read an
2850 old-format file, and it generates new data and writes a new file. This
2851 means that you can't go back to an older release without removing the
2854 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2855 work if there are any delivery delays because "one_time" forces the
2856 parent to be marked "delivered", so its unseen clone is never tried
2857 again. For this reason, Exim now forbids the simultaneous setting of
2860 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2861 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2862 entry if you want to know the details. The fix, however, neglected to
2863 consider the case when local delivery batching is involved. The test for
2864 "previously delivered" was not happening when checking to see if an
2865 address could be batched with a previous (undelivered) one; under
2866 certain circumstances this could lead to multiple deliveries to the same
2869 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2870 in its include files, and this causes problems building Exim.
2872 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2873 header_syntax) but Exim was just ignoring anything given after a slash.
2874 In particular, this caused confusion with an attempt to use "verify =
2875 reverse_host_lookup/defer_ok". An error is now given when options are
2876 supplied for verify items that do not have them. (Maybe reverse_host_
2877 lookup should have a defer_ok option, but that's a different point.)
2879 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2880 defined by RFC 821) to 2048, because there were problems with some AUTH
2881 commands, and RFC 1869 says the size should be increased for extended
2882 SMTP commands that take arguments.
2884 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2887 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2888 "unknown" error; now it says that the functionality isn't in the binary.
2890 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2891 an address' error message when a string expansion fails (syntax or
2892 whatever). Otherwise the password may appear in the log. Following change
2893 PH/42 below, there is no longer a chance of it appearing in a bounce
2896 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2898 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2899 of its addresses was ever tried. (Bugzilla bug #2.)
2901 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2902 the result incorrectly in the debug output. (It correctly added a newline
2903 to what was transported.)
2905 TF/01 Added $received_time.
2907 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2908 commented out examples of how to interface to a virus scanner and to
2909 SpamAssassin. Also added commented examples of av_scanner and
2910 spamd_address settings.
2912 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2913 and controls are allowed in which ACLs. There were a couple of minor
2914 errors. Some of the entries in the conditions table (which is a table of
2915 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2916 negation of where the condition IS allowed.
2918 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2920 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2921 header file does not have a version number, so I've had to invent a new
2922 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2923 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2924 radiusclient), but it was contributed by a Radius user.
2926 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2927 files or directories, for OpenSSL.
2929 PH/19 When an Exim process that is running as root has to create an Exim log
2930 file, it does so in a subprocess that runs as exim:exim so as to get the
2931 ownership right at creation (otherwise, other Exim processes might see
2932 the file with the wrong ownership). There was no test for failure of this
2933 fork() call, which would lead to the process getting stuck as it waited
2934 for a non-existent subprocess. Forks do occasionally fail when resources
2935 run out. I reviewed all the other calls to fork(); they all seem to check
2938 PH/20 When checking for unexpected SMTP input at connect time (before writing
2939 the banner), Exim was not dealing correctly with a non-positive return
2940 from the read() function. If the client had disconnected by this time,
2941 the result was a log entry for a synchronization error with an empty
2942 string after "input=" when read() returned zero. If read() returned -1
2943 (an event I could not check), uninitialized data bytes were printed.
2944 There were reports of junk text (parts of files, etc) appearing after
2947 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2949 PH/22 Added support for macro redefinition, and (re)definition in between
2950 driver and ACL definitions.
2952 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2953 forgetting to use the resulting value; it was using the unexpanded value.
2955 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2956 hadn't been configured. The fix is from Juergen Kreileder, who
2957 understands it better than I do:
2959 "Here's what I see happening with three configured cyrus_sasl
2960 authenticators configured (plain, login, cram-md5):
2962 On startup auth_cyrus_sasl_init() gets called for each of these.
2963 This means three calls to sasl_listmech() without a specified mech_list.
2964 => SASL tests which mechs of all available mechs actually work
2965 => three warnings about OTP not working
2966 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2968 With the patch, sasl_listmech() also gets called three times. But now
2969 SASL's mech_list option is set to the server_mech specified in the the
2970 authenticator. Or in other words, the answer from sasl_listmech()
2971 gets limited to just the mech you're testing for (which is different
2973 => the return list contains just 'plain' or 'login', 'cram-md5' or
2974 nothing depending on the value of ob->server_mech.
2976 I've just tested the patch: Authentication still works fine,
2977 unavailable mechs specified in the exim configuration are still
2978 caught, and the auth.log warnings about OTP are gone."
2980 PH/25 When debugging is enabled, the contents of the command line are added
2981 to the debugging output, even when log_selector=+arguments is not
2984 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2985 answer is "GNU", and only if the return is "GNU/something" is the answer
2988 PH/27 $acl_verify_message is now set immediately after the failure of a
2989 verification in an ACL, and so is available in subsequent modifiers. In
2990 particular, the message can be preserved by coding like this:
2992 warn !verify = sender
2993 set acl_m0 = $acl_verify_message
2995 Previously, $acl_verify_message was set only while expanding "message"
2996 and "log_message" when a very denied access.
2998 PH/28 Modified OS/os.c-Linux with
3000 -#ifndef OS_LOAD_AVERAGE
3001 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3003 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3004 nomenclature these days.)
3006 PH/29 Installed patch from the Sieve maintainer that adds the options
3007 sieve_useraddress and sieve_subaddress to the redirect router.
3009 PH/30 In these circumstances:
3010 . Two addresses routed to the same list of hosts;
3011 . First host does not offer TLS;
3012 . First host accepts first address;
3013 . First host gives temporary error to second address;
3014 . Second host offers TLS and a TLS session is established;
3015 . Second host accepts second address.
3016 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3017 and peerdn, if requested) that were in fact used only for the second
3020 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3021 attention to any local part prefix or suffix that was matched by the
3022 router that accepted the address. It now behaves in the same way as it
3023 does for delivery: the affixes are removed from the local part unless
3024 rcpt_include_affixes is set on the transport.
3026 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3027 timeout during the DATA phase of an incoming message.
3029 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3030 applied a patch sanctioned by the Sieve maintainer.
3032 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3033 the uid or gid is negative. A case of a negative gid caused this to be
3034 noticed. The fix allows for either to be negative.
3036 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3037 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3040 PH/36 The change PH/12 above was broken. Fixed it.
3042 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3043 the grounds that routing the same address twice would always produce the
3044 same answer. This might have been true once, but it is certainly no
3045 longer true now. Routing a child address may depend on the previous
3046 routing that produced that child. Some complicated redirection strategies
3047 went wrong when messages had multiple recipients, and made Exim's
3048 behaviour dependent on the order in which the addresses were given.
3050 I have moved the duplicate checking until after the routing is complete.
3051 Exim scans the addresses that are assigned to local and remote
3052 transports, and removes any duplicates. This means that more work will be
3053 done, as duplicates will always all be routed, but duplicates are
3054 presumably rare, so I don't expect this is of any significance.
3056 For deliveries to pipes, files, and autoreplies, the duplicate checking
3057 still happens during the routing process, since they are not going to be
3060 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3061 It corrects a timeout issue with spamd. This is Ian's comment: "The
3062 background is that sometimes spamd either never reads data from a
3063 connection it has accepted, or it never writes response data. The exiscan
3064 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3065 blindly assumes that writes won't block so it may never time out."
3067 PH/39 Allow G after quota size as well as K and M.
3069 PH/40 The value set for $authenticated_id in an authenticator may not contain
3070 binary zeroes or newlines because the value is written to log lines and
3071 to spool files. There was no check on this. Now the value is run through
3072 the string_printing() function so that such characters are converted to
3073 printable escape sequences.
3075 PH/41 $message_linecount is a new variable that contains the total number of
3076 lines in the message. Compare $body_linecount, which is the count for the
3079 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3080 bounce and delay warning messages, except in certain special cases, which
3083 (a) An SMTP error message from a remote host;
3084 (b) A message specified in a :fail: redirection;
3085 (c) A message specified in a "fail" command in a system filter;
3086 (d) A message specified in a FAIL return from the queryprogram router;
3087 (e) A message specified by the cannot_route_message router option.
3089 In these cases only, Exim does include the error details in bounce and
3090 warning messages. There are also a few cases where bland messages such
3091 as "unrouteable address" or "local delivery error" are given.
3093 PH/43 $value is now also set for the "else" part of a ${run expansion.
3095 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3096 being worked on, but at least Exim now implements the latest version to
3099 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3100 process to complete was treated as a delivery failure, a timeout while
3101 writing the message to the pipe was logged, but erroneously treated as a
3102 successful delivery. Such timeouts include transport filter timeouts. For
3103 consistency with the overall process timeout, these timeouts are now
3104 treated as errors, giving rise to delivery failures by default. However,
3105 there is now a new Boolean option for the pipe transport called
3106 timeout_defer, which, if set TRUE, converts the failures into defers for
3107 both kinds of timeout. A transport filter timeout is now identified in
3110 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3111 systems where "make" and "gmake" are different, calling "gmake" at top
3112 level broke things. I've arranged for the value of $(MAKE) to be passed
3113 from the Makefile to this script so that it can call the same version of
3117 A note about Exim versions 4.44 and 4.50
3118 ----------------------------------------
3120 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3121 changes of various kinds. As a consequence, a big documentation update was
3122 needed. This delayed the release for rather longer than seemed good, especially
3123 in the light of a couple of (minor) security issues. Therefore, the changes
3124 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3125 release. So 4.44 and 4.50 are in effect two different branches that both start
3128 I have left the 4.50 change log unchanged; it contains all the changes since
3129 4.43. The change log for 4.44 is below; many of its items are identical to
3130 those for 4.50. This seems to be the most sensible way to preserve the
3131 historical information.
3137 1. Minor wording change to the doc/README.SIEVE file.
3139 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3140 computation of the current number of files was incorrect.
3142 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3143 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3146 4. Give more explanation in the error message when the command for a transport
3147 filter fails to execute.
3149 5. There are several places where Exim runs a non-Exim command in a
3150 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3151 was being done only for the command run by the queryprogram router. It is
3152 now done for all such subprocesses. The other cases are: ${run, transport
3153 filters, and the commands run by the lmtp and pipe transports.
3155 6. Added CONFIGURE_GROUP build-time option.
3157 7. Some older OS have a limit of 256 on the maximum number of file
3158 descriptors. Exim was using setrlimit() to set 1000 as a large value
3159 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3160 systems. I've change it so that if it can't get 1000, it tries for 256.
3162 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3163 was an oversight, and furthermore, ever since the addition of extra
3164 controls (e.g. 4.43/32), the checks on when to allow different forms of
3165 "control" were broken. There should now be diagnostics for all cases when a
3166 control that does not make sense is encountered.
3168 9. Added the /retain_sender option to "control=submission".
3170 10. $recipients is now available in the predata ACL (oversight).
3172 11. Tidy the search cache before the fork to do a delivery from a message
3173 received from the command line. Otherwise the child will trigger a lookup
3174 failure and thereby defer the delivery if it tries to use (for example) a
3175 cached ldap connection that the parent has called unbind on.
3177 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3178 of $address_data from the recipient verification was clobbered by the
3179 sender verification.
3181 13. The value of address_data from a sender verification is now available in
3182 $sender_address_data in subsequent conditions in the ACL statement.
3184 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3186 15. Added a new option "connect=<time>" to callout options, to set a different
3189 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3190 was its contents. (It was OK if the option was not defined at all.)
3192 17. A "Completed" log line is now written for messages that are removed from
3193 the spool by the -Mrm option.
3195 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3196 information about exactly what failed.
3198 19. Added -dd to debug only the daemon process.
3200 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3201 handles timeouts, both on the server side and network timeouts. Renamed the
3202 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3204 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3205 It is now set to "smtps".
3207 22. $host_address is now set to the target address during the checking of
3208 ignore_target_hosts.
3210 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3211 being passed; this would have caused $sender_host_name to have been used if
3212 matching the list had actually called for a host name (not very likely,
3213 since this list is usually IP addresses). A host name is now passed as
3216 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3217 code that specifies a non-restarting handler (typically sigaction() in
3218 modern systems) in an attempt to fix a rare and obscure crash bug.
3220 25. Narrowed the window for a race in the daemon that could cause it to ignore
3221 SIGCHLD signals. This is not a major problem, because they are used only to
3222 wake it up if nothing else does.
3224 26. A malformed maildirsize file could cause Exim to calculate negative values
3225 for the mailbox size or file count. Odd effects could occur as a result.
3226 The maildirsize information is now recalculated if the size or filecount
3229 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3230 support for a long time. Removed HAVE_SYS_VFS_H.
3232 28. Installed the latest version of exipick from John Jetmore.
3234 29. In an address list, if the pattern was not a regular expression, an empty
3235 subject address (from a bounce message) matched only if the pattern was an
3236 empty string. Non-empty patterns were not even tested. This was the wrong
3237 because it is perfectly reasonable to use an empty address as part of a
3238 database query. An empty address is now tested by patterns that are
3239 lookups. However, all the other forms of pattern expect the subject to
3240 contain a local part and a domain, and therefore, for them, an empty
3241 address still always fails if the pattern is not itself empty.
3243 30. Exim went into a mad DNS loop when attempting to do a callout where the
3244 host was specified on an smtp transport, and looking it up yielded more
3245 than one IP address.
3247 31. Re-factored the code for checking spool and log partition space into a
3248 function that finds that data and another that does the check. The former
3249 is then used to implement four new variables: $spool_space, $log_space,
3250 $spool_inodes, and $log_inodes.
3252 32. The RFC2047 encoding function was originally intended for short strings
3253 such as real names; it was not keeping to the 75-character limit for
3254 encoded words that the RFC imposes. It now respects the limit, and
3255 generates multiple encoded words if necessary. To be on the safe side, I
3256 have increased the buffer size for the ${rfc2047: expansion operator from
3259 33. It is now permitted to omit both strings after an "if" condition; if the
3260 condition is true, the result is "true". As before, when the second string
3261 is omitted, a false condition yields an empty string. This makes it less
3262 cumbersome to write custom ACL and router conditions.
3264 34. Failure to deliver a bounce message always caused it to be frozen, even if
3265 there was an errors_to setting on the router. The errors_to setting is now
3268 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3269 canonical form (fully expanded) before being placed in
3270 $sender_host_address.
3272 36. The table in the code that translates DNS record types into text (T_A to
3273 "A" for instance) was missing entries for NS and CNAME. It is just possible
3274 that this could have caused confusion if both these types were looked up
3275 for the same domain, because the text type is used as part of Exim's
3276 per-process caching. But the chance of anyone hitting this buglet seems
3279 37. The dnsdb lookup has been extended in a number of ways.
3281 (1) There is a new type, "zns", which walks up the domain tree until it
3282 finds some nameserver records. It should be used with care.
3284 (2) There is a new type, "mxh", which is like "mx" except that it returns
3285 just the host names, not the priorities.
3287 (3) It is now possible to give a list of domains (or IP addresses) to be
3288 looked up. The behaviour when one of the lookups defers can be
3289 controlled by a keyword.
3291 (4) It is now possible to specify the separator character for use when
3292 multiple records are returned.
3294 38. The dnslists ACL condition has been extended: it is now possible to supply
3295 a list of IP addresses and/or domains to be looked up in a particular DNS
3298 39. Added log_selector=+queue_time_overall.
3300 40. When running the queue in the test harness, wait just a tad after forking a
3301 delivery process, to get repeatability of debugging output.
3303 41. Include certificate and key file names in error message when GnuTLS fails
3304 to set them up, because the GnuTLS error message doesn't include the name
3305 of the failing file when there is a problem reading it.
3307 42. Allow both -bf and -bF in the same test run.
3309 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3311 44. The "Exiscan patch" is now merged into the mainline Exim source.
3313 45. Sometimes the final signoff response after QUIT could fail to get
3314 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3315 < 0 before doing a fflush(). This bug looks as though it goes back to the
3316 introduction of TLS in release 3.20, but "sometimes" must have been rare
3317 because the tests only now provoked it.
3319 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3320 (this can affect the format of dates).
3322 47. exim_tidydb, when checking for the continued existence of a message for
3323 which it has found a message-specific retry record, was not finding
3324 messages that were in split spool directories. Consequently, it was
3325 deleting retry records that should have stayed in existence.
3327 48. Steve fixed some bugs in eximstats.
3329 49. The SPA authentication driver was not abandoning authentication and moving
3330 on to the next authenticator when an expansion was forced to fail,
3331 contradicting the general specification for all authenticators. Instead it
3332 was generating a temporary error. It now behaves as specified.
3334 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3335 (the order specifies the preference for clients). The order is now AES256,
3336 AES128, 3DES, ARCFOUR128.
3338 51. Small patch to Sieve code - explicitly set From: when generating an
3341 52. Exim crashed if a remote delivery caused a very long error message to be
3342 recorded - for instance if somebody sent an entire SpamAssassin report back
3343 as a large number of 550 error lines. This bug was coincidentally fixed by
3344 increasing the size of one of Exim's internal buffers (big_buffer) that
3345 happened as part of the Exiscan merge. However, to be on the safe side, I
3346 have made the code more robust (and fixed the comments that describe what
3349 53. Now that there can be additional text after "Completed" in log lines (if
3350 the queue_time_overall log selector is set), a one-byte patch to exigrep
3351 was needed to allow it to recognize "Completed" as not the last thing in
3354 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3355 patch that reportedly fixes this has been added. I am not expert enough to
3356 create a test for it. This is what the patch creator wrote:
3358 "I found a little strange behaviour of ldap code when working with
3359 Windows 2003 AD Domain, where users was placed in more than one
3360 Organization Units. When I tried to give exim partial DN, the exit code
3361 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3362 But simultaneously result of request was absolutely normal ldap result,
3363 so I produce this patch..."
3365 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3366 REFERENCE, so I have modified the code to exclude the patch when that macro
3369 55. Some experimental protocols are using DNS PTR records for new purposes. The
3370 keys for these records are domain names, not reversed IP addresses. The
3371 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3372 leaves it alone. Component reversal etc. now happens only for IP addresses.
3375 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3377 57. Double the size of the debug message buffer (to 2048) so that more of very
3378 long debug lines gets shown.
3380 58. The exicyclog utility now does better if the number of log files to keep
3381 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3383 59. Two changes related to the smtp_active_hostname option:
3385 (1) $smtp_active_hostname is now available as a variable.
3386 (2) The default for smtp_banner uses $smtp_active_hostname instead
3387 of $primary_hostname.
3389 60. The host_aton() function is supposed to be passed a string that is known
3390 to be a valid IP address. However, in the case of IPv6 addresses, it was
3391 not checking this. This is a hostage to fortune. Exim now panics and dies
3392 if the condition is not met. A case was found where this could be provoked
3393 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3394 components; fortuitously, this particular loophole had already been fixed
3395 by change 4.50/55 above.
3397 If there are any other similar loopholes, the new check in host_aton()
3398 itself should stop them being exploited. The report I received stated that
3399 data on the command line could provoke the exploit when Exim was running as
3400 exim, but did not say which command line option was involved. All I could
3401 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3402 running as the user.
3405 61. There was a buffer overflow vulnerability in the SPA authentication code
3406 (which came originally from the Samba project). I have added a test to the
3407 spa_base64_to_bits() function which I hope fixes it.
3410 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3411 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3413 63. The daemon start-up calls getloadavg() while still root for those OS that
3414 need the first call to be done as root, but it missed one case: when
3415 deliver_queue_load_max is set with deliver_drop_privilege. This is
3416 necessary for the benefit of the queue runner, because there is no re-exec
3417 when deliver_drop_privilege is set.
3419 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3420 This has been fixed.
3422 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3423 was in use, was not putting the data itself into the right store pool;
3424 consequently, it could be overwritten for a subsequent message in the same
3425 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3428 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3430 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3431 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3432 the function were treating the return as a boolean value, which happened to
3433 work because 0=false and not-0=true, but is not correct code.
3435 68. The host_aton() function was not handling scoped IPv6 addresses (those
3436 with, for example, "%eth0" on the end) correctly.
3438 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3439 negated items (that is, ~something) in unsigned ints. Some compilers
3440 apparently mutter when there is no cast.
3442 70. If an address verification called from an ACL failed, and did not produce a
3443 user-specific message (i.e. there was only a "system" message), nothing was
3444 put in $acl_verify_message. In this situation, it now puts the system
3447 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3448 session; change 4.31/43 added the unwanted input to the log line - except
3449 that it did not do this in the start of session case. It now does.
3451 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3452 This is wrong and can cause the other end to generate a synchronization
3453 error if it is another Exim or anything else that does the synchronization
3454 check. A QUIT command is no longer sent after a timeout.
3456 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3457 during host lookups.
3459 74. The defer_ok option of callout verification was not working if it was used
3460 when verifying addresses in header lines, that is, for this case:
3462 verify = header_sender/callout=defer_ok
3464 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3465 those file descriptors could be used for SMTP connections. If anything
3466 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3467 could be sent to the SMTP client, causing chaos. The daemon now opens
3468 stdin, stdout, and stderr to /dev/null when it puts itself into the
3471 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3472 log by default. The user can override this with suitable Perl magic.
3474 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3475 the log message when discard triggers, was not working for the DATA ACL or
3476 for the non-SMTP ACL.
3478 78. Error message wording change in sieve.c.
3480 79. If smtp_accept_max_per_host was set, the number of connections could be
3481 restricted to fewer than expected, because the daemon was trying to set up
3482 a new connection before checking whether the processes handling previous
3483 connections had finished. The check for completed processes is now done
3484 earlier. On busy systems, this bug wouldn't be noticed because something
3485 else would have woken the daemon, and it would have reaped the completed
3488 80. If a message was submitted locally by a user whose login name contained one
3489 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3490 It caused a spool format error. I have fixed the spool reading code. A
3491 related problem was that the "from" clause in the Received: line became
3492 illegal because of the space(s). It is now covered by ${quote_local_part.
3494 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3497 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3499 83. Patch from Sieve maintainer to fix unterminated string problem in
3500 "vacation" handling.
3502 84. Some minor changes to the Linux configuration files to help with other
3503 OS variants using glibc.
3505 85. One more patch for Sieve to update vacation handling to latest spec.
3508 ----------------------------------------------------
3509 See the note above about the 4.44 and 4.50 releases.
3510 ----------------------------------------------------
3516 1. Change 4.43/35 introduced a bug that caused file counts to be
3517 incorrectly computed when quota_filecount was set in an appendfile
3520 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3521 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3524 3. Give more explanation in the error message when the command for a transport
3525 filter fails to execute.
3527 4. There are several places where Exim runs a non-Exim command in a
3528 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3529 was being done only for the command run by the queryprogram router. It is
3530 now done for all such subprocesses. The other cases are: ${run, transport
3531 filters, and the commands run by the lmtp and pipe transports.
3533 5. Some older OS have a limit of 256 on the maximum number of file
3534 descriptors. Exim was using setrlimit() to set 1000 as a large value
3535 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3536 systems. I've change it so that if it can't get 1000, it tries for 256.
3538 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3539 was an oversight, and furthermore, ever since the addition of extra
3540 controls (e.g. 4.43/32), the checks on when to allow different forms of
3541 "control" were broken. There should now be diagnostics for all cases when a
3542 control that does not make sense is encountered.
3544 7. $recipients is now available in the predata ACL (oversight).
3546 8. Tidy the search cache before the fork to do a delivery from a message
3547 received from the command line. Otherwise the child will trigger a lookup
3548 failure and thereby defer the delivery if it tries to use (for example) a
3549 cached ldap connection that the parent has called unbind on.
3551 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3552 of $address_data from the recipient verification was clobbered by the
3553 sender verification.
3555 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3556 was its contents. (It was OK if the option was not defined at all.)
3558 11. A "Completed" log line is now written for messages that are removed from
3559 the spool by the -Mrm option.
3561 12. $host_address is now set to the target address during the checking of
3562 ignore_target_hosts.
3564 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3565 being passed; this would have caused $sender_host_name to have been used if
3566 matching the list had actually called for a host name (not very likely,
3567 since this list is usually IP addresses). A host name is now passed as
3570 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3571 code that specifies a non-restarting handler (typically sigaction() in
3572 modern systems) in an attempt to fix a rare and obscure crash bug.
3574 15. Narrowed the window for a race in the daemon that could cause it to ignore
3575 SIGCHLD signals. This is not a major problem, because they are used only to
3576 wake it up if nothing else does.
3578 16. A malformed maildirsize file could cause Exim to calculate negative values
3579 for the mailbox size or file count. Odd effects could occur as a result.
3580 The maildirsize information is now recalculated if the size or filecount
3583 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3584 support for a long time. Removed HAVE_SYS_VFS_H.
3586 18. Updated exipick to current release from John Jetmore.
3588 19. Allow an empty sender to be matched against a lookup in an address list.
3589 Previously the only cases considered were a regular expression, or an
3592 20. Exim went into a mad DNS lookup loop when doing a callout where the
3593 host was specified on the transport, if the DNS lookup yielded more than
3596 21. The RFC2047 encoding function was originally intended for short strings
3597 such as real names; it was not keeping to the 75-character limit for
3598 encoded words that the RFC imposes. It now respects the limit, and
3599 generates multiple encoded words if necessary. To be on the safe side, I
3600 have increased the buffer size for the ${rfc2047: expansion operator from
3603 22. Failure to deliver a bounce message always caused it to be frozen, even if
3604 there was an errors_to setting on the router. The errors_to setting is now
3607 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3608 canonical form (fully expanded) before being placed in
3609 $sender_host_address.
3611 24. Updated eximstats to version 1.33
3613 25. Include certificate and key file names in error message when GnuTLS fails
3614 to set them up, because the GnuTLS error message doesn't include the name
3615 of the failing file when there is a problem reading it.
3617 26. Expand error message when OpenSSL has problems setting up cert/key files.
3620 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3621 (this can affect the format of dates).
3623 28. exim_tidydb, when checking for the continued existence of a message for
3624 which it has found a message-specific retry record, was not finding
3625 messages that were in split spool directories. Consequently, it was
3626 deleting retry records that should have stayed in existence.
3628 29. eximstats updated to version 1.35
3629 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3630 1.35 - bugfix such that pie charts by volume are generated correctly
3632 30. The SPA authentication driver was not abandoning authentication and moving
3633 on to the next authenticator when an expansion was forced to fail,
3634 contradicting the general specification for all authenticators. Instead it
3635 was generating a temporary error. It now behaves as specified.
3637 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3638 (the order specifies the preference for clients). The order is now AES256,
3639 AES128, 3DES, ARCFOUR128.
3641 31. Small patch to Sieve code - explicitly set From: when generating an
3644 32. Exim crashed if a remote delivery caused a very long error message to be
3645 recorded - for instance if somebody sent an entire SpamAssassin report back
3646 as a large number of 550 error lines. This bug was coincidentally fixed by
3647 increasing the size of one of Exim's internal buffers (big_buffer) that
3648 happened as part of the Exiscan merge. However, to be on the safe side, I
3649 have made the code more robust (and fixed the comments that describe what
3652 33. Some experimental protocols are using DNS PTR records for new purposes. The
3653 keys for these records are domain names, not reversed IP addresses. The
3654 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3655 leaves it alone. Component reversal etc. now happens only for IP addresses.
3658 34. The host_aton() function is supposed to be passed a string that is known
3659 to be a valid IP address. However, in the case of IPv6 addresses, it was
3660 not checking this. This is a hostage to fortune. Exim now panics and dies
3661 if the condition is not met. A case was found where this could be provoked
3662 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3663 components; fortuitously, this particular loophole had already been fixed
3664 by change 4.50/55 or 4.44/33 above.
3666 If there are any other similar loopholes, the new check in host_aton()
3667 itself should stop them being exploited. The report I received stated that
3668 data on the command line could provoke the exploit when Exim was running as
3669 exim, but did not say which command line option was involved. All I could
3670 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3671 running as the user.
3674 35. There was a buffer overflow vulnerability in the SPA authentication code
3675 (which came originally from the Samba project). I have added a test to the
3676 spa_base64_to_bits() function which I hope fixes it.
3679 36. The daemon start-up calls getloadavg() while still root for those OS that
3680 need the first call to be done as root, but it missed one case: when
3681 deliver_queue_load_max is set with deliver_drop_privilege. This is
3682 necessary for the benefit of the queue runner, because there is no re-exec
3683 when deliver_drop_privilege is set.
3685 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3686 was in use, was not putting the data itself into the right store pool;
3687 consequently, it could be overwritten for a subsequent message in the same
3688 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3691 38. Sometimes the final signoff response after QUIT could fail to get
3692 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3693 < 0 before doing a fflush(). This bug looks as though it goes back to the
3694 introduction of TLS in release 3.20, but "sometimes" must have been rare
3695 because the tests only now provoked it.
3701 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3702 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3703 Now it returns an integer. A number of calls were still expecting a T/F
3704 return. Fortuitously, in all cases, the tests worked in OK situations,
3705 which is the norm. However, things would have gone wrong on any write
3706 failures on the smtp file descriptor. This function is used when sending
3707 messages over SMTP and also when doing verify callouts.
3709 2. When Exim is called to do synchronous delivery of a locally submitted
3710 message (the -odf or -odi options), it no longer closes stderr before doing
3713 3. Implemented the mua_wrapper option.
3715 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3717 5. Implemented the functions header_remove(), header_testname(),
3718 header_add_at_position(), and receive_remove_recipient(), and exported them
3721 6. If an ACL "warn" statement specified the addition of headers, Exim already
3722 inserted X-ACL-Warn: at the start if there was no header name. However, it
3723 was not making this test for the second and subsequent header lines if
3724 there were newlines in the string. This meant that an invalid header could
3725 be inserted if Exim was badly configured.
3727 7. Allow an ACL "warn" statement to add header lines at the start or after all
3728 the Received: headers, as well as at the end.
3730 8. Added the rcpt_4xx retry error code.
3732 9. Added postmaster_mailfrom=xxx to callout verification option.
3734 10. Added mailfrom=xxxx to the callout verification option, for verify=
3737 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3738 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3740 12. Inserted some casts to stop certain compilers warning when using pointer
3741 differences as field lengths or precisions in printf-type calls (mostly
3742 affecting debugging statements).
3744 13. Added optional readline() support for -be (dynamically loaded).
3746 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3747 same clock tick as a message's arrival, so that its received time was the
3748 same as the "first fail" time on the retry record, and that message
3749 remained on the queue past the ultimate address timeout, every queue runner
3750 would try a delivery (because it was past the ultimate address timeout) but
3751 after another failure, the ultimate address timeout, which should have then
3752 bounced the address, did not kick in. This was a "< instead of <=" error;
3753 in most cases the first failure would have been in the next clock tick
3754 after the received time, and all would be well.
3756 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3757 being recognized when the domain list was tested by the match_domain
3758 condition in an expansion string.
3760 16. Added the ${str2b64: operator.
3762 17. Exim was always calling setrlimit() to set a large limit for the number of
3763 processes, without checking whether the existing limit was already
3764 adequate. (It did check for the limit on file descriptors.) Furthermore,
3765 errors from getrlimit() and setrlimit() were being ignored. Now they are
3766 logged to the main and panic logs, but Exim does carry on, to try to do its
3767 job under whatever limits there are.
3769 18. Imported PCRE 5.0.
3771 19. Trivial typo in log message " temporarily refused connection" (the leading
3774 20. If the log selector return_path_on_delivery was set and an address was
3775 redirected to /dev/null, the delivery process crashed because it assumed
3776 that a return path would always be set for a "successful" delivery. In this
3777 case, the whole delivery is bypassed as an optimization, and therefore no
3780 21. Internal re-arrangement: the function for sending a challenge and reading
3781 a response while authentication was assuming a zero-terminated challenge
3782 string. It's now changed to take a pointer and a length, to allow for
3783 binary data in such strings.
3785 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3787 23. Exim was not respecting finduser_retries when seeking the login of the
3788 uid under which it was called; it was always trying 10 times. (The default
3789 setting of finduser_retries is zero.) Also, it was sleeping after the final
3790 failure, which is pointless.
3792 24. Implemented tls_on_connect_ports.
3794 25. Implemented acl_smtp_predata.
3796 26. If the domain in control=submission is set empty, Exim assumes that the
3797 authenticated id is a complete email address when it generates From: or
3798 Sender: header lines.
3800 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3801 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3802 chown and chgrp in /bin and hostname in /usr/bin.
3804 28. Exim was keeping the "process log" file open after each use, just as it
3805 does for the main log. This opens the possibility of it remaining open for
3806 long periods when the USR1 signal hits a daemon. Occasional processlog
3807 errors were reported, that could have been caused by this. Anyway, it seems
3808 much more sensible not to leave this file open at all, so that is what now
3811 29. The long-running daemon process does not normally write to the log once it
3812 has entered its main loop, and it closes the log before doing so. This is
3813 so that log files can straightforwardly be renamed and moved. However,
3814 there are a couple of unusual error situations where the daemon does write
3815 log entries, and I had neglected to close the log afterwards.
3817 30. The text of an SMTP error response that was received during a remote
3818 delivery was being truncated at 512 bytes. This is too short for some of
3819 the long messages that one sometimes sees. I've increased the limit to
3822 31. It is now possible to make retry rules that apply only when a message has a
3823 specific sender, in particular, an empty sender.
3825 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3826 it possible to be selective about when SMTP synchronization is enforced.
3828 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3830 32. Implemented hosts_connection_nolog.
3832 33. Added an ACL for QUIT.
3834 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3837 35. Added mailbox_size and mailbox_filecount to appendfile.
3839 36. Added control = no_multiline_responses to ACLs.
3841 37. There was a bug in the logic of the code that waits for the clock to tick
3842 in the case where the clock went backwards by a substantial amount such
3843 that the microsecond fraction of "now" was more than the microsecond
3844 fraction of "then" (but the whole seconds number was less).
3846 38. Added support for the libradius Radius client library this is found on
3847 FreeBSD (previously only the radiusclient library was supported).
3853 1. When certain lookups returned multiple values in the form name=value, the
3854 quoting of the values was not always being done properly. Specifically:
3855 (a) If the value started with a double quote, but contained no whitespace,
3857 (b) If the value contained whitespace other than a space character (i.e.
3858 tabs or newlines or carriage returns) it was not quoted.
3859 This fix has been applied to the mysql and pgsql lookups by writing a
3860 separate quoting function and calling it from the lookup code. The fix
3861 should probably also be applied to nisplus, ibase and oracle lookups, but
3862 since I cannot test any of those, I have not disturbed their existing code.
3864 2. A hit in the callout cache for a specific address caused a log line with no
3865 reason for rejecting RCPT. Now it says "Previous (cached) callout
3866 verification failure".
3868 3. There was an off-by-one bug in the queryprogram router. An over-long
3869 return line was truncated at 256 instead of 255 characters, thereby
3870 overflowing its buffer with the terminating zero. As well as fixing this, I
3871 have increased the buffer size to 1024 (and made a note to document this).
3873 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3874 when Exim is waiting for an SMTP response from a remote server, Exim
3875 restarts its select() call on the socket, thereby resetting its timeout.
3876 This is not a problem when such interrupts are rare. Somebody set up a cron
3877 job to run exiwhat every 2 minutes, which is less than the normal select()
3878 timeout (5 or 10 minutes). This meant that the select() timeout never
3879 kicked in because it was always reset. I have fixed this by comparing the
3880 time when an interrupt arrives with the time at the start of the first call
3881 to select(). If more time than the timeout has elapsed, the interrupt is
3882 treated as a timeout.
3884 5. Some internal re-factoring in preparation for the addition of Sieve
3885 extensions (by MH). In particular, the "personal" test is moved to a
3886 separate function, and given an option for scanning Cc: and Bcc: (which is
3887 not set for Exim filters).
3889 6. When Exim created an email address using the login of the caller as the
3890 local part (e.g. when creating a From: or Sender: header line), it was not
3891 quoting the local part when it contained special characters such as @.
3893 7. Installed new OpenBSD configuration files.
3895 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3896 try to make them clearer.
3898 9. Callout options, other than the timeout value, were being ignored when
3899 verifying sender addresses in header lines. For example, when using
3901 verify = header_sender/callout=no_cache
3903 the cache was (incorrectly) being used.
3905 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3906 only the Cygwin environment.
3908 11. When return_path_on_delivery was set as a log selector, if different remote
3909 addresses in the same message used different return paths and parallel
3910 remote delivery occurred, the wrong values would sometimes be logged.
3911 (Whenever a remote delivery process finished, the return path value from
3912 the most recently started remote delivery process was logged.)
3914 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3915 lines when AUTH and/or TLS are in use. This is the "received protocol"
3916 field. Exim used to use "asmtp" for authenticated SMTP, without any
3917 indication (in the protocol name) for TLS use. Now it follows the RFC and
3918 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3919 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3920 names appear in log lines as well as in Received: header lines.
3922 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3923 extensions, and comparison tests, and to fix some bugs.
3925 14. Changes to the "personal" filter test:
3927 (1) The test was buggy in that it was just doing the equivalent of
3928 "contains" tests on header lines. For example, if a user's address was
3929 anne@some.where, the "personal" test would incorrectly be true for
3931 To: susanne@some.where
3933 This test is now done by extracting each address from the header in turn,
3934 and checking the entire address. Other tests that are part of "personal"
3935 are now done using regular expressions (for example, to check local parts
3936 of addresses in From: header lines).
3938 (2) The list of non-personal local parts in From: addresses has been
3939 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3940 taken from the Sieve specification recommendations.
3942 (3) If the message contains any header line starting with "List-" it is
3943 treated as non-personal.
3945 (4) The test for "circular" in the Subject: header line has been removed
3946 because it now seems ill-conceived.
3948 15. Minor typos in src/EDITME comments corrected.
3950 16. Installed latest exipick from John Jetmore.
3952 17. If headers_add on a router specified a text string that was too long for
3953 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3954 of string_sprintf() is now avoided.
3956 18. $message_body_size was not set (it was always zero) when running the DATA
3957 ACL and the local_scan() function.
3959 19. For the "mail" command in an Exim filter, no default was being set for
3960 the once_repeat time, causing a random time value to be used if "once" was
3961 specified. (If the value happened to be <= 0, no repeat happened.) The
3962 default is now 0s, meaning "never repeat". The "vacation" command was OK
3963 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3964 (I found it when inspecting the code).
3966 20. There is now an overall timeout for performing a callout verification. It
3967 defaults to 4 times the callout timeout, which applies to individual SMTP
3968 commands during the callout. The overall timeout applies when there is more
3969 than one host that can be tried. The timeout is checked before trying the
3970 next host. This prevents very long delays if there are a large number of
3971 hosts and all are timing out (e.g. when the network connections are timing
3972 out). The value of the overall timeout can be changed by specifying an
3973 additional sub-option for "callout", called "maxwait". For example:
3975 verify = sender/callout=5s,maxwait=20s
3977 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3978 to the end before writing, but this should make it even safer).
3980 22. Exim was forgetting that it had advertised PIPELINING for the second and
3981 subsequent messages on an SMTP connection. It was also not resetting its
3982 memory on STARTTLS and an internal HELO.
3984 23. When Exim logs an SMTP synchronization error within a session, it now
3985 records whether PIPELINING has been advertised or not.
3987 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3988 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3989 rather than long int.
3991 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3993 26. Added the never_mail option to autoreply.
3999 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4000 crash if the getsockname() call failed; this can happen if a connection is
4001 closed very soon after it is established. The problem was simply in the
4002 order in which certain operations were done, causing Exim to try to write
4003 to the SMTP stream before it had set up the file descriptor. The bug has
4004 been fixed by making things happen in the correct order.
4010 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4011 before the connection was closed, thus losing the rejection response.
4013 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4014 some early Solaris releases, but causes trouble in current releases where
4015 socklen_t is defined.
4017 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4020 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4023 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4024 that releasing the top of it at the end releases what was used for sub-
4025 expansions (unless the block got too big). However, discard this block if
4026 the first thing is a variable or header, so that we can use its block when
4027 it is dynamic (useful for very large $message_headers, for example).
4029 6. Lookups now cache *every* query, not just the most recent. A new, separate
4030 store pool is used for this. It can be recovered when all lookup caches are
4031 flushed. Lookups now release memory at the end of their result strings.
4032 This has involved some general refactoring of the lookup sources.
4034 7. Some code has been added to the store_xxx() functions to reduce the amount
4035 of flapping under certain conditions.
4037 8. log_incoming_interface used to affect only the <= reception log lines. Now
4038 it causes the local interface and port to be added to several more SMTP log
4039 lines, for example "SMTP connection from", and rejection lines.
4041 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4043 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4045 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4046 could be overwritten at the end of the current message (or the start of a
4047 new message if it was set in a HELO ACL). The value is now preserved for
4048 the duration of the SMTP connection.
4050 12. If a transport had a headers_rewrite setting, and a matching header line
4051 contained an unqualified address, that address was qualified, even if it
4052 did not match any rewriting rules. The underlying bug was that the values
4053 of the flags that permit the existence of unqualified sender and recipient
4054 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4055 non-local messages, and by -bnq for local messages) were not being
4056 preserved with the message after it was received.
4058 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4059 "next input=" as part of the text comprising the host identity instead of
4060 the correct text. The code was using the same buffer for two different
4061 strings. However, depending on which order the printing function evaluated
4062 its arguments, the bug did not always show up. Under Linux, for example, my
4063 test suite worked just fine.
4065 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4066 doesn't work with some older versions of Perl. It has been changed to "my",
4067 which in any case is probably the better facility to use.
4069 15. A really picky compiler found some instances of statements for creating
4070 error messages that either had too many or two few arguments for the format
4073 16. The size of the buffer for calls to the DNS resolver has been increased
4074 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4075 for addresses that have a lot of PTR records. This alleviates a problem; it
4076 does not fully solve it.
4078 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4079 buffer now truncates the list and logs the incident, which is the same
4080 action as happens when Exim is looking up a host name and its aliases.
4081 Previously in this situation something unpredictable would happen;
4082 sometimes it was "internal error: store_reset failed".
4084 18. If a server dropped the connection unexpectedly when an Exim client was
4085 using GnuTLS and trying to read a response, the client delivery process
4086 crashed while trying to generate an error log message.
4088 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4089 string, for example:
4091 warn message = H1: something\nH2: something
4093 the text was added as a single header line from Exim's point of view
4094 though it ended up OK in the delivered message. However, searching for the
4095 second and subsequent header lines using $h_h2: did not work. This has been
4096 fixed. Similarly, if a system filter added multiple headers in this way,
4097 the routers could not see them.
4099 20. Expanded the error message when iplsearch is called with an invalid key to
4100 suggest using net-iplsearch in a host list.
4102 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4103 ACLs are no longer actually imposed (and a message to that effect is
4106 22. If a "gecos" field in a passwd entry contained escaped characters, in
4107 particular, if it contained a \" sequence, Exim got it wrong when building
4108 a From: or a Sender: header from that name. A second bug also caused
4109 incorrect handling when an unquoted " was present following a character
4110 that needed quoting.
4112 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4113 was not being matched caselessly.
4115 24. Arranged for all hyphens in the exim.8 source to be escaped with
4118 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4119 properly. Recipient callout cache records were still being keyed to include
4120 the sender, even when use_sender was set false. This led to far more
4121 callouts that were necessary. The sender is no longer included in the key
4122 when use_sender is false.
4124 26. Added "control = submission" modifier to ACLs.
4126 27. Added the ${base62d: operator to decode base 62 numbers.
4128 28. dnsdb lookups can now access SRV records.
4130 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4131 the configuration file.
4133 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4134 (-v) mode. This makes the output for a verbose queue run more intelligible.
4136 31. Added a use_postmaster feature to recipient callouts.
4138 32. Added the $body_zerocount variable, containing the number of binary zero
4139 bytes in the message body.
4141 33. The time of last modification of the "new" subdirectory is now used as the
4142 "mailbox time last read" when there is a quota error for a maildir
4145 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4147 35. Added +ignore_unknown as a special item in host lists.
4149 36. Code for decoding IPv6 addresses in host lists is now included, even if
4150 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4151 address was recognized as an IP address, but was then not correctly decoded
4152 into binary, causing unexpected and incorrect effects when compared with
4159 1. Very minor rewording of debugging text in manualroute to say "list of
4160 hosts" instead of "hostlist".
4162 2. If verify=header_syntax was set, and a header line with an unqualified
4163 address (no domain) and a large number of spaces between the end of the
4164 name and the colon was received, the reception process suffered a buffer
4165 overflow, and (when I tested it) crashed. This was caused by some obsolete
4166 code that should have been removed. The fix is to remove it!
4168 3. When running in the test harness, delay a bit after writing a bounce
4169 message to get a bit more predictability in the log output.
4171 4. Added a call to search_tidyup() just before forking a reception process. In
4172 theory, someone could use a lookup in the expansion of smtp_accept_max_
4173 per_host which, without the tidyup, could leave open a database connection.
4175 5. Added the variables $recipient_data and $sender_data which get set from a
4176 lookup success in an ACL "recipients" or "senders" condition, or a router
4177 "senders" option, similar to $domain_data and $local_part_data.
4179 6. Moved the writing of debug_print from before to after the "senders" test
4182 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4183 problems for message scanning, either using a data ACL, or using
4184 local_scan() because the Received: header was not generated till after they
4185 were called (in order to set the time as the time of reception completion).
4186 I have revised the way this works. The header is now generated after the
4187 body is received, but before the ACL or local_scan() are called. After they
4188 are run, the timestamp in the header is updated.
4194 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4195 before starting a queue runner without re-exec. This happened only when
4196 deliver_drop_privilege was set or when the Exim user was set to root. The
4197 effect of the bug was that timeouts during subsequent deliveries caused
4198 crashes instead of being properly handled. The handler is now left at its
4199 default (and expected) setting.
4201 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4202 message, again when deliver_drop_privilege is set or Exim is run as root.
4203 The bug described in (1) was not present in this case, but the tidying up
4204 of the other signals was missing. I have made the two cases consistent.
4206 3. The ignore_target_hosts setting on a manualroute router was being ignored
4207 for hosts that were looked up using the /MX notation.
4209 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4212 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4213 operated on the sender address. After changing the $sender_address to <>
4214 for the sender address verify, Exim was re-instated it as the original
4215 (before rewriting) address, but remembering that it had rewritten it, so it
4216 wasn't rewriting it again. This bug also had the effect of breaking the
4217 sender address verification caching when the sender address was rewritten.
4219 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4220 This has been changed so that if the ip literal address matches
4221 ignore_target_hosts, the router declines.
4223 7. Added expansion conditions match_domain, match_address, and match_local_
4224 part (NOT match_host).
4226 8. The placeholder for the Received: header didn't have a length field set.
4228 9. Added code to Exim itself and to exim_lock to test for a specific race
4229 condition that could lead to file corruption when using MBX delivery. The
4230 issue is with the lockfile that is created in /tmp. If this file is removed
4231 after a process has opened it but before that process has acquired a lock,
4232 there is the potential for a second process to recreate the file and also
4233 acquire a lock. This could lead to two Exim processes writing to the file
4234 at the same time. The added code performs the same test as UW imapd; it
4235 checks after acquiring the lock that its file descriptor still refers to
4236 the same named file.
4238 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4239 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4242 11. Added the smtp_active_hostname option. If used, this will typically be made
4243 to depend on the incoming interface address. Because $interface_address is
4244 not set up until the daemon has forked a reception process, error responses
4245 that can happen earlier (such as "too many connections") no longer contain
4248 12. If an expansion in a condition on a "warn" statement fails because a lookup
4249 defers, the "warn" statement is abandoned, and the next ACL statement is
4250 processed. Previously this caused the whole ACL to be aborted.
4252 13. Added the iplsearch lookup type.
4254 14. Added ident_timeout as a log selector.
4256 15. Added tls_certificate_verified as a log selector.
4258 16. Added a global option tls_require_ciphers (compare the smtp transport
4259 option of the same name). This controls incoming TLS connections.
4261 17. I finally figured out how to make tls_require_ciphers do a similar thing
4262 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4263 before starting the TLS session.
4265 18. Tabs are now shown as \t in -bP output.
4267 19. If the log selector return_path_on_delivery was set, Exim crashed when
4268 bouncing a message because it had too many Received: header lines.
4270 20. If two routers both had headers_remove settings, and the first one included
4271 a superfluous trailing colon, the final name in the first list and the
4272 first name in the second list were incorrectly joined into one item (with a
4273 colon in the middle).
4279 1. Added -C and -D options to the exinext utility, mainly to make it easier
4280 to include in the automated testing, but these could be helpful when
4281 multiple configurations are in use.
4283 2. The exinext utility was not formatting the output nicely when there was
4284 an alternate port involved in the retry record key, nor when there was a
4285 message id as well (for retries that were specific to a specific message
4286 and a specific host). It was also confused by IPv6 addresses, because of
4287 the additional colons they contain. I have fixed the IPv4 problem, and
4288 patched it up to do a reasonable job for IPv6.
4290 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4291 delivery, the log line now contains "pipelined" if PIPELINING was used.
4293 4. An SMTP transport process used to panic and die if the bind() call to set
4294 an explicit outgoing interface failed. This has been changed; it is now
4295 treated in the same way as a connect() failure.
4297 5. A reference to $sender_host_name in the part of a conditional expansion
4298 that was being skipped was still causing a DNS lookup. This no longer
4301 6. The def: expansion condition was not recognizing references to header lines
4302 that used bh_ and bheader_.
4304 7. Added the _cache feature to named lists.
4306 8. The code for checking quota_filecount in the appendfile transport was
4307 allowing one more file than it should have been.
4309 9. For compatibility with Sendmail, the command line option
4317 and sets the incoming protocol and host name (for trusted callers). The
4318 host name and its colon can be omitted when only the protocol is to be set.
4319 Note the Exim already has two private options, -pd and -ps, that refer to
4320 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4321 "s", but I don't think that's a major issue.
4323 10. A number of refactoring changes to the code, none of which should affect
4326 (a) The number of logging options was getting close to filling up the
4327 32-bit word that was used as a bit map. I have split them into two classes:
4328 those that are passed in the argument to log_write(), and those that are
4329 only ever tested independently outside of that function. These are now in
4330 separate 32-bit words, so there is plenty of room for expansion again.
4331 There is no change in the user interface or the logging behaviour.
4333 (b) When building, for example, log lines, the code previously used a
4334 macro that called string_cat() twice, in order to add two strings. This is
4335 not really sufficiently general. Furthermore, there was one instance where
4336 it was actually wrong because one of the argument was used twice, and in
4337 one call a function was used. (As it happened, calling the function twice
4338 did not affect the overall behaviour.) The macro has been replaced by a
4339 function that can join an arbitrary number of extra strings onto a growing
4342 (c) The code for expansion conditions now uses a table and a binary chop
4343 instead of a serial search (which was left over from when there were very
4344 few conditions). Also, it now recognizes conditions like "pam" even when
4345 the relevant support is not compiled in: a suitably worded error message is
4346 given if an attempt is made to use such a condition.
4348 11. Added ${time_interval:xxxxx}.
4350 12. A bug was causing one of the ddress fields not to be passed back correctly
4351 from remote delivery subprocesses. The field in question was not being
4352 subsequently used, so this caused to problems in practice.
4354 13. Added new log selectors queue_time and deliver_time.
4356 14. Might have fixed a bug in maildirsizefile handling that threw up
4357 "unexpected character" debug warnings, and recalculated the data
4358 unnecessarily. In any case, I expanded the warning message to give more
4361 15. Added the message "Restricted characters in address" to the statements in
4362 the default ACL that block characters like @ and % in local parts.
4364 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4365 Three changes have been made:
4367 (a) There was a serious bug; a negative response to MAIL caused the whole
4368 recipient domain to be cached as invalid, thereby blocking all messages
4369 to all local parts at the same domain, from all senders. This bug has
4370 been fixed. The domain is no longer cached after a negative response to
4371 MAIL if the sender used is not empty.
4373 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4376 (c) A new callout option, "use_sender" has been added for people who want
4377 the modified behaviour.
4383 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4386 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4387 indeed breaks things for older releases.
4389 3. Added additional logging to the case where there is a problem reading data
4390 from a filter that is running in a subprocess using a pipe, in order to
4391 try to track down a specific problem.
4393 4. Testing facility fudge: when running in the test harness and attempting
4394 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4395 getting "No route to host". Convert this to a timeout.
4397 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4400 6. Some OS don't have socklen_t but use size_t instead. This affects the
4401 fifth argument of getsockopt() amongst other things. This is now
4402 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4403 can be set for individual OS. I have set it for SunOS5, OSF1, and
4404 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4405 some earlier ones do not.
4407 7. Change 4.30/15 was not doing the test caselessly.
4409 8. The standard form for an IPv6 address literal was being rejected by address
4410 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4411 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4412 this, as well as the form without the "IPv6" on the front (but only when
4413 address literals are enabled, of course).
4415 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4417 10. Exim crashed if a message with an empty sender address specified by -f
4418 encountered a router with an errors_to setting. This could be provoked only
4419 by a command such as
4423 where an empty string was supplied; "<>" did not hit this bug.
4425 11. Installed PCRE release 4.5.
4427 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4428 remained set. It is now erased.
4430 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4431 times from message ids (which are base 36 rather than the normal 62).
4433 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4434 were being counted as actual protocol errors, and logged if the log
4435 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4436 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4437 and DATA following a set of rejected RCPTs do not count as protocol errors.
4438 In other words, Exim assumes they were pipelined, though this may not
4439 actually be the case. Of course, in all cases the client gets an
4440 appropriate error code.
4442 15. If a lookup fails in an ACL condition, a message about the failure may
4443 be available; it is used if testing the ACL cannot continue, because most
4444 such messages specify what the cause of the deferral is. However, some
4445 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4446 that caused an old message to be retained and used if a later statement
4447 caused a defer, replacing the real cause of the deferral.
4449 16. If an IP address had so many PTR records that the DNS lookup buffer
4450 was not large enough to hold them, Exim could crash while trying to process
4451 the truncated data. It now detects and logs this case.
4453 17. Further to 4.21/58, another change has been made: if (and only if) the
4454 first line of a message (the first header line) ends with CRLF, a bare LF
4455 in a subsequent header line has a space inserted after it, so as not to
4456 terminate the header.
4458 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4459 unnecessarily, used atoi() instead of strtol(), and didn't check the
4460 termination when getting file sizes from file names by regex.
4462 19. Completely re-implemented the support for maildirsize files, in the light
4463 of a number of problems with the previous contributed implementation
4464 (4.30/29). In particular:
4466 . If the quota is zero, the maildirsize file is maintained, but no quota is
4469 . If the maildir directory does not exist, it is created before any attempt
4470 to write a maildirsize file.
4472 . The quota value in the file is just a cache; if the quota is changed in
4473 the transport, the new value overrides.
4475 . A regular expression is available for excluding directories from the
4478 20. The autoreply transport checks the characters in options that define the
4479 message's headers; it allows continued headers, but it was checking with
4480 isspace() after an embedded newline instead of explicitly looking for a
4483 21. If all the "regular" hosts to which an address was routed had passed their
4484 expiry times, and had not reached their retry times, the address was
4485 bounced, even if fallback hosts were defined. Now Exim should go on to try
4488 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4489 equivalent code in the SMTP transport. Some hosts send humungous responses
4490 to HELO/EHLO, more than 1024 it seems.
4492 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4493 gives compiler warnings in some environments. I've now done it "properly",
4496 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4497 (because of problems with the built-in one) was declared to return uschar *
4498 instead of char *, causing compiler failure.
4500 25. Fixed a file descriptor leak when processing alias/forward files.
4502 26. Fixed a minor format string issue in dbfn.c.
4504 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4506 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4507 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4510 29. When a sender address is verified, it is cached, to save repeating the test
4511 when there is more than one recipient in a message. However, when the
4512 verification involves a callout, it is possible for different callout
4513 options to be set for different recipients. It is too complicated to keep
4514 track of this in the cache, so now Exim always runs a verification when a
4515 callout is required, relying on the callout cache for the optimization.
4516 The overhead is duplication of the address routing, but this should not be
4519 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4520 to be verified with callout=postmaster, and the main callout worked but the
4521 postmaster check failed, the verification correctly failed. However, if a
4522 subsequent RCPT command asked for sender verification *without* the
4523 postmaster check, incorrect caching caused this verification also to fail,
4526 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4527 it was not caching the DNS options (qualify_single, search_parents) that
4528 were used when the lookup failed. A subsequent lookup with different
4529 options therefore always gave the same answer, though there were cases
4530 where it should not have. (Example: a "domains = !$mx_any" option on a
4531 dnslookup router: the "domains" option is always processed without any
4532 widening, but the router might have qualify_single set.) Now Exim uses the
4533 cached value only when the same options are set.
4535 32. Added John Jetmore's "exipick" utility to the distribution.
4537 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4538 than a timeout (e.g. a certificate is required, and is not provided), an
4539 Exim server now closes the connection immediately. Previously it waited for
4540 the client to close - but if the client is SSL, it seems that they each
4541 wait for each other, leading to a delay before one of them times out.
4543 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4544 maintained 0.8.x compatibility because I don't think many are using it, and
4545 it is clearly obsolete.
4547 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4550 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4551 client certificate was expired. A simple patch fixes this, though I don't
4552 understand the full logic of why the verify callback is called multiple
4555 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4556 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4557 which causes problems with some clients (such as the Certicom SSL Plus
4558 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4559 disables the coutermeasure allowing Eudora to connect."
4561 38. Exim was not checking that a write() to a log file succeeded. This could
4562 lead to Bad Things if a log got too big, in particular if it hit a file
4563 size limit. Exim now panics and dies if it cannot write to a log file, just
4564 as it does if it cannot open a log file.
4566 39. Modified OS/Makefile-Linux so that it now contains
4568 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4570 The two -D definitions ensure that Exim is compiled with large file
4571 support, which makes it possible to handle log files that are bigger than
4574 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4575 instance) a domain was checked against a named list that involved a lookup,
4576 causing $domain_data to be set, then another domain was checked against the
4577 same list, then the first domain was re-checked, the value of $domain_data
4578 after the final check could be wrong. In particular, if the second check
4579 failed, it could be set empty. This bug probably also applied to
4582 41. The strip_trailing_dot option was not being applied to the address given
4583 with the -f command-line option.
4585 42. The code for reading a message's header from the spool was incrementing
4586 $received_count, but never initializing it. This meant that the value was
4587 incorrect (doubled) while delivering a message in the same process in which
4588 it was received. In the most common configuration of Exim, this never
4589 happens - a fresh exec is done - but it can happen when
4590 deliver_drop_privilege is set.
4592 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4593 it now includes up to 150 characters of the unexpected data in the log
4596 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4597 and building data strings. The size of both of these buffers was 10 000
4598 bytes - far larger than anybody would *ever* want, thought I. Needless to
4599 say, somebody hit the limit. I have increased the maximum line length to
4600 20 000 and the maximum data length of concatenated lines to 100 000. I have
4601 also fixed two bugs, because there was no checking on these buffers. Tsk,
4602 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4603 buffer is too small.
4605 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4606 lsearch lookups. Now it does.
4608 46. When parsing a route_list item in a manualroute router, a fixed-length
4609 buffer was used for the list of hosts. I made this 1024 bytes long,
4610 thinking that nobody would ever have a list of hosts that long. Wrong.
4611 Somebody had a whole pile of complicated expansion conditions, and the
4612 string was silently truncated, leading to an expansion error. It turns out
4613 that it is easier to change to an unlimited length (owing to other changes
4614 that have happened since this code was originally written) than to build
4615 structure for giving a limitation error. The length of the item that
4616 expands into the list of hosts is now unlimited.
4618 47. The lsearch lookup could not handle data where the length of text line was
4619 more than 4095 characters. Such lines were truncated, leading to shortened
4620 data being returned. It should now handle lines of any length.
4622 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4623 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4626 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4627 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4628 no longer happens in comment lines. A list of replacements is now placed
4629 at the head of all of the source files, except those whose only change is
4630 to replace PERL_COMMAND in the very first #! line.
4632 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4633 messages on the queue, with a bottom-up merge sort, using code contributed
4634 by Michael Haardt. This should make operations like -bp somewhat faster on
4635 large queues. It won't affect queue runners, except when queue_run_in_order
4638 51. Installed eximstats 1.31 in the distribution.
4640 52. Added support for SRV lookups to the dnslookup router.
4642 53. If an ACL referred to $message_body or $message_body_end, the value was not
4643 reset for any messages that followed in the same SMTP session.
4645 54. The store-handling optimization for building very long strings was not
4646 differentiating between the different store pools. I don't think this
4647 actually made any difference in practice, but I've tidied it.
4649 55. While running the routers to verify a sender address, $sender_address
4650 was still set to the sender address. This is wrong, because when routing to
4651 send a bounce to the sender, it would be empty. Therefore, I have changed
4652 it so that, while verifying a sender address, $sender_address is set to <>.
4653 (There is no change to what happens when verifying a recipient address.)
4655 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4656 target A or AAAA records (if not already returned) without resetting the
4657 qualify_single or search_parents options of the DNS resolver. These are
4658 inappropriate in this case because the targets of MX and SRV records must
4659 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4660 target that, when qualified, matched something in the local domain. These
4661 two options are now turned off when doing these lookups.
4663 57. It seems that at least some releases of Reiserfs (which does not have the
4664 concept of a fixed number of inodes) returns zero and not -1 for the
4665 number of available inodes. This interacted badly with check_spool_inodes,
4666 which assumed that -1 was the "no such thing" setting. What I have done is
4667 to check that the total number of inodes is greater than zero before doing
4668 the test of how many are available.
4670 58. When a "warn" ACL statement has a log_message modifier, the message is
4671 remembered, and not repeated. This is to avoid a lot of repetition when a
4672 message has many recipients that cause the same warning to be written.
4673 Howewer, Exim was preserving the list of already written lines for an
4674 entire SMTP session, which doesn't seem right. The memory is now reset if a
4675 new message is started.
4677 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4678 debugging output unless log_rewrite was also set.
4680 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4681 of (int)(handle) when we know that handle contains (void *)(-1).
4683 61. The Exim daemon panic-logs an error return when it closes the incoming
4684 connection. However "connection reset by peer" seems to be common, and
4685 isn't really an error worthy of noting specially, so that particular error
4688 62. When Exim is trying to find all the local interfaces, it used to panic and
4689 die if the ioctl to get the interface flags failed. However, it seems that
4690 on at least one OS (Solaris 9) it is possible to have an interface that is
4691 included in the list of interfaces, but for which you get a failure error
4692 for this call. This happens when the interface is not "plumbed" into a
4693 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4694 failure of the "get flags" call assumes that the interface is down.
4696 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4697 makes life easier for people who are doing arithmetic on fields extracted
4698 from dates, where you often get leading zeros that should not be
4699 interpreted as octal.
4701 64. Added qualify_domain to the redirect router, to override the global
4704 65. If a pathologically long header line contained very many addresses (the
4705 report of this problem mentioned 10 000) and each of them was rewritten,
4706 Exim could use up a very large amount of memory. (It kept on making new
4707 copies of the header line as it rewrote, and never released the old ones.)
4708 At the expense of a bit more processing, the header rewriting function has
4709 been changed so that it no longer eats memory in this way.
4711 66. The generation of the Received: header has been moved from the time that a
4712 message starts to be received, to the time that it finishes. The timestamp
4713 in the Received: header should now be very close to that of the <= log
4714 line. There are two side-effects of this change:
4716 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4717 logged header lines no longer include the local Received: line, because
4718 it has not yet been created. The same applies to a copy of the message
4719 that is returned to a non-SMTP sender when a message is rejected.
4721 (b) When a filter file is tested using -bf, no additional Received: header
4722 is added to the test message. After some thought, I decided that this
4725 This change does not affect the value of $received_for. It is still set
4726 after address rewriting, but before local_scan() is called.
4728 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4730 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4731 gave an unhelpful panic error message, and a defer error. I have managed to
4732 change this behaviour so that it now rejects any supplied certificate,
4733 which seems right, as the list of acceptable certificates is empty.
4735 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4736 gave an unhelpful defer error. I have not managed to make this reject any
4737 supplied certificates, but the error message it gives is "no certificate
4738 supplied", which is not helpful.
4740 70. exigrep's output now also includes lines that are not associated with any
4741 message, but which match the given pattern. Implemented by a patch from
4742 Martin Sluka, which also tidied up the Perl a bit.
4744 71. Recipient callout verification, like sender verification, was using <> in
4745 the MAIL FROM command. This isn't really the right thing, since the actual
4746 sender may affect whether the remote host accepts the recipient or not. I
4747 have changed it to use the actual sender in the callout; this means that
4748 the cache record is now keyed on a recipient/sender pair, not just the
4749 recipient address. There doesn't seem to be a real danger of callout loops,
4750 since a callout by the remote host to check the sender would use <>.
4751 [SEE ABOVE: changed after hitting problems.]
4753 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4754 temporary errors. However, in the case of such a code being given after
4755 the end of a data transmission (i.e. after ".") Exim was failing to write
4756 a retry record for the message. (Yes, there was some broken host that was
4757 actually sending 8xx at this point.)
4759 73. An unknown lookup type in a host list could cause Exim to panic-die when
4760 the list was checked. (An example that provoked this was putting <; in the
4761 middle of a list instead of at the start.) If this happened during a DATA
4762 ACL check, a -D file could be left lying around. This kind of configuration
4763 error no longer causes Exim to die; instead it causes a defer errror. The
4764 incident is still logged to the main and panic logs.
4766 74. Buglet left over from Exim 3 conversion. The message "too many messages
4767 in one connection" was written to the rejectlog but not the mainlog, except
4768 when address rewriting (yes!) was being logged.
4770 75. Added write_rejectlog option.
4772 76. When a system filter was run not as root (that is, when system_filter_user
4773 was set), the values of the $n variables were not being returned to the
4774 main process; thus, they were not subsequently available in the $sn
4777 77. Added +return_path_on_delivery log selector.
4779 78. A connection timeout was being treated differently from recipients deferred
4780 when testing hosts_max_try with a message that was older than the host's
4781 retry timeout. (The host should not be counted, thus allowing all hosts to
4782 be tried at least once before bouncing.) This may have been the cause of an
4783 occasionally reported bug whereby a message would remain on the queue
4784 longer than the retry timeout, but would be bounced if a delivery was
4785 forced. I say "may" because I never totally pinned down the problem;
4786 setting up timeout/retry tests is difficult. See also the next item.
4788 79. The ultimate address timeout was not being applied to errors that involved
4789 a combination of host plus message (for example, a timeout on a MAIL
4790 command). When an address resolved to a number of possible hosts, and they
4791 were not all tried for each delivery (e.g. because of hosts_max_try), a
4792 message could remain on the queue longer than the retry timeout.
4794 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4797 81. Fixed an obscure SMTP outgoing bug which required at least the following
4798 conditions: (a) there was another message waiting for the same server;
4799 (b) the server returned 5xx to all RCPT commands in the first message so
4800 that the message was not completed; (c) the server dropped the connection
4801 or gave a negative response to the RSET that Exim sends to abort the
4802 transaction. The observed case was a dropped connection after DATA that had
4803 been sent in pipelining mode. That is, the server had advertised PIPELINING
4804 but was not implementing it correctly. The effect of the bug was incorrect
4805 behaviour, such as trying another host, and this could lead to a crash.
4811 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4812 and daemon.c were passed as pointers to ints; they should have been
4813 pointers to socklen_t variables (which are typically unsigned ints).
4815 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4818 3. Fixed a really odd bug that affected only the testing scheme; patching a
4819 certain fixed string in the binary changed the value of another string that
4820 happened to be identical to the end of the original first string.
4822 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4823 name", it returns that address as the IP address. On some operating
4824 systems (e.g. Solaris), it also passes back the IP address string as the
4825 "host name". However, on others (e.g. Linux), it passes back an empty
4826 string. Exim wasn't checking for this, and was changing the host name to an
4827 empty string, assuming it had been canonicized.
4829 5. Although rare, it is permitted to have more than one PTR record for a given
4830 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4831 all the names associated with an address, because they do in Solaris.
4832 However, it seems that they do not in Linux for data that comes from the
4833 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4834 I found this out when I moved to a new Linux workstation and tried to run
4835 the Exim test suite.
4837 To get round this problem I have changed the code so that it now does its
4838 own call to the DNS to look up PTR records when searching for a host name.
4839 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4840 addresses that are only in /etc/hosts are still found.
4842 This behaviour is, however, controlled by an option called host_lookup_
4843 order, which defaults to "bydns:byaddr". If people want to use the other
4844 order, or indeed, just use one or the other means of lookup, they can
4845 specify it in this variable.
4847 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4848 some operating systems, this comes back from gethostbyaddr() as an empty
4849 string, and this is what Exim used to test for. However, it seems that in
4850 other systems, "." is yielded. Exim now tests for this case too.
4852 7. The values of check_spool_space and check_log_space are now held internally
4853 as a number of kilobytes instead of an absolute number of bytes. If a
4854 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4855 kilobyte. This means that much larger values can be stored.
4857 8. Exim monitor: an attempt to get the action menu when not actually pointing
4858 at a message produces an empty menu entitled "No message selected". This
4859 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4860 no entries in it ("Shell widget menu has zero width and/or height"). So I
4861 have added a single, blank menu entry in this case.
4863 9. Added ${quote_local_part.
4865 10. MIME decoding is now applied to the contents of Subject: header lines when
4868 11. Now that a reference to $sender_host_address automatically causes a reverse
4869 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4870 host lookup before query-style lookups in lists that might use this
4871 variable. This has therefore been abolished, and the "net-" prefix is no
4872 longer necessary for query-style lookups.
4874 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4875 have been a typo for LFLAGS, so it has been changed.
4877 13. The install script calls Exim with "-C /dev/null" in order to find the
4878 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4879 to be output. Howeve, since Exim outputs its version number before the
4880 error, it didn't break the script. It just looked ugly. I fixed this by
4881 always allowing "-C /dev/null" if the caller is root.
4883 14. Ignore overlarge ACL variable number when reading spool file - insurance
4884 against a later release with more variables having written the file.
4886 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4887 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4888 this, as well as the form without the "IPv6" on the front.
4890 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4891 OS/Makefile-Darwin file.
4893 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4894 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4896 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4897 variable. However, some OS use ints for inodes. I've added cast to long int
4898 to get rid of the compiler warning.
4900 19. I had forgotten to lock out "/../" in configuration file names when
4901 ALT_CONFIG_PREFIX was set.
4903 20. Routers used for verification do not need to specify transports. However,
4904 if such a router generated a host list, and callout was configured, Exim
4905 crashed, because it could not find a port number from the (non-existent)
4906 transport. It now assumes port 25 in this circumstance.
4908 21. Added the -t option to exigrep.
4910 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4911 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4912 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4913 them is set, LOOKUP_LSEARCH is forced.
4915 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4917 24. Added sender and host information to the "rejected by local_scan()" log
4918 line; previously there was no indication of these.
4920 25. Added .include_if_exists.
4922 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4923 Linux. It turns out that not all file systems support this. Apparently some
4924 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4925 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4926 not supported on the file descriptor, is now ignored when Exim is trying to
4927 sync a directory. This applies only to Linux.
4929 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4931 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4932 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4935 29. Added support for maildirsize files from supplied patch (modified a bit).
4937 30. The use of :fail: followed by an empty string could lead Exim to respond to
4938 sender verification failures with (e.g.):
4940 550 Verification failed for <xxx>
4941 550 Sender verify failed
4943 where the first response line was missing the '-' that indicates it is not
4944 the final line of the response.
4946 31. The loop for finding the name of the user that called Exim had a hardwired
4947 limit of 10; it now uses the value of finduser_retries, which is used for
4948 all other user lookups.
4950 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4953 33. Exim was neglecting to zero errno before one call of strtol() when
4954 expanding a string and expecting an integer value. On some systems this
4955 resulted in spurious "integer overflow" errors. Also, it was casting the
4956 result into an int without checking.
4958 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4959 did not work. The code looks as if it has *never* worked, though it appears
4960 to have been documented since at least releast 1.62. I have made it work.
4962 35. The "timeout_DNS" error in retry rules, also documented since at least
4963 1.62, also never worked. As it isn't clear exactly what this means, and
4964 clearly it isn't a major issue, I have abolished the feature by treating it
4965 as "timeout", and writing a warning to the main and panic logs.
4967 36. The display of retry rules for -brt wasn't always showing the error code
4970 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4971 timeout_connect_A, timeout_connect_MX.
4973 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4974 to the empty sender.
4976 39. The daemon was not analysing the content of -oX till after it had closed
4977 stderr and disconnected from the controlling terminal. This meant that any
4978 syntax errors were only noted on the panic log, and the return code from
4979 the command was 0. By re-arranging the code a little, I've made the
4980 decoding happen first, so such errors now appear on stderr, and the return
4981 code is 1. However, the actual setting up of the sockets still happens in
4982 the disconnected process, so errors there are still only recorded on the
4985 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4986 connections (as happens on some IP stacks) was logged at start up time as
4987 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4988 it from "IPv6 and IPv4", which means that two separate sockets are being
4991 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4992 says whether AF_INET or AF_INET6 was passed as an argument.
4994 42. Exiwhat output was messed up when time zones were included in log
4997 43. Exiwhat now gives more information about the daemon's listening ports,
4998 and whether -tls-on-connect was used.
5000 44. The "port" option of the smtp transport is now expanded.
5002 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5003 silently ignored. Now an error message is written to the main and panic
5006 46. There's a new ACL modifier called "logwrite" which writes to a log file
5007 as soon as it is encountered.
5009 47. Added $local_user_uid and $local_user_gid at routing time.
5011 48. Exim crashed when trying to verify a sender address that was being
5014 49. Exim was recognizing only a space character after ".include". It now also
5015 recognizes a tab character.
5017 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5018 extracting the relevant information from the specification. The man page no
5019 longer contains scrambled data for the -d option, and I've added a section
5020 at the front about calling Exim under different names.
5022 51. Added "extra_headers" argument to the "mail" command in filter files.
5024 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5027 53. Installed eximstats 1.29.
5029 54. Added transport_filter_timeout as a generic transport option.
5031 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5032 Cc: header lines. This was required by RFC 822, but it not required by RFC
5035 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5036 incoming messages that did not have them. Now it does so only if the
5037 message originates locally, that is, if there is no associated remote host
5038 address. When Resent- header lines are present, this applies to the Resent-
5039 lines rather than the non-Resent- lines.
5041 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5042 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5044 58. Messages for configuration errors now include the name of the main
5045 configuration file - useful now that there may be more than one file in a
5046 list (.included file names were always shown).
5048 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5049 for those rare installations that do not start the daemon as root or run it
5050 setuid root. I've cut out the call to initgroups() if the daemon is not
5053 60. The Exim user and group can now be bound into the binary as text strings
5054 that are looked up at the start of Exim's processing.
5056 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5058 62. Added $mailstore_basename variable.
5060 63. Installed patch to sieve.c from Michael Haardt.
5062 64. When Exim failed to open the panic log after failing to open the main log,
5063 the original message it was trying to log was written to stderr and debug
5064 output, but if they were not available (the usual case in production), it
5065 was lost. Now it is written to syslog before the two lines that record the
5066 failures to open the logs.
5068 65. Users' Exim filters run in subprocesses under the user's uid. It is
5069 possible for a "deliver" command or an alias in a "personal" command to
5070 provoke an address rewrite. If logging of address rewriting is configured,
5071 this fails because the process is not running as root or exim. There may be
5072 a better way of dealing with this, but for the moment (because 4.30 needs
5073 to be released), I have disabled address rewrite logging when running a
5074 filter in a non-root, non-exim process.
5080 1. The buildconfig auxiliary program wasn't quoting the value set for
5081 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5082 not defined. This bug was masked in 4.22 by the effect that was fixed in
5085 2. Some messages that were rejected after a message id was allocated were
5086 shown as "incomplete" by exigrep. It no longer does this for messages that
5087 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5089 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5090 have allow_domain_literals set, the ID did not get logged in the <= line.
5091 Domain literals are now always recognized in Message-ID: header lines.
5093 4. The first argument for a ${extract expansion item is the key name or field
5094 number. Leading and trailing spaces in this item were not being ignored,
5095 causing some misleading effects.
5097 5. When deliver_drop_privilege was set, single queue runner processes started
5098 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5099 same command in the process it spins off) were not dropping privilege.
5101 6. When the daemon running as "exim" started a queue runner, it always
5102 re-executed Exim in the spun-off process. This is a waste of effort when
5103 deliver_drop_privilege is set. The new process now just calls the
5104 queue-runner function directly.
5110 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5113 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5114 ignored. Though the use of -oP was forcing the writing of a pid file, it
5115 was always written to the default place.
5117 3. If the message "no IP address found for host xxxx" is generated during
5118 incoming verification, it is now followed by identification of the incoming
5119 connection (so you can more easily find what provoked it).
5121 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5123 5. Added some features to "harden" Exim a bit more against certain attacks:
5125 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5126 be put in Local/Makefile. This is like the never_users runtime option,
5127 but it cannot be overridden. The default setting is "root".
5129 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5130 prefix string with which any file named in a -C command line option
5133 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5134 is retained for -C and -D only if the caller of Exim is root. Without
5135 it, the exim user may also use -C and -D and retain privilege.
5137 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5138 command line option is disabled.
5140 6. Macro names set by the -D option must start with an upper case letter, just
5141 like macro names defined in the configuration file.
5143 7. Added "dereference=" facility to LDAP.
5145 8. Two instances of the typo "uknown" in the source files are fixed.
5147 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5148 the Configure-Makefile script screwed up while processing it.
5150 10. Incorporated PCRE 4.4.
5152 11. The SMTP synchronization check was not operating right at the start of an
5153 SMTP session. For example, it could not catch a HELO sent before the client
5154 waited for the greeting. There is now a check for outstanding input at the
5155 point when the greeting is written. Because of the duplex, asynchronous
5156 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5157 way, but not yet received, when the check is performed.
5159 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5160 on TCP/IP sockets, because this apparently causes some broken clients to
5163 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5164 unchanged) from the Cygwin maintainer.
5166 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5167 was supported instead of testing for mbx. Effectively a typo.
5169 15. The spa authenticator server code was not checking that the input it
5170 received was valid base64.
5172 16. The debug output line for the "set" modifier in ACLs was not showing the
5173 name of the variable that was being set.
5175 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5177 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5178 lookup on its own. Something else was needed to trigger the lookup. For
5179 example, a match in host_lookup or the need for a host name in a host list.
5180 Now, if $sender_host_name is referenced and the host name has not yet been
5181 looked up, a lookup is performed. If the lookup fails, the variable remains
5182 empty, and $host_lookup_failed is set to "1".
5184 19. Added "eqi" as a case-independent comparison operator.
5186 20. The saslauthd authentication condition could segfault if neither service
5187 nor realm was specified.
5189 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5190 error message that was logged was misleading, and incoming SMTP
5191 connections were dropped. The message is now more accurate, and temporary
5192 errors are given to SMTP connections.
5194 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5195 (except RSET) with a 421 error, until QUIT is received. However, it was
5196 failing to send a response to QUIT.
5198 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5199 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5200 ACL is now run after the helo_try_verify_hosts code.
5202 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5203 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5204 case-independent, so other case variants are also recognized). Apparently
5205 some systems use these upper case variants.
5207 25. If more than two messages were waiting for the same host, and a transport
5208 filter was specified for the transport, Exim sent two messages over the
5209 same TCP/IP connection, and then failed with "socket operation on non-
5210 socket" when it tried to send the third.
5212 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5214 27. The extern definition of crypt16() in expand.c was not being excluded when
5215 the OS had its own crypt16() function.
5217 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5218 as a preferred synonym for return_size_limit, both as an option and as an
5221 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5223 30. Changed the default configuration ACL to relax the local part checking rule
5224 for addresses that are not in any local domains. For these addresses,
5225 slashes and pipe symbols are allowed within local parts, but the sequence
5226 /../ is explicitly forbidden.
5228 31. SPA server authentication was not clearing the challenge buffer before
5231 32. log_message in a "warn" ACL statement was writing to the reject log as
5232 well as to the main log, which contradicts the documentation and doesn't
5233 seem right (because no rejection is happening). So I have stopped it.
5235 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5236 However, I am unable to do any testing of this.
5238 34. Fixed an infelicity in the appendfile transport. When checking directories
5239 for a mailbox, to see if any needed to be created, it was accidentally
5240 using path names with one or more superfluous leading slashes; tracing
5241 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5243 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5244 contents are added to the log line that is written for every discarded
5245 recipient. (Previously a log_message setting was ignored.)
5247 36. The ${quote: operator now quotes the string if it is empty.
5249 37. The install script runs exim in order to find its version number. If for
5250 some reason other than non-existence or emptiness, which it checks, it
5251 could not run './exim', it was installing it with an empty version number,
5252 i.e. as "exim-". This error state is now caught, and the installation is
5255 38. An argument was missing from the function that creates an error message
5256 when Exim fails to connect to the socket for saslauthd authentication.
5257 This could cause Exim to crash, or give a corrupted message.
5259 39. Added isip, isip4, and isip6 to ${if conditions.
5261 40. The ACL variables $acl_xx are now saved with the message, and can be
5262 accessed later in routers, transports, and filters.
5264 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5265 strings in the file are not string-expanded.
5267 42. If a MAIL command specified a SIZE value that was too large to fit into an
5268 int variable, the check against message_size_limit failed. Such values are
5269 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5270 day this will have to be increased, but I don't think I want to be around
5271 when emails are that large.
5278 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5279 iconv() is not standard in FreeBSD.
5281 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5282 IPv6 enabled. The observed symptom was a segmentation fault on return from
5283 the function os_common_find_running_interfaces() in src/os.c.
5285 3. In the check_special_case() function in daemon.c I had used "errno" as an
5286 argument name, which causes warnings on some systems. This was basically a
5287 typo, since it was named "eno" in the comments!
5289 4. The code that waits for the clock to tick (at a resolution of some fraction
5290 of a second) so as to ensure message-id uniqueness was always waiting for
5291 at least one whole tick, when it could have waited for less. [This is
5292 almost certainly not relevant at current processor speeds, where it is
5293 unlikely to ever wait at all. But we try to future-proof.]
5295 5. The function that sleeps for a time interval that includes fractions of a
5296 second contained a race. It did not block SIGALRM between setting the
5297 timer, and suspending (a couple of lines later). If the interval was short
5298 and the sigsuspend() was delayed until after it had expired, the suspension
5299 never ended. On busy systems this could lead to processes getting stuck for
5302 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5303 process, before it forks any delivery processes. The open lookup caching
5304 mechanism meant that the open file or database connection was passed into
5305 the delivery process. The problem was that delivery processes always tidy
5306 up cached lookup data. This could cause a problem for the next delivery
5307 process started by the queue runner, because the external queue runner
5308 process does not know about the closure. So the next delivery process
5309 still has data in the lookup cache. In the case of a file lookup, there was
5310 no problem because closing a file descriptor in a subprocess doesn't affect
5311 the parent. However, if the lookup was caching a connection to a database,
5312 the connection was closed, and the second delivery process was likely to
5313 see errors such as "PGSQL: query failed: server closed the connection
5314 unexpectedly". The problem has been fixed by closing all cached lookups
5315 in a queue runner before running a delivery process.
5317 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5318 seem to have the "const" qualifier which it has on other OS. I've
5321 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5322 *of the same type* (client or server) with the same public name that an
5323 error should be diagnosed.
5325 9. When Exim looked up a host name for an IP address, but failed to find the
5326 original IP address when looking up the host name (a safety check), it
5327 output the message "<ip address> does not match any IP for NULL", which was
5328 confusing, to say the least. The bug was that the host name should have
5329 appeared instead of "NULL".
5331 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5332 user that is built into the binary, and the -C or -D options is used, root
5333 privilege is dropped before the configuration file is read. In addition,
5334 logging is switched to stderr instead of the normal log files. If the
5335 configuration then re-defines the Exim user, the unprivileged environment
5336 is probably not what is expected, so Exim logs a panic warning message (but
5339 However, if deliver_drop_privilege is set, the unprivileged state may well
5340 be exactly what is intended, so the warning has been cut out in that case,
5341 and Exim is allowed to try to write to its normal log files.
5347 1. smtp_return_error_details was not giving details for temporary sender
5348 or receiver verification errors.
5350 2. Diagnose a configuration error if two authenticators have the same public
5353 3. Exim used not to create the message log file for a message until the first
5354 delivery attempt. This could be confusing when incoming messages were held
5355 for policy or load reasons. The message log file is now created at the time
5356 the message is received, and an initial "Received" line is written to it.
5358 4. The automatically generated man page for command line options had a minor
5359 bug that caused no ill effects; however, a more serious problem was that
5360 the procedure for building the man page automatically didn't always
5361 operate. Consequently, release 4.20 contains an out-of-date version. This
5362 shouldn't happen again.
5364 5. When building Exim with embedded Perl support, the script that builds the
5365 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5366 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5368 6. The freeze_tell option was not being used for messages that were frozen on
5369 arrival, either by an ACL or by local_scan().
5371 7. Added the smtp_incomplete_transaction log selector.
5373 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5374 was accepting AUTH without a new EHLO.
5376 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5377 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5380 10. Logging of TCP/IP connections (when configured) now happens in the main
5381 daemon process instead of the child process, so that the TCP/IP connection
5382 count is more accurate (but it can never be perfect).
5384 11. The use of "drop" in a nested ACL was not being handled correctly in the
5385 outer ACL. Now, if condition failure induced by the nested "drop" causes
5386 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5387 or "require"), the connection is dropped.
5389 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5390 that yield "discard" can now be used with an "accept" or a "discard" verb,
5391 but an error is generated for any others (because I can't see a useful way
5392 to define what should happen).
5394 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5395 are now processed in the same way as lines in the Exim configuration file.
5396 In particular, continuation lines are supported.
5398 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5400 15. Added -ti meaning -t -i.
5402 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5403 domains, and warn by logging if others are found.
5405 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5406 returned from GIFCONF when Exim is trying to find the list of interfaces on
5407 a host. The code in os.c has been modified to copy each ifreq to an aligned
5408 structure in all cases.
5410 Also, in some cases, the returned ifreq's were being copied to a 'struct
5411 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5412 means the last couple of bytes of an IPv6 address could be chopped if the
5413 ifreq contained only a normal sockaddr (14 bytes storage).
5415 18. Named domain lists were not supported in the hosts_treat_as_local option.
5416 An entry such as +xxxx was not recognized, and was treated as a literal
5419 19. Ensure that header lines added by a DATA ACL are included in the reject log
5420 if the ACL subsequently rejects the message.
5422 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5423 MD5 (which is deprecated).
5425 21. When testing a filter file using -bf, Exim was writing a message when it
5426 took the sender from a "From " line in the message, but it was not doing so
5427 when it took $return_path from a Return-Path: header line. It now does.
5429 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5430 with a valid header line field name (a series of printing characters
5431 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5433 23. Changed "disc" in the source to "disk" to conform to the documentation and
5434 the book and for uniformity.
5436 24. Ignore Sendmail's -Ooption=value command line item.
5438 25. When execve() failed while trying to run a command in a pipe transport,
5439 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5440 could be confused with a return value of 69 from the command itself. This
5441 has been changed to 127, the value the shell returns if it is asked to run
5442 a non-existent command. The wording for the related log line suggests a
5443 non-existent command as the problem.
5445 26. If received_header_text expands to an empty string, do not add a Received:
5446 header line to the message. (Well, it adds a token one on the spool, but
5447 marks it "old" so that it doesn't get used or transmitted.)
5449 27. Installed eximstats 1.28 (addition of -nt option).
5451 28. There was no check for failure on the call to getsockname() in the daemon
5452 code. This can fail if there is a shortage of resources on the system, with
5453 ENOMEM, for example. A temporary error is now given on failure.
5455 29. Contrary to the C standard, it seems that in some environments, the
5456 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5457 program. Exim now does this explicitly; it affects the formatting of
5458 timestamps using strftime().
5460 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5461 complaints. I've now initialized all the variables, to avoid this.
5463 32. Header lines added by a system filter were not being "seen" during
5464 transport-time rewrites.
5466 33. The info_callback() function passed to OpenSSL is set up with type void
5467 (*)(SSL *, int, int), as described somewhere. However, when calling the
5468 function (actually a macro) that sets it up, the type void(*)() is
5469 expected. I've put in a cast to prevent warnings from picky compilers.
5471 34. If a DNS black list lookup found a CNAME record, but there were no A
5472 records associated with the domain it pointed at, Exim crashed.
5474 35. If a DNS black list lookup returned more than one A record, Exim ignored
5475 all but the first. It now scans all returned addresses if a particular IP
5476 value is being sought. In this situation, the contents of the
5477 $dnslist_value variable are a list of all the addresses, separated by a
5480 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5481 to accept a host name and all its aliases if the forward lookup for any of
5482 them yielded the IP address of the incoming connection. Now it accepts only
5483 those names whose forward lookup yields the correct IP address. Any other
5484 names are discarded. This closes a loophole whereby a rogue DNS
5485 administrator could create reverse DNS records to break through a
5486 wildcarded host restriction in an ACL.
5488 37. If a user filter or a system filter that ran in a subprocess used any of
5489 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5490 the wrong values were passed to the pipe command ($thisaddress had the
5491 value of $0, $0 had the value of $1, etc). This bug was introduced by
5492 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5494 38. Improved the line breaking for long SMTP error messages from ACLs.
5495 Previously, if there was no break point between 40 and 75 characters, Exim
5496 left the rest of the message alone. Two changes have been made: (a) I've
5497 reduced the minimum length to 35 characters; (b) if it can't find a break
5498 point between 35 and 75 characters, it looks ahead and uses the first one
5499 that it finds. This may give the occasional overlong line, but at least the
5500 remaining text gets split now.
5502 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5503 file descriptors might be low, and that setting 1000 would always raise it.
5504 It turns out that in some environments, the limit is already over 1000 and
5505 that lowering it causes trouble. So now Exim takes care not to decrease it.
5507 40. When delivering a message, the value of $return_path is set to $sender_
5508 address at the start of routing (routers may change the value). By an
5509 oversight, this default was not being set up when an address was tested by
5510 -bt or -bv, which affected the outcome if any router or filter referred to
5513 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5514 log only when "message" or "log_message" are set. However, if one of the
5515 conditions was an address verification, or a call to a nested ACL, the
5516 messages generated by the underlying test were being passed through. This
5517 no longer happens. The underlying message is available in $acl_verify_
5518 message for both "message" and "log_message" expansions, so it can be
5519 passed through if needed.
5521 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5522 new expansion $bh_ to give the encoded byte string without charset
5523 translation. Translation happens only if iconv() is available; HAVE_ICONV
5524 indicates this at build time. HEADERS_CHARSET gives the charset to
5525 translate to; headers_charset can change it in the configuration, and
5526 "headers charset" can change it in an individual filter file.
5528 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5529 that creates RFC 2047 encoded "words" labels them as that charset instead
5530 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5531 expansion operator; (ii) when Exim creates a From: line for a local
5532 message; (iii) when a header line is rewritten to include a "phrase" part.
5534 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5535 buggy, causing it to skip the first lines of messages whose message ID
5536 ended in 'D'. This would not have bitten before Exim release 4.14, because
5537 message IDs were unlikely to end in 'D' before then. The effect was to have
5538 incorrect size information for certain domains.
5540 45. #include "config.h" was missing at the start of the crypt16.c module. This
5541 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5544 46. If there was a timeout during a "random" callout check, Exim treated it as
5545 a failure of the random address, and carried on sending RSET and the real
5546 address. If the delay was just some slowness somewhere, the response to the
5547 original RCPT would be taken as a response to RSET and so on, causing
5548 mayhem of various kinds.
5550 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5551 when I implemented it. It didn't allow for the fact that some option values
5552 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5553 the right test for positive values.
5555 48. Domain names in DNS records are case-independent. Exim always looks them up
5556 in lower case. Some resolvers return domain names in exactly the case they
5557 appear in the zone file, that is, they may contain uppercase letters. Not
5558 all resolvers do this - some return always lower case. Exim was treating a
5559 change of case by a resolver as a change of domain, similar to a widening
5560 of a domain abbreviation. This triggered its re-routing code and so it was
5561 trying to route what was effectively the same domain again. This normally
5562 caused routing to fail (because the router wouldn't handle the domain
5563 twice). Now Exim checks for this case specially, and just changes the
5564 casing of the domain that it ultimately uses when it transmits the message
5567 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5570 50. If a filter generated a file delivery with a non-absolute name (possible if
5571 no home directory exists for the router), the forbid_file option was not
5574 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5575 the existing equality matching.
5577 52. Exim was using ints instead of ino_t variables in some places where it was
5578 dealing with inode numbers.
5580 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5581 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5582 called TMPDIR, and if it finds it is different, it changes its value.
5584 54. The smtp_printf() function is now made available to local_scan() so
5585 additional output lines can be written before returning. There is also an
5586 smtp_fflush() function to enable the detection of a dropped connection.
5587 The variables smtp_input and smtp_batched_input are exported to
5590 55. Changed the default runtime configuration: the message "Unknown user"
5591 has been removed from the ACL, and instead placed on the localuser router,
5592 using the cannot_route_message feature. This means that any verification
5593 failures that generate their own messages won't get overridden. Similarly,
5594 the "Unrouteable address" message that was in the ACL for unverifiable
5595 relay addresses has also been removed.
5597 56. Added hosts_avoid_esmtp to the smtp transport.
5599 57. The exicyclog script was not checking for the esoteric option
5600 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5601 will work only if exicyclog is run under the appropriate euid.
5603 58. Following a discussion on the list, the rules by which Exim recognises line
5604 endings on incoming messages have been changed. The -dropcr and drop_cr
5605 options are now no-ops, retained only for backwards compatibility. The
5606 following line terminators are recognized: LF CRLF CR. However, special
5607 processing applies to CR:
5609 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5610 nor a local message in the state where . is a terminator.
5612 (ii) If a bare CR is encountered in a header line, an extra space is added
5613 after the line terminator so as not to end the header. The reasoning
5614 behind this is that bare CRs in header lines are most likely either
5615 to be mistakes, or people trying to play silly games.
5617 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5618 was being incorrectly given as 18 bytes larger than it should have been.
5619 This is a VOB (very old bug).
5621 60. This may never have affected anything current, but just in case it has:
5622 When the local host is found other than at the start of a list of hosts,
5623 the local host, those with the same MX, and any that follow, are discarded.
5624 When the list in question was part of a longer list of hosts, the following
5625 hosts (not currently being processed) were also being discarded. This no
5626 longer happens. I'm not sure if this situation could ever has previously
5629 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5632 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5633 header. This is something that is recommended in a new Internet Draft, and
5634 is something that is documented as being done by Sendmail. There are two
5635 possible values. For messages generated by the autoreply transport, Exim
5638 Auto-Submitted: auto-replied
5640 whereas for all other generated messages (e.g. bounces) it adds
5642 Auto-Submitted: auto-generated
5644 63. The "personal" condition in filters now includes a test for the
5645 Auto-Submitted: header. If it contains the string "auto-" the message it
5646 not considered personal.
5648 64. Added rcpt_include_affixes as a generic transport option.
5650 65. Added queue_only_override (default true).
5652 66. Added the syslog_duplication option.
5654 67. If what should have been the first header line of a message consisted of
5655 a space followed by a colon, Exim was mis-interpreting it as a header line.
5656 It isn't of course - it is syntactically invalid and should therefore be
5657 treated as the start of the message body. The misbehaviour could have
5658 caused a number of strange effects, including loss of data in subsequent
5659 header lines, and spool format errors.
5661 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5662 client host had authenticated. This control can now be exercised by an ACL
5663 for more flexibility.
5665 69. By default, callouts do not happen when testing with -bh. There is now a
5666 variant, -bhc, which does actually run the callout code, including
5667 consulting and updating the callout cache.
5669 70. Added support for saslauthd authentication, courtesy of Alexander
5672 71. If statvfs() failed on the spool or log directories while checking their
5673 size for availability, Exim confusingly gave the error "space shortage".
5674 Furthermore, in debugging mode it crashed with a floating point exception.
5675 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5676 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5677 serious problem, Exim now writes to the main and panic logs when this
5678 happens, with details of the failure. It then refuses to accept the
5679 incoming message, giving the message "spool directory problem" or "log
5680 directory problem" with a 421 code for SMTP messages.
5682 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5683 0, 1, and 2 exist, because some OS complain for execs without them (see
5684 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5685 descriptors. However, the code omitted to check that the open succeeded,
5686 causing mysterious errors if for some reason the permissions on /dev/null
5687 got screwed. Now Exim writes a message to the main and panic logs, and
5688 bombs out if it can't open /dev/null.
5690 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5691 interact so that it is all more flexible. It is supposed to remain
5692 backwards compatible. Also added extra_local_interfaces.
5694 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5695 to bomb out with an assertion failure - to the client this appears as a
5696 connection drop. This problem occurs in the part of the code that was taken
5697 from the Samba project. Fortunately, the assertion is in a very simple
5698 function, so I have fixed this by reproducing the function inline in the
5699 one place where it is called, and arranging for authentication to fail
5700 instead of killing the process with assert().
5702 75. The SPA client code was not working when the server requested OEM rather
5703 than Unicode encoding.
5705 76. Added code to make require_files with a specific uid setting more usable in
5706 the case where statting the file as root fails - usually a non-root-mounted
5707 NFS file system. When this happens and the failure is EACCES, Exim now
5708 forks a subprocess and does the per-uid checking as the relevant uid.
5710 77. Added process_log_path.
5712 78. If log_file_path was not explicitly set, a setting of check_log_space or
5713 check_log_inodes was ignored.
5715 79. If a space check for the spool or log partitions fails, the incident is now
5716 logged. Of course, in the latter case the data may get lost...
5718 80. Added the %p formatting code to string_format() so that it can be used to
5719 print addresses in debug_print(). Adjusted all the address printing in the
5720 debugging in store.c to use %p rather than %d.
5722 81. There was a concern that a line of code in smtp_in.c could overflow a
5723 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5724 initially expressed, the concern was not well-founded, because trailing
5725 spaces are removed early. However, if the trailing spaces were followed by
5726 a NULL, they did not get removed, so the overflow was possible. Two fixes
5729 (a) I re-wrote the offending code in a cleaner fashion.
5730 (b) If an incoming SMTP command contains a NULL character, it is rejected
5733 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5734 runs initgroups(), so that if the Exim user is in any additional groups,
5735 they will be used during message reception.
5741 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob