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.
44 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
46 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
48 JH/02 Support "G" suffix to numbers in ${if comparisons.
50 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
52 NM/01 Bugzilla 1197 - Spec typo
53 Bugzilla 1196 - Spec examples corrections
55 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
57 PP/09 Add gnutls_enable_pkcs11 option.
59 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
60 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
62 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
63 "acl {{name}{arg}...}", and optional args on acl condition
66 JH/05 Permit multiple router/transport headers_add/remove lines.
71 PP/01 Handle short writes when writing local log-files.
72 In practice, only affects FreeBSD (8 onwards).
73 Bugzilla 1053, with thanks to Dmitry Isaikin.
75 NM/01 Bugzilla 949 - Documentation tweak
77 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
80 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
82 PP/02 Implemented gsasl authenticator.
84 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
86 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
87 `pkg-config foo` for cflags/libs.
89 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
90 with rest of GSASL and with heimdal_gssapi.
92 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
93 `pkg-config foo` for cflags/libs for the TLS implementation.
95 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
96 properties get this fed in as external SSF. A number of robustness
97 and debugging improvements to the cyrus_sasl authenticator.
99 PP/08 cyrus_sasl server now expands the server_realm option.
101 PP/09 Bugzilla 1214 - Log authentication information in reject log.
102 Patch by Jeremy Harris.
104 PP/10 Added dbmjz lookup type.
106 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
108 PP/12 MAIL args handles TAB as well as SP, for better interop with
109 non-compliant senders.
110 Analysis and variant patch by Todd Lyons.
112 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
113 Bug report from Lars Müller <lars@samba.org> (via SUSE),
114 Patch from Dirk Mueller <dmueller@suse.com>
116 PP/13 tls_peerdn now print-escaped for spool files.
117 Observed some $tls_peerdn in wild which contained \n, which resulted
118 in spool file corruption.
120 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
121 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
122 or write after TLS renegotiation, which otherwise led to messages
125 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
126 as a tracking header (ie: a signed header comes before the signature).
127 Patch from Wolfgang Breyha.
129 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
130 comma-sep list; embedded commas doubled.
132 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
134 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
136 Report and patch from Dmitry Banschikov.
138 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
139 Removed SSL_clear() after SSL_new() which led to protocol negotiation
140 failures. We appear to now support TLS1.1+ with Exim.
142 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
143 lets Exim select keys and certificates based upon TLS SNI from client.
144 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
145 before an outbound SMTP session. New log_selector, +tls_sni.
147 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
148 NULL dereference. Report and patch from Alun Jones.
150 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
151 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
152 Not seeing resolver debug output on NetBSD, but suspect this is a
153 resolver implementation change.
155 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
156 Left warnings. Added "eximon gdb" invocation mode.
158 PP/21 Defaulting "accept_8bitmime" to true, not false.
160 PP/22 Added -bw for inetd wait mode support.
162 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
163 locate the relevant includes and libraries. Made this the default.
165 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
166 Bugzilla 1246, report and most of solution from Tomasz Kusy.
168 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
169 This may cause build issues on older platforms.
171 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
172 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
173 gnutls_require_mac & gnutls_require_protocols (no longer supported).
174 Added SNI support via GnuTLS too.
175 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
177 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
179 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
180 Applied second patch from Janne, implementing suggestion to default
181 multiple-strings-in-record handling to match SPF spec.
183 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
185 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
186 read-only, out of scope).
187 Patch from Wolfgang Breyha, report from Stuart Northfield.
189 PP/29 Fix three issues highlighted by clang analyser static analysis.
190 Only crash-plausible issue would require the Cambridge-specific
191 iplookup router and a misconfiguration.
192 Report from Marcin Mirosław.
194 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
196 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
197 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
198 As part of this, removing so much warning spew let me fix some minor
199 real issues in debug logging.
201 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
202 assignment on my part. Fixed.
204 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
205 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
206 Janne Snabb (who went above and beyond: thank you).
208 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
209 string otherwise requires a connection and a bunch more work and it's
210 relatively easy to get wrong. Should also expose TLS library linkage
213 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
214 64-bit ${eval} (JH/03).
216 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
217 GNU libc to support some of the 64-bit stuff, should not lead to
218 conflicts. Defined before os.h is pulled in, so if a given platform
219 needs to override this, it can.
221 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
222 protection layer was required, which is not implemented.
223 Bugzilla 1254, patch from Wolfgang Breyha.
225 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
226 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
227 tls_dhparam take prime identifiers. Also unbreak combination of
228 OpenSSL+DH_params+TLSSNI.
230 PP/39 Disable SSLv2 by default in OpenSSL support.
236 PP/01 Solaris build fix for Oracle's LDAP libraries.
237 Bugzilla 1109, patch from Stephen Usher.
239 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
241 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
244 TF/02 Fix a couple more cases where we did not log the error message
245 when unlink() failed. See also change 4.74-TF/03.
247 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
248 lock up or crash if it happened to be inside a call to libc when it
249 got a SIGUSR1 from exiwhat.
251 The SIGUSR1 handler appends the current process status to the process
252 log which is later printed by exiwhat. It used to use the general
253 purpose logging code to do this, but several functions it calls are
254 not safe for signals.
256 The new output code in the SIGUSR1 handler is specific to the process
257 log, and simple enough that it's easy to inspect for signal safety.
258 Removing some special cases also simplifies the general logging code.
259 Removing the spurious timestamps from the process log simplifies
262 TF/04 Improved ratelimit ACL condition.
264 The /noupdate option has been deprecated in favour of /readonly which
265 has clearer semantics. The /leaky, /strict, and /readonly update modes
266 are mutually exclusive. The update mode is no longer included in the
267 database key; it just determines when the database is updated. (This
268 means that when you upgrde Exim will forget old rate measurements.)
270 Exim now checks that the per_* options are used with an update mode that
271 makes sense for the current ACL. For example, when Exim is processing a
272 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
273 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
274 must specify per_mail/readonly. If you omit the update mode it defaults to
275 /leaky where that makes sense (as before) or /readonly where required.
277 The /noupdate option is now undocumented but still supported for
278 backwards compatibility. It is equivalent to /readonly except that in
279 ACLs where /readonly is required you may specify /leaky/noupdate or
280 /strict/noupdate which are treated the same as /readonly.
282 A useful new feature is the /count= option. This is a generalization
283 of the per_byte option, so that you can measure the throughput of other
284 aggregate values. For example, the per_byte option is now equivalent
285 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
287 The per_rcpt option has been generalized using the /count= mechanism
288 (though it's more complicated than the per_byte equivalence). When it is
289 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
290 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
291 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
292 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
293 Note that using per_rcpt with a non-readonly update mode in more than
294 one ACL will cause the recipients to be double-counted. (The per_mail
295 and per_byte options don't have this problem.)
297 The handling of very low rates has changed slightly. If the computed rate
298 is less than the event's count (usually one) then this event is the first
299 after a long gap. In this case the rate is set to the same as this event's
300 count, so that the first message of a spam run is counted properly.
302 The major new feature is a mechanism for counting the rate of unique
303 events. The new per_addr option counts the number of different
304 recipients that someone has sent messages to in the last time period. It
305 behaves like per_rcpt if all the recipient addresses are different, but
306 duplicate recipient addresses do not increase the measured rate. Like
307 the /count= option this is a general mechanism, so the per_addr option
308 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
309 example, measure the rate that a client uses different sender addresses
310 with the options per_mail/unique=$sender_address. There are further
311 details in the main documentation.
313 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
315 TF/06 Removed a few PCRE remnants.
317 TF/07 Automatically extract Exim's version number from tags in the git
318 repository when doing development or release builds.
320 PP/02 Raise smtp_cmd_buffer_size to 16kB.
321 Bugzilla 879. Patch from Paul Fisher.
323 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
324 Heavily based on revision 40f9a89a from Simon Arlott's tree.
327 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
329 PP/05 Variable $av_failed, true if the AV scanner deferred.
330 Bugzilla 1078. Patch from John Horne.
332 PP/06 Stop make process more reliably on build failure.
333 Bugzilla 1087. Patch from Heiko Schlittermann.
335 PP/07 Make maildir_use_size_file an _expandable_ boolean.
336 Bugzilla 1089. Patch from Heiko Schlittermann.
338 PP/08 Handle ${run} returning more data than OS pipe buffer size.
339 Bugzilla 1131. Patch from Holger Weiß.
341 PP/09 Handle IPv6 addresses with SPF.
342 Bugzilla 860. Patch from Wolfgang Breyha.
344 PP/10 GnuTLS: support TLS 1.2 & 1.1.
346 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
349 PP/11 match_* no longer expand right-hand-side by default.
350 New compile-time build option, EXPAND_LISTMATCH_RHS.
351 New expansion conditions, "inlist", "inlisti".
353 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
355 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
357 PP/14 fix log_write() format string regression from TF/03.
358 Bugzilla 1152. Patch from Dmitry Isaikin.
364 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
366 PP/02 Harmonised TLS library version reporting; only show if debugging.
367 Layout now matches that introduced for other libraries in 4.74 PP/03.
369 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
371 PP/04 New "dns_use_edns0" global option.
373 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
376 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
377 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
379 TK/01 Updated PolarSSL code to 0.14.2.
380 Bugzilla 1097. Patch from Andreas Metzler.
382 PP/07 Catch divide-by-zero in ${eval:...}.
385 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
388 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
389 format-string attack -- SECURITY: remote arbitrary code execution.
391 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
392 time unintentionally subject to list matching rules, letting the header
393 cause arbitrary Exim lookups (of items which can occur in lists, *not*
394 arbitrary string expansion). This allowed for information disclosure.
396 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
397 INT_MIN/-1 -- value coerced to INT_MAX.
403 NM/01 Workround for PCRE version dependency in version reporting
406 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
407 This fixes portability to compilers other than gcc, notably
408 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
410 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
411 makefiles for portability to HP-UX and POSIX correctness.
413 PP/01 Permit LOOKUP_foo enabling on the make command-line.
414 Also via indented variable definition in the Makefile.
415 (Debugging by Oliver Heesakkers).
417 PP/02 Restore caching of spamd results with expanded spamd_address.
418 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
420 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
421 Improves build reliability. Fix from: Frank Elsner
423 NM/02 Fix wide character breakage in the rfc2047 coding
424 Fixes bug 1064. Patch from Andrey N. Oktyabrski
426 NM/03 Allow underscore in dnslist lookups
427 Fixes bug 1026. Patch from Graeme Fowler
429 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
430 Code patches from Adam Ciarcinski of NetBSD.
432 NM/04 Fixed exiqgrep to cope with mailq missing size issue
435 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
436 is logged, to avoid truncation. Patch from John Horne.
438 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
439 Patch from Jakob Hirsch.
441 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
442 SQL string expansion failure details.
443 Patch from Andrey Oktyabrski.
445 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
446 Patch from Simon Arlott.
448 PP/09 New lookups functionality failed to compile on old gcc which rejects
449 extern declarations in function scope.
450 Patch from Oliver Fleischmann
452 PP/10 Use sig_atomic_t for flags set from signal handlers.
453 Check getgroups() return and improve debugging.
454 Fixed developed for diagnosis in bug 927 (which turned out to be
457 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
458 Patch from Mark Zealey.
460 PP/12 Bugzilla 1056: Improved spamd server selection.
461 Patch from Mark Zealey.
463 PP/13 Bugzilla 1086: Deal with maildir quota file races.
464 Based on patch from Heiko Schlittermann.
466 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
467 Patch from Uwe Doering, sign-off by Michael Haardt.
469 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
470 variable declaration deep within a block. Bug and patch from
473 PP/15 lookups-Makefile IRIX compatibilty coercion.
475 PP/16 Make DISABLE_DKIM build knob functional.
477 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
478 Patch by Simon Arlott
480 TF/03 Fix valgrind.h portability to C89 compilers that do not support
481 variable argument macros. Our copy now differs from upstream.
487 TF/01 Failure to get a lock on a hints database can have serious
488 consequences so log it to the panic log.
490 TF/02 Log LMTP confirmation messages in the same way as SMTP,
491 controlled using the smtp_confirmation log selector.
493 TF/03 Include the error message when we fail to unlink a spool file.
495 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
496 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
497 for maintaining out-of-tree patches for some time.
499 PP/01 Bugzilla 139: Documentation and portability issues.
500 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
501 Handle per-OS dynamic-module compilation flags.
503 PP/02 Let /dev/null have normal permissions.
504 The 4.73 fixes were a little too stringent and complained about the
505 permissions on /dev/null. Exempt it from some checks.
506 Reported by Andreas M. Kirchwitz.
508 PP/03 Report version information for many libraries, including
509 Exim version information for dynamically loaded libraries. Created
510 version.h, now support a version extension string for distributors
511 who patch heavily. Dynamic module ABI change.
513 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
514 privilege escalation vulnerability whereby the Exim run-time user
515 can cause root to append content of the attacker's choosing to
518 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
521 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
522 If dropping privileges for untrusted macros, we disabled normal logging
523 on the basis that it would fail; for the Exim run-time user, this is not
524 the case, and it resulted in successful deliveries going unlogged.
525 Fixed. Reported by Andreas Metzler.
531 PP/01 Date: & Message-Id: revert to normally being appended to a message,
532 only prepend for the Resent-* case. Fixes regression introduced in
533 Exim 4.70 by NM/22 for Bugzilla 607.
535 PP/02 Include check_rfc2047_length in configure.default because we're seeing
536 increasing numbers of administrators be bitten by this.
538 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
540 PP/03 Bugzilla 994: added openssl_options main configuration option.
542 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
544 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
546 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
548 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
549 without a peer certificate, leading to a segfault because of an
550 assumption that peers always have certificates. Be a little more
551 paranoid. Problem reported by Martin Tscholak.
553 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
554 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
555 NB: ClamAV planning to remove STREAM in "middle of 2010".
556 CL also introduces -bmalware, various -d+acl logging additions and
557 more caution in buffer sizes.
559 PP/09 Implemented reverse_ip expansion operator.
561 PP/10 Bugzilla 937: provide a "debug" ACL control.
563 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
565 PP/12 Bugzilla 973: Implement --version.
567 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
569 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
571 PP/15 Bugzilla 816: support multiple condition rules on Routers.
573 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
574 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
575 ignore trailing whitespace.
577 JJ/02 prevent non-panic DKIM error from being sent to paniclog
579 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
582 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
583 Notification from Dr Andrew Aitchison.
585 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
586 ExtendedDetectionInfo response format.
587 Notification from John Horne.
589 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
592 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
593 XSL and documented dependency on system catalogs, with examples of how
596 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
599 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
600 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
601 configuration file which is writeable by the Exim user or group.
603 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
604 of configuration files to cover files specified with the -C option if
605 they are going to be used with root privileges, not just the default
608 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
609 option (effectively making it always true).
611 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
612 files to be used while preserving root privileges.
614 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
615 that rogue child processes cannot use them.
617 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
618 run-time user, instead of root.
620 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
621 Exim run-time user without dropping privileges.
623 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
624 result string, instead of calling string_vformat() twice with the same
627 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
628 for other users. Others should always drop root privileges if they use
629 -C on the command line, even for a whitelisted configure file.
631 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
633 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
639 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
640 $data_path, and $header_path variables; fixed documentation bugs and
643 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
644 exipick to access non-standard spools, including the "frozen" queue
647 NM/01 Bugzilla 965: Support mysql stored procedures.
648 Patch from Alain Williams
650 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
652 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
653 Patch from Andreas Metzler
655 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
656 Patch from Kirill Miazine
658 NM/05 Bugzilla 671: Added umask to procmail example.
660 JJ/03 installed exipick 20100323.0, fixing doc bug
662 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
663 directory. Notification and patch from Dan Rosenberg.
665 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
667 TK/02 Improve log output when DKIM signing operation fails.
669 MH/01 Treat the transport option dkim_domain as a colon separated
670 list, not as a single string, and sign the message with each element,
671 omitting multiple occurences of the same signer.
673 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
674 Bugzilla 985, 986. Patch by Simon Arlott
676 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
677 Patch by Simon Arlott
679 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
680 MBX locking. Notification from Dan Rosenberg.
686 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
688 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
690 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
692 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
694 NM/04 Bugzilla 915: Flag broken perl installation during build.
700 TK/01 Added patch by Johannes Berg that expands the main option
701 "spamd_address" if it starts with a dollar sign.
703 TK/02 Write list of recipients to X-Envelope-Sender header when building
704 the mbox-format spool file for content scanning (suggested by Jakob
707 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
708 (http://www.dcc-servers.net/) support via dccifd. Activated by
709 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
711 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
712 by Mark Daniel Reidel <mr@df.eu>.
714 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
715 When building exim an external PCRE library is now needed -
716 PCRE is a system library on the majority of modern systems.
717 See entry on PCRE_LIBS in EDITME file.
719 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
720 conversation. Added nologin parameter to request.
721 Patch contributed by Kirill Miazine.
723 TF/01 Do not log submission mode rewrites if they do not change the address.
725 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
727 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
728 log files in place. Contributed by Roberto Lima.
730 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
732 TF/03 Bugzilla 615: When checking the local_parts router precondition
733 after a local_part_suffix or local_part_prefix option, Exim now
734 does not use the address's named list lookup cache, since this
735 contains cached lookups for the whole local part.
737 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
738 Robert Millan. Documentation is in experimental-spec.txt.
740 TF/04 Bugzilla 668: Fix parallel build (make -j).
742 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
744 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
745 Patch provided by Jan Srzednicki.
747 TF/05 Leading white space used to be stripped from $spam_report which
748 wrecked the formatting. Now it is preserved.
750 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
751 that they are available at delivery time.
753 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
755 TF/08 TLS error reporting now respects the incoming_interface and
756 incoming_port log selectors.
758 TF/09 Produce a more useful error message if an SMTP transport's hosts
759 setting expands to an empty string.
761 NM/06 Bugzilla 744: EXPN did not work under TLS.
762 Patch provided by Phil Pennock.
764 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
765 Patch provided by Richard Godbee.
767 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
768 acl_smtp_notquit, added index entry.
770 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
771 Patch provided by Eugene Bujak.
773 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
774 accept(). Patch provided by Maxim Dounin.
776 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
777 Patch provided by Phil Pennock.
779 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
781 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
782 Patch provided by Brad "anomie" Jorsch.
784 NM/14 Bugzilla 622: Added timeout setting to transport filter.
785 Patch provided by Dean Brooks.
787 TK/05 Add native DKIM support (does not depend on external libraries).
789 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
790 Patch provided by Graeme Fowler.
792 NM/16 Bugzilla 851: Documentation example syntax fix.
794 NM/17 Changed NOTICE file to remove references to embedded PCRE.
796 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
799 NM/19 Bugzilla 745: TLS version reporting.
800 Patch provided by Phil Pennock.
802 NM/20 Bugzilla 167: bool: condition support.
803 Patch provided by Phil Pennock.
805 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
806 clients. Patch provided by Phil Pennock.
808 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
809 Patch provided by Brad "anomie" Jorsch.
811 NM/23 Bugzilla 687: Fix misparses in eximstats.
812 Patch provided by Heiko Schlittermann.
814 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
815 Patch provided by Heiko Schlittermann.
817 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
818 plus update to original patch.
820 NM/26 Bugzilla 799: Documentation correction for ratelimit.
822 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
823 Patch provided by David Brownlee.
825 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
827 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
829 NM/30 Bugzilla 888: TLS documentation bugfixes.
831 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
833 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
834 Unlike the original bugzilla I have changed all shell scripts in src tree.
836 NM/33 Bugzilla 898: Transport filter timeout fix.
837 Patch by Todd Rinaldo.
839 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
840 Patch by Serge Demonchaux.
842 NM/35 Bugzilla 39: Base64 decode bug fixes.
843 Patch by Jakob Hirsch.
845 NM/36 Bugzilla 909: Correct connect() call in dcc code.
847 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
849 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
851 NM/39 Bugzilla 911: Fixed MakeLinks build script.
857 TK/01 Add preliminary DKIM support. Currently requires a forked version of
858 ALT-N's libdkim that I have put here:
859 http://duncanthrax.net/exim-experimental/
861 Note to Michael Haardt: I had to rename some vars in sieve.c. They
862 were called 'true' and it seems that C99 defines that as a reserved
863 keyword to be used with 'bool' variable types. That means you could
864 not include C99-style headers which use bools without triggering
865 build errors in sieve.c.
867 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
868 as mailq or other aliases. Changed the --help handling significantly
869 to do whats expected. exim_usage() emits usage/help information.
871 SC/01 Added the -bylocaldomain option to eximstats.
873 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
875 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
877 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
883 PH/01 Another patch from the Sieve maintainer.
885 PH/02 When an IPv6 address is converted to a string for single-key lookup
886 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
887 dots are used instead of colons so that keys in lsearch files need not
888 contain colons. This was done some time before quoting was made available
889 in lsearch files. However, iplsearch files do require colons in IPv6 keys
890 (notated using the quote facility) so as to distinguish them from IPv4
891 keys. This meant that lookups for IP addresses in host lists did not work
892 for iplsearch lookups.
894 This has been fixed by arranging for IPv6 addresses to be expressed with
895 colons if the lookup type is iplsearch. This is not incompatible, because
896 previously such lookups could never work.
898 The situation is now rather anomolous, since one *can* have colons in
899 ordinary lsearch keys. However, making the change in all cases is
900 incompatible and would probably break a number of configurations.
902 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
905 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
906 conversion specification without a maximum field width, thereby enabling
907 a rogue spamd server to cause a buffer overflow. While nobody in their
908 right mind would setup Exim to query an untrusted spamd server, an
909 attacker that gains access to a server running spamd could potentially
910 exploit this vulnerability to run arbitrary code as the Exim user.
912 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
913 $primary_hostname instead of what libspf2 thinks the hosts name is.
915 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
916 a directory entry by the name of the lookup key. Previously, if a
917 symlink pointed to a non-existing file or a file in a directory that
918 Exim lacked permissions to read, a lookup for a key matching that
919 symlink would fail. Now it is enough that a matching directory entry
920 exists, symlink or not. (Bugzilla 503.)
922 PH/03 The body_linecount and body_zerocount variables are now exported in the
925 PH/04 Added the $dnslist_matched variable.
927 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
928 This means they are set thereafter only if the connection becomes
931 PH/06 Added the client_condition to authenticators so that some can be skipped
932 by clients under certain conditions.
934 PH/07 The error message for a badly-placed control=no_multiline_responses left
935 "_responses" off the end of the name.
937 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
939 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
940 (without spaces) instead of just copying the configuration text.
942 PH/10 Added the /noupdate option to the ratelimit ACL condition.
944 PH/11 Added $max_received_linelength.
946 PH/12 Added +ignore_defer and +include_defer to host lists.
948 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
949 way in which PCRE > 7.0 is built.
951 PH/14 Implemented queue_only_load_latch.
953 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
954 MAIL command. The effect was to mangle the value on 64-bit systems.
956 PH/16 Another patch from the Sieve maintainer.
958 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
960 PH/18 If a system quota error occurred while trying to create the file for
961 a maildir delivery, the message "Mailbox is full" was not appended to the
962 bounce if the delivery eventually timed out. Change 4.67/27 below applied
963 only to a quota excession during the actual writing of the file.
965 PH/19 It seems that peer DN values may contain newlines (and other non-printing
966 characters?) which causes problems in log lines. The DN values are now
967 passed through string_printing() before being added to log lines.
969 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
970 and InterBase are left for another time.)
972 PH/21 Added message_body_newlines option.
974 PH/22 Guard against possible overflow in moan_check_errorcopy().
976 PH/23 POSIX allows open() to be a macro; guard against that.
978 PH/24 If the recipient of an error message contained an @ in the local part
979 (suitably quoted, of course), incorrect values were put in $domain and
980 $local_part during the evaluation of errors_copy.
986 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
987 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
990 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
991 issue a MAIL command.
993 PH/02 In an ACL statement such as
995 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
997 if a client was not listed at all, or was listed with a value other than
998 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
999 the condition was not true (as it should be), so access was not denied.
1000 The bug was that the ! inversion was incorrectly passed on to the second
1001 item. This has been fixed.
1003 PH/03 Added additional dnslists conditions == and =& which are different from
1004 = and & when the dns lookup returns more than one IP address.
1006 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1007 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1009 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1010 FSYNC, which compiles an option called disable_fsync that allows for
1011 bypassing fsync(). The documentation is heavily laced with warnings.
1013 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1015 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1016 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1017 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1018 including adding "make clean"; (3) Added -fPIC when compiling the test
1019 dynamically loaded module, to get rid of a warning.
1021 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1022 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1023 = 0s. The bug is otherwise harmless.
1025 PH/07 There was a bug in the dovecot authenticator such that the value of
1026 $auth1 could be overwritten, and so not correctly preserved, after a
1027 successful authentication. This usually meant that the value preserved by
1028 the server_setid option was incorrect.
1030 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1032 PH/09 Installed PCRE release 7.0.
1034 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1035 run for batched SMTP input. It is now run at the start of every message
1036 in the batch. While fixing this I discovered that the process information
1037 (output by running exiwhat) was not always getting set for -bs and -bS
1038 input. This is fixed, and it now also says "batched" for BSMTP.
1040 PH/11 Added control=no_pipelining.
1042 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1043 patch, slightly modified), and move the expansion of helo_data till after
1044 the connection is made in the smtp transport (so it can use these
1047 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1049 PH/14 Added log_selector = +pid.
1051 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1053 PH/16 Add ${if forany and ${if forall.
1055 PH/17 Added dsn_from option to vary the From: line in DSNs.
1057 PH/18 Flush SMTP output before performing a callout, unless control =
1058 no_callout_flush is set.
1060 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1061 was true (the default) a successful delivery failed to delete the retry
1062 item, thus causing premature timeout of the address. The bug is now
1065 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1067 PH/21 Long custom messages for fakedefer and fakereject are now split up
1068 into multiline reponses in the same way that messages for "deny" and
1069 other ACL rejections are.
1071 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1072 with slight modification.
1074 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1075 draft, changing the syntax and factoring some duplicate code".
1077 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1078 for deliveries of the second and subsequent messages over the same SMTP
1081 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1082 ${reduce, with only minor "tidies".
1084 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1086 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1087 expansion side effects.
1089 PH/27 When a message times out after an over-quota error from an Exim-imposed
1090 quota, the bounce message says "mailbox is full". This message was not
1091 being given when it was a system quota that was exceeded. It now should
1094 MH/03 Made $recipients available in local_scan(). local_scan() already has
1095 better access to the recipient list through recipients_list[], but
1096 $recipients can be useful in postmaster-provided expansion strings.
1098 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1099 in the case of a MAIL command with additional options following the
1100 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1101 were accidentally chopped off.
1103 PH/29 SMTP synchronization checks are implemented when a command is read -
1104 there is a check that no more input is waiting when there shouldn't be
1105 any. However, for some commands, a delay in an ACL can mean that it is
1106 some time before the response is written. In this time, more input might
1107 arrive, invalidly. So now there are extra checks after an ACL has run for
1108 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1109 pipelining has not been advertised.
1111 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1113 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1114 returned in the SMTP response when smtp_return_error_details was false.
1115 This has been fixed.
1117 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1118 instead of the C I/O that was originally supplied, because problems were
1119 reported on Solaris.
1121 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1122 Exim which did not show up earlier: it was assuming that a call to
1123 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1124 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1125 was a macro that became an assignment, so it seemed to work. This has
1126 changed to a proper function call with a void return, hence the compile
1127 error. Exim's code has been fixed.
1129 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1132 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1134 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1136 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1137 because it turns out to be too big or there is a timeout) there is no
1138 "Completed" line in the log. When some messages of this type were
1139 selected by exigrep, they were listed as "not completed". Others were
1140 picked up by some special patterns. I have improved the selection
1141 criteria to be more general.
1143 PH/38 The host_find_failed option in the manualroute router can now be set
1144 to "ignore", to completely ignore a host whose IP address cannot be
1145 found. If all hosts are ignored, the behaviour is controlled by the new
1146 host_all_ignored option.
1148 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1149 homing or because of multiple MX records with /mx) generated more than
1150 one IP address, and the following item turned out to be the local host,
1151 all the secondary addresses of the first item were incorrectly removed
1152 from the list, along with the local host and any following hosts (which
1153 is what is supposed to happen).
1155 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1156 whoever called Exim into the -H file. In the case of the daemon it was
1157 behaving confusingly. When first started, it used values for whoever
1158 started the daemon, but after a SIGHUP it used the Exim user (because it
1159 calls itself on a restart). I have changed the code so that it now always
1162 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1163 message are rejected with the same error (e.g. no authentication or bad
1164 sender address), and a DATA command is nevertheless sent (as can happen
1165 with PIPELINING or a stupid MUA), the error message that was given to the
1166 RCPT commands is included in the rejection of the DATA command. This is
1167 intended to be helpful for MUAs that show only the final error to their
1170 PH/42 Another patch from the Sieve maintainer.
1172 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1173 Eximstats - Fixed some broken HTML links and added missing column headers
1175 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1176 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1178 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1180 PH/43 Yet another patch from the Sieve maintainer.
1182 PH/44 I found a way to check for a TCP/IP connection going away before sending
1183 the response to the final '.' that terminates a message, but only in the
1184 case where the client has not sent further data following the '.'
1185 (unfortunately, this is allowed). However, in many cases there won't be
1186 any further data because there won't be any more messages to send. A call
1187 to select() can be used: if it shows that the input is "ready", there is
1188 either input waiting, or the socket has been closed. An attempt to read
1189 the next input character can distinguish the two cases. Previously, Exim
1190 would have sent an OK response which the client would never have see.
1191 This could lead to message repetition. This fix should cure that, at
1192 least in a lot of common cases.
1194 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1195 advertised in response to EHLO.
1201 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1202 fixed by 4.65/MH/01 (is this a record?) are fixed:
1204 (i) An empty string was always treated as zero by the numeric comparison
1205 operators. This behaviour has been restored.
1207 (ii) It is documented that the numeric comparison operators always treat
1208 their arguments as decimal numbers. This was broken in that numbers
1209 starting with 0 were being interpreted as octal.
1211 While fixing these problems I realized that there was another issue that
1212 hadn't been noticed. Values of message_size_limit (both the global option
1213 and the transport option) were treated as octal if they started with 0.
1214 The documentation was vague. These values are now always treated as
1215 decimal, and I will make that clear in the documentation.
1221 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1222 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1225 MH/01 Don't check that the operands of numeric comparison operators are
1226 integers when their expansion is in "skipping" mode (fixes bug
1227 introduced by 4.64-PH/07).
1229 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1230 child addresses, Exim now panics and dies. Previously, because the count
1231 is held in a short int, deliveries were likely to be lost. As such a
1232 large number of recipients for a single message is ridiculous
1233 (performance will be very, very poor), I have chosen to impose a limit
1234 rather than extend the field.
1240 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1241 leftover -K file (the existence of which was triggered by #402).
1242 While we were at it, introduced process PID as part of the -K
1243 filename. This should rule out race conditions when creating
1246 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1247 processing considerably. Previous code took too long for large mails,
1248 triggering a timeout which in turn triggers #401.
1250 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1251 in the DK code in transports.c. sendfile() is not really portable,
1252 hence the _LINUX specificness.
1254 TF/01 In the add_headers option to the mail command in an Exim filter,
1255 there was a bug that Exim would claim a syntax error in any
1256 header after the first one which had an odd number of characters
1259 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1260 callout verification, Exim cached a "reject" for the entire domain. This
1261 is correct for most verifications, but it is not correct for a recipient
1262 verification with use_sender or use_postmaster set, because in that case
1263 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1264 case of MAIL FROM:<> rejection from other early rejections (e.g.
1265 rejection of HELO). When verifying a recipient using a non-null MAIL
1266 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1267 Whatever the result of the callout, the value of the domain cache is
1268 left unchanged (for any other kind of callout, getting as far as trying
1269 RCPT means that the domain itself is ok).
1271 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1274 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1275 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1276 session, the daemon ignores ECONNECT errors and logs others; it now
1277 ignores EPIPE as well.
1279 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1280 (quoted-printable decoding).
1282 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1283 later the small subsequent patch to fix an introduced bug.
1285 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1287 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1289 PH/08 An error is now given if message_size_limit is specified negative.
1291 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1292 to be given (somewhat) arbitrary names.
1294 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1297 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1298 miscellaneous code fixes
1300 PH/10 Added the log_reject_target ACL modifier to specify where to log
1303 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1304 hostname. This is wrong, because it relates to the incoming message (and
1305 probably the interface on which it is arriving) and not to the outgoing
1306 callout (which could be using a different interface). This has been
1307 changed to use the value of the helo_data option from the smtp transport
1308 instead - this is what is used when a message is actually being sent. If
1309 there is no remote transport (possible with a router that sets up host
1310 addresses), $smtp_active_hostname is used.
1312 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1313 tweaks were necessary in order to get it to work (see also 21 below):
1314 (a) The code assumed that strncpy() returns a negative number on buffer
1315 overflow, which isn't the case. Replaced with Exim's string_format()
1317 (b) There were several signed/unsigned issues. I just did the minimum
1318 hacking in of casts. There is scope for a larger refactoring.
1319 (c) The code used strcasecmp() which is not a standard C function.
1320 Replaced with Exim's strcmpic() function.
1321 (d) The code set only $1; it now sets $auth1 as well.
1322 (e) A simple test gave the error "authentication client didn't specify
1323 service in request". It would seem that Dovecot has changed its
1324 interface. Fortunately there's a specification; I followed it and
1325 changed what the client sends and it appears to be working now.
1327 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1330 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1331 address A is aliased to B and C, where B exists and C does not. Without
1332 -v the output is "A verified" because verification stops after a
1333 successful redirection if more than one address is generated. However,
1334 with -v the child addresses are also verified. Exim was outputting "A
1335 failed to verify" and then showing the successful verification for C,
1336 with its parentage. It now outputs "B failed to verify", showing B's
1337 parentage before showing the successful verification of C.
1339 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1340 look up a TXT record in a specific list after matching in a combined
1343 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1344 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1345 they consult the DNS. I had assumed they would set it the way they
1346 wanted; and indeed my experiments on Linux seem to show that in some
1347 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1348 To be on the safe side, however, I have now made the interface to
1349 host_find_byname() similar to host_find_bydns(), with an argument
1350 containing the DNS resolver options. The host_find_byname() function now
1351 sets these options at its start, just as host_find_bydns() does. The smtp
1352 transport options dns_qualify_single and dns_search_parents are passed to
1353 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1354 of host_find_byname() use the default settings of RES_DEFNAMES
1355 (qualify_single) but not RES_DNSRCH (search_parents).
1357 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1358 spool_read_header() do less string testing, by means of a preliminary
1359 switch on the second character of optional "-foo" lines. (This is
1360 overdue, caused by the large number of possibilities that now exist.
1361 Originally there were few.) While I was there, I also converted the
1362 str(n)cmp tests so they don't re-test the leading "-" and the first
1363 character, in the hope this might squeeze out yet more improvement.
1365 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1366 flag allowing group syntax was set by the header_syntax check but not
1367 turned off, possible causing trouble later; (2) The flag was not being
1368 set at all for the header_verify test, causing "group"-style headers to
1369 be rejected. I have now set it in this case, and also caused header_
1370 verify to ignore an empty address taken from a group. While doing this, I
1371 came across some other cases where the code for allowing group syntax
1372 while scanning a header line wasn't quite right (mostly, not resetting
1373 the flag correctly in the right place). These bugs could have caused
1374 trouble for malformed header lines. I hope it is now all correct.
1376 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1377 with the "reply" argument non-NULL. The code, however (which originally
1378 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1379 but it didn't always do it. This confused somebody who was copying the
1380 code for some other use. I have removed all the tests.
1382 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1383 feature that was used to support insecure browsers during the U.S. crypto
1384 embargo. It requires special client support, and Exim is probably the
1385 only MTA that supported it -- and would never use it because real RSA is
1386 always available. This code has been removed, because it had the bad
1387 effect of slowing Exim down by computing (never used) parameters for the
1388 RSA_EXPORT functionality.
1390 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1391 authenticator to fail if there's a tab character in the incoming data
1392 (there should never be unless someone is messing about, as it's supposed
1393 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1394 if the connection is using TLS or if the remote IP is the same as the
1395 local IP, and the "valid-client-cert option" if a client certificate has
1398 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1399 authenticators. This can be used for authorization after authentication
1400 succeeds. (In the case of plaintext, it servers for both authentication
1403 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1404 if any retry times were supplied.
1406 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1407 connection, where there is no client IP address to check. In this
1408 situation, the verify now always succeeds.
1410 PH/25 Applied John Jetmore's -Mset patch.
1412 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1414 PH/27 In a string expansion for a processed (not raw) header when multiple
1415 headers of the same name were present, leading whitespace was being
1416 removed from all of them, but trailing whitespace was being removed only
1417 from the last one. Now trailing whitespace is removed from each header
1418 before concatenation. Completely empty headers in a concatenation (as
1419 before) are ignored.
1421 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1422 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1424 PH/29 [Removed. This was a change that I later backed out, and forgot to
1425 correct the ChangeLog entry (that I had efficiently created) before
1426 committing the later change.]
1428 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1429 address errors (4xx response to RCPT) over the same connection as other
1430 messages routed to the same hosts. Such deliveries are always "forced",
1431 so retry times are not inspected. This resulted in far too many retries
1432 for the affected addresses. The effect occurred only when there were more
1433 hosts than the hosts_max_try setting in the smtp transport when it had
1434 the 4xx errors. Those hosts that it had tried were not added to the list
1435 of hosts for which the message was waiting, so if all were tried, there
1436 was no problem. Two fixes have been applied:
1438 (i) If there are any address or message errors in an SMTP delivery, none
1439 of the hosts (tried or untried) are now added to the list of hosts
1440 for which the message is waiting, so the message should not be a
1441 candidate for sending over the same connection that was used for a
1442 successful delivery of some other message. This seems entirely
1443 reasonable: after all the message is NOT "waiting for some host".
1444 This is so "obvious" that I'm not sure why it wasn't done
1445 previously. Hope I haven't missed anything, but it can't do any
1446 harm, as the worst effect is to miss an optimization.
1448 (ii) If, despite (i), such a delivery is accidentally attempted, the
1449 routing retry time is respected, so at least it doesn't keep
1450 hammering the server.
1452 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1453 in ${readsocket because some servers need this prod.
1455 PH/32 Added some extra debug output when updating a wait-xxx database.
1457 PH/33 The hint "could be header name not terminated by colon", which has been
1458 given for certain expansion errors for a long time, was not being given
1459 for the ${if def:h_colon_omitted{... case.
1461 PH/34 The spec says: "With one important exception, whenever a domain list is
1462 being scanned, $domain contains the subject domain." There was at least
1463 one case where this was not true.
1465 PH/35 The error "getsockname() failed: connection reset by peer" was being
1466 written to the panic log as well as the main log, but it isn't really
1467 panic-worthy as it just means the connection died rather early on. I have
1468 removed the panic log writing for the ECONNRESET error when getsockname()
1471 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1472 runs only) independently of the message's sender address. This meant
1473 that, if the 4xx error was in fact related to the sender, a different
1474 message to the same recipient with a different sender could confuse
1475 things. In particualar, this can happen when sending to a greylisting
1476 server, but other circumstances could also provoke similar problems.
1477 I have changed the default so that the retry time for these errors is now
1478 based a combination of the sender and recipient addresses. This change
1479 can be overridden by setting address_retry_include_sender=false in the
1482 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1483 remote server are returned as part of bounce messages. This was not
1484 happening for LMTP over a pipe (the lmtp transport), but now it is the
1485 same for both kinds of LMTP.
1487 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1488 in header lines that were in fact CNAMEs. This is no longer the case.
1490 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1491 and queue runs started by the daemon processed all messages. This has
1492 been fixed so that -R and -S can now usefully be given with -q<time>.
1494 PH/40 Import PCRE release 6.7 (fixes some bugs).
1496 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1498 PH/42 Give an error if -q is specified more than once.
1500 PH/43 Renamed the variables $interface_address and $interface_port as
1501 $received_ip_address and $received_port, to make it clear that these
1502 values apply to message reception, and not to the outgoing interface when
1503 a message is delivered. (The old names remain recognized, of course.)
1505 PH/44 There was no timeout on the connect() call when using a Unix domain
1506 socket in the ${readsocket expansion. There now is.
1508 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1509 be meaningful with "accept".
1511 SC/01 Eximstats V1.43
1512 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1514 SC/02 Eximstats V1.44
1515 Use a glob alias rather than an array ref in the generated
1516 parser. This improves both readability and performance.
1518 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1519 Collect SpamAssassin and rejection statistics.
1520 Don't display local sender or destination tables unless
1521 there is data to show.
1522 Added average volumes into the top table text output.
1524 SC/04 Eximstats V1.46
1525 Collect data on the number of addresses (recipients)
1526 as well as the number of messages.
1528 SC/05 Eximstats V1.47
1529 Added 'Message too big' to the list of mail rejection
1530 reasons (thanks to Marco Gaiarin).
1532 SC/06 Eximstats V1.48
1533 Mainlog lines which have GMT offsets and are too short to
1534 have a flag are now skipped.
1536 SC/07 Eximstats V1.49 (Alain Williams)
1537 Added the -emptyok flag.
1539 SC/08 Eximstats V1.50
1540 Fixes for obtaining the IP address from reject messages.
1542 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1543 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1544 whitesspace changes from 4.64-PH/27
1546 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1549 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1550 are found, allow negative numbers in numeric criteria)
1552 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1554 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1555 to match changes made in 4.64-PH/43
1557 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1559 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1560 "domain" is actually an IP address, and force a failure. This locks out
1561 those revolvers/nameservers that support "A-for-A" lookups, in
1562 contravention of the specifications.
1564 PH/48 When a host name was looked up from an IP address, and the subsequent
1565 forward lookup of the name timed out, the host name was left in
1566 $sender_host_name, contrary to the specification.
1568 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1569 restricted to single-key lookups, Exim was not diagnosing an error if
1570 * or *@ was used with a query-style lookup.
1572 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1574 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1575 long ago, but noone interested enough thought of it. Let's just say that
1576 the "1.1" means that there are some new functions that weren't there at
1577 some point in the past.
1579 PH/51 Error processing for expansion failure of helo_data from an smtp
1580 transport during callout processing was broken.
1582 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1583 tested/used via the -bh/-bhc/-bs options.
1585 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1586 bug, fixed in subsequent PCRE releases).
1588 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1589 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1591 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1597 SC/01 Use a glob alias rather than an array ref in eximstats generated
1598 parser. This improves both readability and performance.
1600 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1601 Don't display local sender or destination tables in eximstats unless
1602 there is data to show.
1603 Added average volumes into the eximstats top table text output.
1605 SC/03 Collect data on the number of addresses (recipients) as well
1606 as the number of messages in eximstats.
1608 TF/01 Correct an error in the documentation for the redirect router. Exim
1609 does (usually) call initgroups() when daemonizing.
1611 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1612 with consistent privilege compared to when running as a daemon.
1614 TF/03 Note in the spec that $authenticated_id is not set for local
1615 submissions from trusted users.
1617 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1618 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1620 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1621 by adding some example configuration directives to the default
1622 configuration file. A little bit of work is required to uncomment the
1623 directives and define how usernames and passwords are checked, but
1624 there is now a framework to start from.
1626 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1627 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1628 without this. I don't know how relevant this is to other LDAP libraries.
1630 PH/02 Add the verb name to the "unknown ACL verb" error.
1632 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1634 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1636 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1637 directory not expanded when it should be if an expanded home directory
1638 was set for the address (which is overridden by the transport).
1640 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1643 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1644 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1645 because it is too late at that time, and has no effect.
1647 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1648 security issue with \' (bugzilla #107). I could not use the
1649 PQescapeStringConn() function, because it needs a PGconn value as one of
1652 PH/08 When testing addresses using -bt, indicate those final addresses that
1653 are duplicates that would not cause an additional delivery. At least one
1654 person was confused, thinking that -bt output corresponded to deliveries.
1655 (Suppressing duplicates isn't a good idea as you lose the information
1656 about possibly different redirections that led to the duplicates.)
1658 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1659 systems where poll() doesn't work, in particular OS X.
1661 PH/10 Added more information to debugging output for retry time not reached.
1663 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1664 operations in malware.c.
1666 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1669 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1670 syslog_duplication set false, log lines that would normally be written
1671 both the the main log and to the reject log were not written to syslog at
1674 PH/14 In the default configuration, change the use of "message" in ACL warn
1675 statements to "add_header".
1677 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1678 not followed by a command (e.g. "seen endif").
1680 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1681 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1684 PH/17 Added extra conditions to the default value of delay_warning_condition
1688 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1689 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1690 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1693 The Auto-Submitted: and various List- headers are standardised, whereas I
1694 don't think Precedence: ever was.
1696 PH/18 Refactored debugging code in route_finduser() to show more information,
1697 in particular, the error code if getpwnam() issues one.
1699 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1700 This is apparently needed in addition to the PH/07 change above to avoid
1701 any possible encoding problems.
1703 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1704 but not after initializing Perl.
1706 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1707 output them only if debugging. By default they are written stderr,
1708 apparently, which is not desirable.
1710 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1713 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1716 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1718 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1719 authenticated or an ident call has been made. Suppress the default
1720 values for $authenticated_id and $authenticated_sender (but permit -oMai
1721 and -oMas) when testing with -bh.
1723 PH/24 Re-jigged the order of the tests in the default configuration so that the
1724 tests for valid domains and recipients precede the DNS black list and CSA
1725 tests, on the grounds that those ones are more expensive.
1727 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1728 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1729 This bug exists in every version of Exim that I still have, right back to
1732 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1733 However, an attempt to turn on case-sensitivity in a regex key by
1734 including (?-i) didn't work because the subject string was already
1735 lowercased, and the effects were non-intuitive. It turns out that a
1736 one-line patch can be used to allow (?-i) to work as expected.
1742 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1743 other effects) broke the use of negated acl sub-conditions.
1745 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1748 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1749 "Deny" causes Exim to reject the incoming connection with a 554 error.
1750 Unfortunately, if there is a major crisis, such as a disk failure,
1751 tcp-wrappers gives "deny", whereas what one would like would be some
1752 kind of temporary error. A kludge has been added to help with this.
1753 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1754 554 error is used if errno is still zero or contains ENOENT (which occurs
1755 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1758 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1760 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1761 errors. Otherwise a message that provokes a temporary error (when other
1762 messages do not) can cause a whole host to time out.
1764 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1765 addresses were routed directly to files or pipes from a redirect router.
1766 File deliveries just didn't batch; pipe deliveries might have suffered
1769 PH/06 A failure to get a lock for a hints database would erroneously always say
1770 "Failed to get write lock", even when it was really a read lock.
1772 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1773 of 0600. This has been changed to use the value of the lockfile_mode
1774 option (which defaults to 0600).
1776 PH/08 Applied small patch from the Sieve maintainer.
1778 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1779 folder from quota calculations, a direct delivery into this folder messed
1780 up the contents of the maildirsize file. This was because the regex was
1781 used only to exclude .Trash (or whatever) when the size of the mailbox
1782 was calculated. There was no check that a delivery was happening into an
1783 excluded directory. This bug has been fixed by ignoring all quota
1784 processing for deliveries into excluded directories.
1786 PH/10 Added the maildirfolder_create_regex option to appendfile.
1792 PH/01 The code for finding all the local interface addresses on a FreeBSD
1793 system running IPv6 was broken. This may well have applied to all BSD
1794 systems, as well as to others that have similar system calls. The broken
1795 code found IPv4 interfaces correctly, but gave incorrect values for the
1796 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1797 that it would not match correctly against @[] and not recognize the IPv6
1800 PH/02 The ipliteral router was not recognizing addresses of the form user@
1801 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1803 PH/03 Added disable_ipv6.
1805 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1806 decoded form, because it is most often used to construct To: headers
1807 lines in autoreplies, and the decoded form may well be syntactically
1808 invalid. However, $reply_address has leading white space removed, and all
1809 newlines turned into spaces so that the autoreply transport does not
1812 PH/05 If group was specified without a user on a router, and no group or user
1813 was specified on a transport, the group from the router was ignored.
1815 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1816 for visible compile-time settings that can be used to change these
1817 numbers, for those that want even more. Backwards compatibility with old
1818 spool files has been maintained. However, going back to a previous Exim
1819 release will lost any variables that are in spool files.
1821 PH/07 Two small changes when running in the test harness: increase delay when
1822 passing a TCP/IP connection to a new process, in case the original
1823 process has to generate a bounce, and remove special handling of
1824 127.0.0.2 (sic), which is no longer necessary.
1826 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1827 be the same on different OS.
1829 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1832 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1833 whether --show-vars was specified or not
1835 JJ/02 exipick: Added support for new ACL variable spool format introduced
1838 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1839 syntactically invalid From: or Reply-to: line, and a filter used this to
1840 generate an autoreply, and therefore failed to obtain an address for the
1841 autoreply, Exim could try to deliver to a non-existent relative file
1842 name, causing unrelated and misleading errors. What now happens is that
1843 it logs this as a hard delivery error, but does not attempt to create a
1846 PH/11 The exinext utility has a -C option for testing purposes, but although
1847 the given file was scanned by exinext itself; it wasn't being passed on
1848 when Exim was called.
1850 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1851 an end-of-file indication when reading a command response.
1853 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1854 compiled. In many other places in Exim, IPv6 addresses are always
1855 recognized, so I have changed this. It also means that IPv4 domain
1856 literals of the form [IPV4:n.n.n.n] are now always recognized.
1858 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1859 used if the router is not running as root, for example, when verifying at
1860 ACL time, or when using -bh. The debugging output from this situation was
1861 non-existent - all you got was a failure to exec. I have made two
1864 (a) Failures to set uid/gid, the current directory, or a process leader
1865 in a subprocess such as that created by queryprogram now generate
1866 suitable debugging ouput when -d is set.
1868 (b) The queryprogram router detects when it is not running as root,
1869 outputs suitable debugging information if -d is set, and then runs
1870 the subprocess without attempting to change uid/gid.
1872 PH/15 Minor change to Makefile for building test_host (undocumented testing
1875 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1876 additional section of a DNS packet that returns MX or SRV records.
1877 Instead, it always explicitly searches for A/AAAA records. This avoids
1878 major problems that occur when a DNS server includes only records of one
1879 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1880 fixed another bug: if SRV records were looked up and the corresponding
1881 address records were *not* found in the additional section, the port
1882 values from the SRV records were lost.
1884 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1885 using the correct key (the original address) when searching the retry
1886 rules in order to find which one to use for generating the retry hint.
1888 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1889 adding the default one. Similarly, if it contains a Reply-To: header, the
1890 errors_reply_to option, if set, is not used.
1892 PH/19 When calculating a retry time, Exim used to measure the "time since
1893 failure" by looking at the "first failed" field in the retry record. Now
1894 it does not use this if it is later than than the arrival time of the
1895 message. Instead it uses the arrival time. This makes for better
1896 behaviour in cases where some deliveries succeed, thus re-setting the
1897 "first failed" field. An example is a quota failure for a huge message
1898 when small messages continue to be delivered. Without this change, the
1899 "time since failure" will always be short, possible causing more frequent
1900 delivery attempts for the huge message than are intended.
1901 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1903 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1904 $1, $2, $3) because the numerical variables can be reset during some
1905 expansion items (e.g. "match"), thereby losing the authentication data.
1907 PH/21 Make -bV show the size of off_t variables so that the test suite can
1908 decide whether to run tests for quotas > 2G.
1910 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1911 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1912 filecount value is greater than 2G or if a quota value is greater than 2G
1913 on a system where the size of off_t is not greater than 4, a panic error
1916 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1917 never match. The debug and -bh output now contains an explicit error
1918 message indicating a malformed IPv4 address or mask.
1920 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1921 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1922 PH/23 above applies.
1924 PH/25 Do not write to syslog when running in the test harness. The only
1925 occasion when this arises is a failure to open the main or panic logs
1926 (for which there is an explicit test).
1928 PH/26 Added the /no_tell option to "control=freeze".
1930 PH/27 If a host name lookup failed very early in a connection, for example, if
1931 the IP address matched host_lookup and the reverse lookup yielded a name
1932 that did not have a forward lookup, an error message of the form "no IP
1933 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1934 could be logged. Now it outputs the IP address instead of "NULL".
1936 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1937 allows the sender and the authenticated sender to be set when
1938 submitting a message from within Exim. Since child_open_exim() is
1939 documented for local_scan(), the new function should be too.
1941 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1942 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1943 results in an empty string is now treated as unset.
1945 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1947 PH/31 Added sender_verify_fail logging option.
1949 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1950 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1951 tidied the source and removed it altogether.
1953 PH/33 When a queue run was abandoned because the load average was too high, a
1954 log line was always written; now it is written only if the queue_run log
1955 selector is set. In addition, the log line for abandonment now contains
1956 information about the queue run such as the pid. This is always present
1957 in "start" and "stop" lines but was omitted from the "abandon" line.
1959 PH/34 Omit spaces between a header name and the colon in the error message that
1960 is given when verify = headers_syntax fails (if there are lots of them,
1961 the message gets confusing).
1963 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1964 names, as there are now some PTR records that contain slashes. This check
1965 is only to protect against broken name servers that fall over on strange
1966 characters, so the fact that it applies to all lookups doesn't matter.
1968 PH/36 Now that the new test suite is complete, we can remove some of the
1969 special code in Exim that was needed for the old test suite. For example,
1970 sorting DNS records because real resolvers return them in an arbitrary
1971 order. The new test suite's fake resolver always returns records in the
1974 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1975 bounces) except when queue_only is set, to avoid logging races between
1976 the different processes.
1978 PH/38 Panic-die if .include specifies a non-absolute path.
1980 PH/39 A tweak to the "H" retry rule from its user.
1982 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1983 a label. They prevented compilation on older perls.
1985 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1986 a warning to be raised on newish perls.
1988 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1989 on queue. Changes to match documented behaviour of showing count of
1990 messages matching specified criteria.
1992 PH/40 Changed the default ident timeout from 30s to 5s.
1994 PH/41 Added support for the use of login_cap features, on those BSD systems
1995 that have them, for controlling the resources used by pipe deliveries.
1997 PH/42 The content-scanning code uses fopen() to create files in which to put
1998 message data. Previously it was not paying any attention to the mode of
1999 the files. Exim runs with umask(0) because the rest of the code creates
2000 files with open(), and sets the required mode explicitly. Thus, these
2001 files were ending up world-writeable. This was not a big issue, because,
2002 being within the spool directory, they were not world-accessible. I have
2003 created a function called modefopen, which takes an additional mode
2004 argument. It sets umask(777), creates the file, chmods it to the required
2005 mode, then resets the umask. All the relevant calls to fopen() in the
2006 content scanning code have been changed to use this function.
2008 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2009 to 24 hours. This avoids potential overflow problems when processing G
2010 and H retry rules. I suspect nobody ever tinkers with this value.
2012 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2014 PH/45 When the plaintext authenticator is running as a client, the server's
2015 challenges are checked to ensure they are valid base64 strings. By
2016 default, the authentication attempt is cancelled if an invalid string is
2017 received. Setting client_ignore_invalid_base64 true ignores these errors.
2018 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2019 they are received. Thus, the responses can be made to depend on the
2020 challenges. If an invalid string is ignored, an empty string is placed in
2023 PH/46 Messages that are created by the autoreply transport now contains a
2024 References: header, in accordance with RFCs 2822 and 3834.
2026 PH/47 Added authenticated_sender_force to the smtp transport.
2028 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2030 PH/49 Installed latest patch from the Sieve maintainer.
2032 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2033 was also set, the appendfile transport was unnecessarily scanning a
2034 directory of message files (e.g. for maildir delivery) to find the count
2035 of files (along with the size), even though it did not need this
2036 information. It now does the scan only if it needs to find either the
2037 size of the count of files.
2039 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2041 PH/52 Two bugs concerned with error handling when the smtp transport is
2044 (i) Exim was not creating retry information for temporary errors given
2045 for individual recipients after the DATA command when the smtp transport
2046 was used in LMTP mode. This meant that they could be retried too
2047 frequently, and not timed out correctly.
2049 (ii) Exim was setting the flag that allows error details to be returned
2050 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2051 recipients that were returned after the DATA command.
2053 PH/53 This is related to PH/52, but is more general: for any failing address,
2054 when detailed error information was permitted to be returned to the
2055 sender, but the error was temporary, then after the final timeout, only
2056 "retry timeout exceeded" was returned. Now it returns the full error as
2057 well as "retry timeout exceeded".
2059 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2060 do this, and (what is worse) MTAs that accept it.
2062 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2063 will now be deprecated.
2065 PH/56 New os.c-cygwin from the Cygwin maintainer.
2067 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2068 formats (previously only available in exim formats via -bpr, -bpru,
2069 and -bpra. Now also available in native and exiqgrep formats)
2071 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2072 with very large, slow to parse queues
2074 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2076 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2078 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2079 responses to authentication challenges, though it was showing the
2080 challenges; (ii) I've removed the CR characters from the debug output for
2083 PH/58 Allow for the insertion of a newline as well as a space when a string
2084 is turned into more than one encoded-word during RFC 2047 encoding. The
2085 Sieve code now uses this.
2087 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2088 data_4xx, lost_connection, tls_required.
2090 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2091 message was being sent as an SMTP response.
2093 PH/61 Add -l and -k options to exicyclog.
2095 PH/62 When verifying, if an address was redirected to one new address, so that
2096 verification continued, and the new address failed or deferred after
2097 having set something in $address_data, the value of $address_data was not
2098 passed back to the ACL. This was different to the case when no
2099 redirection occurred. The value is now passed back in both cases.
2101 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2102 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2103 use login_cap.h, so on its own it isn't the distinguishing feature. The
2104 new name refers directly to the setclassresources() function.
2106 PH/65 Added configuration files for NetBSD3.
2108 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2110 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2111 is preferred over IPv4.
2113 PH/68 The bounce_return_message and bounce_return_body options were not being
2114 honoured for bounces generated during the reception of non-SMTP messages.
2115 In particular, this applied to messages rejected by the ACL. This bug has
2116 been fixed. However, if bounce_return_message is true and bounce_return_
2117 body is false, the headers that are returned for a non-SMTP message
2118 include only those that have been read before the error was detected.
2119 (In the case of an ACL rejection, they have all been read.)
2121 PH/69 The HTML version of the specification is now built in a directory called
2122 spec_html instead of spec.html, because the latter looks like a path with
2123 a MIME-type, and this confuses some software.
2125 PH/70 Catch two compiler warnings in sieve.c.
2127 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2128 function verify_get_ident() calls ip_connect() to connect a socket, but
2129 if the "connect()" function timed out, ip_connect() used to close the
2130 socket. However, verify_get_ident() also closes the socket later, and in
2131 between Exim writes to the log, which may get opened at this point. When
2132 the socket was closed in ip_connect(), the log could get the same file
2133 descriptor number as the socket. This naturally causes chaos. The fix is
2134 not to close the socket in ip_connect(); the socket should be closed by
2135 the function that creates it. There was only one place in the code where
2136 this was missing, in the iplookup router, which I don't think anybody now
2137 uses, but I've fixed it anyway.
2139 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2140 well as to direct DNS lookups. Otherwise the handling of names in host
2141 lists is inconsistent and therefore confusing.
2147 PH/01 Two changes to the default runtime configuration:
2149 (1) Move the checks for relay_from_hosts and authenticated clients from
2150 after to before the (commented out) DNS black list checks.
2152 (2) Add control=submission to the relay_from_hosts and authenticated
2153 clients checks, on the grounds that messages accepted by these
2154 statements are most likely to be submissions.
2156 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2158 (1) Generate an error if the third argument for the ${prvs expansion is
2161 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2164 (3) Reset the variables that are obtained from the first argument of
2165 ${prvscheck and used in the second argument before leaving the code,
2166 because their memory is reclaimed, so using them afterwards may do
2169 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2170 one (it's much easier than Tom thought :-).
2172 (5) Because of (4), we can now allow for the use of $prvscheck_result
2173 inside the third argument.
2175 PH/03 For some reason, the default setting of PATH when running a command from
2176 a pipe transport was just "/usr/bin". I have changed it to
2179 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2180 anything to be listed in the output from -bV.
2182 PH/05 When a filter generated an autoreply, the entire To: header line was
2183 quoted in the delivery log line, like this:
2185 => >A.N.Other <ano@some.domain> <original@ddress> ...
2187 This has been changed so that it extracts the operative address. There
2188 may be more than one such address. If so, they are comma-separated, like
2191 => >ano@some.domain,ona@other.domain <original@ddress> ...
2193 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2194 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2195 not being looked up in the rDNS to get a host name, Exim was showing the
2196 IP address twice in Received: lines, even though the IP addresses were
2197 identical. For example:
2199 Received: from [1.2.3.4] (helo=[1.2.3.4])
2201 However, if the real host name was known, it was omitting the HELO data
2202 if it matched the actual IP address. This has been tidied up so that it
2203 doesn't show the same IP address twice.
2205 PH/07 When both +timestamp and +memory debugging was on, the value given by
2206 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2207 called by the debug printing, thereby overwriting the timestamp buffer.
2208 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2210 PH/08 When the original message was included in an autoreply transport, it
2211 always said "this is a copy of the message, including all the headers",
2212 even if body_only or headers_only was set. It now gives an appropriate
2215 PH/09 Applied a patch from the Sieve maintainer which:
2217 o fixes some comments
2218 o adds the (disabled) notify extension core
2219 o adds some debug output for the result of if/elsif tests
2220 o points to the current vacation draft in the documentation
2221 and documents the missing references header update
2225 o fixes a bug in processing the envelope test (when testing
2226 multiple envelope elements, the last element determinted the
2229 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2230 Electronic Mail") by including:
2232 Auto-submitted: auto-generated
2234 in the messages that it generates (bounce messages and others, such as
2235 warnings). In the case of bounce messages for non-SMTP mesages, there was
2236 also a typo: it was using "Auto_submitted" (underscore instead of
2237 hyphen). Since every message generated by Exim is necessarily in response
2238 to another message, thes have all been changed to:
2240 Auto-Submitted: auto-replied
2242 in accordance with these statements in the RFC:
2244 The auto-replied keyword:
2246 - SHOULD be used on messages sent in direct response to another
2247 message by an automatic process,
2249 - MUST NOT be used on manually-generated messages,
2251 - MAY be used on Delivery Status Notifications (DSNs) and Message
2252 Disposition Notifications (MDNs),
2254 - MUST NOT be used on messages generated by automatic or periodic
2255 processes, except for messages which are automatic responses to
2258 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2259 to the default Received: header definition.
2261 PH/12 Added log selector acl_warn_skipped (default on).
2263 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2264 variables because (a) they are in the wrong storage pool and (b) even if
2265 they were copied, it wouldn't work properly because of the caching.
2267 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2268 checking when decoding. Apparently there are clients that generate
2269 overlong encoded strings. Why am I not surprised?
2271 PH/15 If the first argument of "${if match_address" was not empty, but did not
2272 contain an "@" character, Exim crashed. Now it writes a panic log message
2273 and treats the condition as false.
2275 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2277 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2278 extension "envelope-auth". The code is finished and in agreement with
2279 other implementations, but there is no documentation so far and in fact,
2280 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2281 not changing the active code.
2283 Print executed "if" and "elsif" statements when debugging is used. This
2284 helps a great deal to understand what a filter does.
2286 Document more things not specified clearly in RFC3028. I had all this
2287 sorted out, when out of a sudden new issues came to my mind. Oops."
2289 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2292 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2293 canonical form (as documented). However, after a host name lookup from
2294 the IP address, check_host() was doing a simple string comparison with
2295 addresses acquired from the DNS when checking that the found name did
2296 have the original IP as one of its addresses. Since any found IPv6
2297 addresses are likely to be in abbreviated form, the comparison could
2298 fail. Luckily, there already exists a function for doing the comparison
2299 by converting both addresses to binary, so now that is used instead of
2300 the text comparison.
2302 PH/20 There was another similar case to PH/19, when a complete host name was
2303 given in a host list; looking up its IP address could give an abbreviated
2304 form, whereas the current host's name might or might not be abbreviated.
2305 The same fix has been applied.
2311 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2312 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2315 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2316 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2318 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2320 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2321 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2322 Also, for "Auto-Submitted", treat anything other than "no" as
2323 non-personal, in accordance with RFC 3834. (Previously it treated
2324 anything starting "auto-" as non-personal.)
2326 TF/01 The control=submission/name=... option had a problem with syntax
2327 errors if the name included a slash character. The /name= option
2328 now slurps the rest of the string, so it can include any characters
2329 but it must come last in the list of options (after /sender_retain
2332 PH/05 Some modifications to the interface to the fake nameserver for the new
2340 TK/01 Added the "success_on_redirect" address verification option. See
2341 NewStuff for rationale and an example.
2343 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2345 PH/02 Patch to exigrep to allow it to work on syslog lines.
2347 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2348 fread() to skip over the body file's header line, because in Cygwin the
2349 header line is locked and is inaccessible.
2351 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2352 co-exist for some time) to make it clear that it is the Exim ID that is
2353 referenced, not the Message-ID: header line.
2355 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2356 string_format() function, because snprintf() does not exist on all
2359 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2360 ${stat: expansion item.
2362 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2363 protocol synchronization error", to keep the pedants happy.
2365 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2366 well as for IRIX systems, when gcc is being used. See the host.c source
2369 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2371 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2374 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2375 transport and to the smtp transport in LMTP mode.
2377 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2379 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2380 run a filter in a subprocess. This could lead to confusion in subsequent
2381 lookups in the parent process. There should also be a search_tidyup() at
2382 the end of the subprocess.
2384 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2385 only if the host matched helo_try_verify_hosts, which caused the
2386 verification to occur when the EHLO/HELO command was issued. The ACL just
2387 tested the remembered result. Now, if a previous verification attempt has
2388 not happened, "verify = helo" does it there and then.
2390 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2392 TK/03 Fix log output including CR from clamd.
2394 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2395 exist provoked a memory error which could cause a segfault.
2397 PH/15 Installed PCRE 6.2
2399 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2401 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2402 of the problem. Specifically, suggested +O2 rather than +O1 for the
2405 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2407 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2408 record, the port was not correctly logged when the outgoing_port log
2409 selector was set (it logged the transort's default port).
2411 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2412 fallback_hosts, and "hosts" in the smtp transport.
2414 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2415 host errors such as "Connection refused".
2417 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2418 authentication with radiusclient 0.4.9:
2420 - Error returned from rc_read_config was caught wrongly
2421 - Username/password not passed on to radius server due to wrong length.
2423 The presumption is that some radiusclient API changes for 4.51/PH/17
2424 were not taken care of correctly. The code is still untested by me (my
2425 Linux distribution still has 0.3.2 of radiusclient), but it was
2426 contributed by a Radius user.
2428 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2429 expanding the "port" option of the smtp transport.
2431 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2432 while reading a MIME header. Thanks to Tom Hughes for a patch.
2434 PH/24 Include config.h inside local_scan.h so that configuration settings are
2437 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2438 that in an ACL for RCPT (for example), you can examine exactly what was
2441 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2442 commands, but it was not correctly comparing the address with the actual
2443 client host address. Thus, it would show the EHLO address in Received:
2444 header lines when this was not necessary.
2446 PH/27 Added the % operator to ${eval:}.
2448 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2449 it should be ignoring failures (because with -C, for example, it has lost
2450 privilege). It wasn't ignoring creation failures other than "already
2453 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2456 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2457 because an input line was too long, either on its own, or by virtue of
2458 too many continuations, the temporary file was not being removed, and the
2459 return code was incorrect.
2461 PH/31 Missing "BOOL" in function definition in filtertest.c.
2463 PH/32 Applied Sieve patches from the maintainer.
2465 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2467 PH/33 Added "verify = not_blind".
2469 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2470 Local/Makefile (with some defaults set). These are used in built scripts
2471 such as exicyclog, but they have never been used in the exim_install
2472 script (though there are many overriding facilities there). I have
2473 arranged that the exim_install script now takes note of these two
2476 PH/35 Installed configuration files for Dragonfly.
2478 PH/36 When a locally submitted message by a trusted user did not contain a
2479 From: header, and the sender address was obtained from -f or from an SMTP
2480 MAIL command, and the trusted user did not use -F to supply a sender
2481 name, $originator_name was incorrectly used when constructing a From:
2482 header. Furthermore, $originator_name was used for submission mode
2483 messages from external hosts without From: headers in a similar way,
2484 which is clearly wrong.
2486 PH/37 Added control=suppress_local_fixups.
2488 PH/38 When log_selector = +received_sender was set, and the addition of the
2489 sender made the log line's construction buffer exactly full, or one byte
2490 less than full, an overflow happened when the terminating "\n" was
2493 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2494 when the result of a list match is failure because a DNS lookup failed.
2496 PH/40 RM_COMMAND is now used in the building process.
2498 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2499 the "build-* directories that it finds.
2501 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2502 address in a domain literal was a prefix of an interface address.
2504 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2505 when verifying a sender address, unless rewrite_headers is false.
2507 PH/44 Wrote a long comment about why errors_to addresses are verified as
2508 recipients, not senders.
2510 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2511 the ratelimit ACL was added.
2513 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2515 PH/46 Added extra information about PostgreSQL errors to the error string.
2517 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2518 avoiding the need to install special zones in a real server. This is
2519 backwards compatible; if it can't find the fake resolver, it drops back.
2520 Thus, both old and new test suites can be run.
2522 TF/02 Added util/ratelimit.pl
2524 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2525 clock is set back in time.
2527 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2528 Candler <B.Candler@pobox.com>.
2530 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2531 OK for addresses that are the result of redirections.
2533 PH/48 A number of further additions for the benefit of the new test suite,
2534 including a fake gethostbyname() that interfaces to the fake DNS resolver
2537 TF/06 The fix for widen_domains has also been applied to qualify_single and
2538 search_parents which are the other dnslookup options that can cause
2541 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2544 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2546 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2547 service" if it encountered an MX record with an empty target hostname.
2548 The message is now "an MX or SRV record indicated no SMTP service".
2550 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2551 if the DNS of the sending site is misconfigured. This is quite a
2552 common situation. This change restores the behaviour of treating a
2553 helo verification defer as a failure.
2555 PH/51 If self=fail was set on a router, the bounce message did not include the
2556 actual error message.
2562 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2564 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2565 command itself ended in error, the underlying message about the transport
2566 filter timeout was being overwritten with the pipe command error. Now the
2567 underlying error message should be appended to the second error message.
2569 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2571 PH/02 Reduce the amount of output that "make" produces by default. Full output
2572 can still be requested.
2574 PH/03 The warning log line about a condition test deferring for a "warn" verb
2575 was being output only once per connection, rather than after each
2576 occurrence (because it was using the same function as for successful
2577 "warn" verbs). This seems wrong, so I have changed it.
2579 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2580 it should not have, which might have caused a crash in the right
2581 circumstances, but probably never did.
2583 PH/04 Installed a modified version of Tony Finch's patch to make submission
2584 mode fix the return path as well as the Sender: header line, and to
2585 add a /name= option so that you can make the user's friendly name appear
2588 TF/03 Added the control = fakedefer ACL modifier.
2590 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2591 Mark Lowes for thorough testing.
2593 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2595 TK/03 Merged latest SRS patch from Miles Wilton.
2597 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2598 with the definition in sysexits.h (which is #included earlier).
2599 Fortunately, Exim does not actually use EX_OK. The code used to try to
2600 preserve the sysexits.h value, by assumimg that macro definitions were
2601 scanned for macro replacements. I have been disabused of this notion,
2602 so now the code just undefines EX_OK before #including unistd.h.
2604 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2605 in the smtp transport. When a block could not be written in a single
2606 write() function, the timeout was being re-applied to each part-write.
2607 This seems wrong - if the receiver was accepting one byte at a time it
2608 would take for ever. The timeout is now adjusted when this happens. It
2609 doesn't have to be particularly precise.
2611 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2612 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2614 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2615 without a domain if the check to <postmaster@domain> fails.
2617 SC/01 Eximstats: added -xls and the ability to specify output files
2618 (patch written by Frank Heydlauf).
2620 SC/02 Eximstats: use FileHandles for outputing results.
2622 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2624 SC/04 Eximstats: fixed display of large numbers with -nvr option
2626 SC/05 Eximstats: fixed merging of reports with empty tables.
2628 SC/06 Eximstats: added the -include_original_destination flag
2630 SC/07 Eximstats: removed tabs and trailing whitespace.
2632 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2634 TK/06 MBOX spool code: Add real "From " MBOX separator line
2635 so the .eml file is really in mbox format (even though
2636 most programs do not really care). Patch from Alex Miller.
2638 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2639 The latter is generated from $received_to and is only set if the
2640 message has one envelope recipient. SA can use these headers,
2641 obviously out-of-the-box. Patch from Alex Miller.
2643 PH/08 The ${def test on a variable was returning false if the variable's
2644 value was "0", contrary to what the specification has always said!
2645 The result should be true unless the variable is empty.
2647 PH/09 The syntax error of a character other than { following "${if
2648 def:variable_name" (after optional whitespace) was not being diagnosed.
2649 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2650 accidental colon was present, for example, could give incorrect results.
2652 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2653 result is used (not including appendfile, where other changes are about
2656 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2657 This involved changing a lot of size variables from int to off_t. It
2658 should work with maildirs and everything.
2660 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2661 spamd dying while we are connected to it.
2663 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2666 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2667 function for BDB 4.3.
2669 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2671 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2672 the sources for PCRE have been re-arranged and more files are now
2675 PH/15 The code I had for printing potentially long long variables in PH/11
2676 above was not the best (it lost precision). The length of off_t variables
2677 is now inspected at build time, and an appropriate printing format (%ld
2678 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2679 to be "long long int" or "long int". This is needed for the internal
2680 formatting function string_vformat().
2682 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2683 the configuration file to be ":syslog", then the script "guesses" where
2684 the logs files are, rather than using the compiled in default. In our
2685 case the guess is not the same as the compiled default, so the script
2686 suddenly stopped working when I started to use syslog. The patch checks
2687 to see if log_file_path is "". If so, it attempts to read it from exim
2688 with no configuration file to get the compiled in version, before it
2689 falls back to the previous guessing code."
2691 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2692 implementing BATV in an Exim configuration. See NewStuff for the gory
2695 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2696 Makefile that are specific to HP-UX.
2698 PH/18 If the "use_postmaster" option was set for a recipient callout together
2699 with the "random" option, the postmaster address was used as the MAIL
2700 FROM address for the random test, but not for the subsequent recipient
2701 test. It is now used for both.
2703 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2704 patch removes a few documentation additions to RFC 3028, because the
2705 latest draft now contains them. It adds the new en;ascii-case comparator
2706 and a new error check for 8bit text in MIME parts. Comparator and
2707 require names are now matched exactly. I enabled the subaddress
2708 extension, but it is not well tested yet (read: it works for me)."
2710 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2711 rework some of the code of TK/09 above to avoid the hardwired use of
2712 "%lld" and "long long". Replaced the call to snprintf() with a call to
2715 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2716 records point to non-existent hosts", "retry timeout exceeded", and
2717 "retry time not reached for any host after a long failure period".
2719 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2720 experimental DomainKeys support:
2722 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2723 (2) On an error such as an illegally used "control", the wrong name for
2724 the control was given.
2726 These problems did NOT occur unless DomainKeys support was compiled.
2728 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2730 PH/24 Added ${if match_ip condition.
2732 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2733 databases so that it will be absolutely obvious if a crash occurs in the
2734 DB library. This is a regular occurrence (often caused by mis-matched
2737 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2738 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2739 picked up on a user's system that detects such things. There doesn't seem
2740 to be a gcc warning option for this - only an attribute that has to be
2741 put on the function's prototype. It seems that in Fedora Core 4 they have
2742 set this on a number of new functions. No doubt there will be more in due
2745 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2746 specify a transport. However, if an address that was verified by such a
2747 router was the subject of a callout, Exim crashed because it tried to
2748 read the rcpt_include_affixes from the non-existent transport. Now it
2749 just assumes that the setting of that option is false. This bug was
2750 introduced by 4.51/PH/31.
2752 PH/28 Changed -d+all to exclude +memory, because that information is very
2753 rarely of interest, but it makes the output a lot bigger. People tend to
2754 do -d+all out of habit.
2756 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2757 code in os-type was giving problems when libc.so lives in lib64, like on
2760 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2761 aren't the modern standard, and it seems that some systems' include files
2762 don't always have them. Exim was already checking for some of the newer
2763 ones like T_AAAA, and defining it itself. I've added checks for all the
2764 record types that Exim uses.
2766 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2767 not automatically generating a new one, as it is supposed to. This
2768 prevented TLS from working. If the file did exist, but contained invalid
2769 data, a new version was generated, as expected. It was only the case of a
2770 non-existent file that was broken.
2772 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2773 with a change in libdomainkeys > 0.64.
2775 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2776 from DNS. If the selector record carries the flag, it now has
2777 precedence over the domain-wide flag.
2779 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2781 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2782 the use of a port name that isn't defined in /etc/services) occurred, the
2783 message was deferred as in a normal delivery, and thus remained on the
2784 spool, instead of being failed because of the mua_wrapper setting. This
2785 is now fixed, and I tidied up some of the mua_wrapper messages at the
2788 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2789 the messages in a hash of arrays rather than using individual hashes.
2790 This is a bit cleaner and results in dramatic memory savings, albeit
2791 at a slight CPU cost.
2793 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2794 as requested by Marc Sherman.
2796 SC/10 Eximstats: added histograms for user specified patterns as requested
2799 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2801 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2802 fopen() in the content-scanning modules that did not already have it.
2808 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2809 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2811 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2813 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2815 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2816 to test Sieve filters that use "vacation".
2818 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2819 that changes the way the GnuTLS parameters are stored in the cache file.
2820 The new format can be generated externally. For backward compatibility,
2821 if the data in the cache doesn't make sense, Exim assumes it has read an
2822 old-format file, and it generates new data and writes a new file. This
2823 means that you can't go back to an older release without removing the
2826 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2827 work if there are any delivery delays because "one_time" forces the
2828 parent to be marked "delivered", so its unseen clone is never tried
2829 again. For this reason, Exim now forbids the simultaneous setting of
2832 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2833 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2834 entry if you want to know the details. The fix, however, neglected to
2835 consider the case when local delivery batching is involved. The test for
2836 "previously delivered" was not happening when checking to see if an
2837 address could be batched with a previous (undelivered) one; under
2838 certain circumstances this could lead to multiple deliveries to the same
2841 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2842 in its include files, and this causes problems building Exim.
2844 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2845 header_syntax) but Exim was just ignoring anything given after a slash.
2846 In particular, this caused confusion with an attempt to use "verify =
2847 reverse_host_lookup/defer_ok". An error is now given when options are
2848 supplied for verify items that do not have them. (Maybe reverse_host_
2849 lookup should have a defer_ok option, but that's a different point.)
2851 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2852 defined by RFC 821) to 2048, because there were problems with some AUTH
2853 commands, and RFC 1869 says the size should be increased for extended
2854 SMTP commands that take arguments.
2856 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2859 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2860 "unknown" error; now it says that the functionality isn't in the binary.
2862 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2863 an address' error message when a string expansion fails (syntax or
2864 whatever). Otherwise the password may appear in the log. Following change
2865 PH/42 below, there is no longer a chance of it appearing in a bounce
2868 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2870 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2871 of its addresses was ever tried. (Bugzilla bug #2.)
2873 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2874 the result incorrectly in the debug output. (It correctly added a newline
2875 to what was transported.)
2877 TF/01 Added $received_time.
2879 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2880 commented out examples of how to interface to a virus scanner and to
2881 SpamAssassin. Also added commented examples of av_scanner and
2882 spamd_address settings.
2884 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2885 and controls are allowed in which ACLs. There were a couple of minor
2886 errors. Some of the entries in the conditions table (which is a table of
2887 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2888 negation of where the condition IS allowed.
2890 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2892 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2893 header file does not have a version number, so I've had to invent a new
2894 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2895 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2896 radiusclient), but it was contributed by a Radius user.
2898 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2899 files or directories, for OpenSSL.
2901 PH/19 When an Exim process that is running as root has to create an Exim log
2902 file, it does so in a subprocess that runs as exim:exim so as to get the
2903 ownership right at creation (otherwise, other Exim processes might see
2904 the file with the wrong ownership). There was no test for failure of this
2905 fork() call, which would lead to the process getting stuck as it waited
2906 for a non-existent subprocess. Forks do occasionally fail when resources
2907 run out. I reviewed all the other calls to fork(); they all seem to check
2910 PH/20 When checking for unexpected SMTP input at connect time (before writing
2911 the banner), Exim was not dealing correctly with a non-positive return
2912 from the read() function. If the client had disconnected by this time,
2913 the result was a log entry for a synchronization error with an empty
2914 string after "input=" when read() returned zero. If read() returned -1
2915 (an event I could not check), uninitialized data bytes were printed.
2916 There were reports of junk text (parts of files, etc) appearing after
2919 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2921 PH/22 Added support for macro redefinition, and (re)definition in between
2922 driver and ACL definitions.
2924 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2925 forgetting to use the resulting value; it was using the unexpanded value.
2927 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2928 hadn't been configured. The fix is from Juergen Kreileder, who
2929 understands it better than I do:
2931 "Here's what I see happening with three configured cyrus_sasl
2932 authenticators configured (plain, login, cram-md5):
2934 On startup auth_cyrus_sasl_init() gets called for each of these.
2935 This means three calls to sasl_listmech() without a specified mech_list.
2936 => SASL tests which mechs of all available mechs actually work
2937 => three warnings about OTP not working
2938 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2940 With the patch, sasl_listmech() also gets called three times. But now
2941 SASL's mech_list option is set to the server_mech specified in the the
2942 authenticator. Or in other words, the answer from sasl_listmech()
2943 gets limited to just the mech you're testing for (which is different
2945 => the return list contains just 'plain' or 'login', 'cram-md5' or
2946 nothing depending on the value of ob->server_mech.
2948 I've just tested the patch: Authentication still works fine,
2949 unavailable mechs specified in the exim configuration are still
2950 caught, and the auth.log warnings about OTP are gone."
2952 PH/25 When debugging is enabled, the contents of the command line are added
2953 to the debugging output, even when log_selector=+arguments is not
2956 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2957 answer is "GNU", and only if the return is "GNU/something" is the answer
2960 PH/27 $acl_verify_message is now set immediately after the failure of a
2961 verification in an ACL, and so is available in subsequent modifiers. In
2962 particular, the message can be preserved by coding like this:
2964 warn !verify = sender
2965 set acl_m0 = $acl_verify_message
2967 Previously, $acl_verify_message was set only while expanding "message"
2968 and "log_message" when a very denied access.
2970 PH/28 Modified OS/os.c-Linux with
2972 -#ifndef OS_LOAD_AVERAGE
2973 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2975 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2976 nomenclature these days.)
2978 PH/29 Installed patch from the Sieve maintainer that adds the options
2979 sieve_useraddress and sieve_subaddress to the redirect router.
2981 PH/30 In these circumstances:
2982 . Two addresses routed to the same list of hosts;
2983 . First host does not offer TLS;
2984 . First host accepts first address;
2985 . First host gives temporary error to second address;
2986 . Second host offers TLS and a TLS session is established;
2987 . Second host accepts second address.
2988 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2989 and peerdn, if requested) that were in fact used only for the second
2992 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2993 attention to any local part prefix or suffix that was matched by the
2994 router that accepted the address. It now behaves in the same way as it
2995 does for delivery: the affixes are removed from the local part unless
2996 rcpt_include_affixes is set on the transport.
2998 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2999 timeout during the DATA phase of an incoming message.
3001 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3002 applied a patch sanctioned by the Sieve maintainer.
3004 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3005 the uid or gid is negative. A case of a negative gid caused this to be
3006 noticed. The fix allows for either to be negative.
3008 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3009 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3012 PH/36 The change PH/12 above was broken. Fixed it.
3014 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3015 the grounds that routing the same address twice would always produce the
3016 same answer. This might have been true once, but it is certainly no
3017 longer true now. Routing a child address may depend on the previous
3018 routing that produced that child. Some complicated redirection strategies
3019 went wrong when messages had multiple recipients, and made Exim's
3020 behaviour dependent on the order in which the addresses were given.
3022 I have moved the duplicate checking until after the routing is complete.
3023 Exim scans the addresses that are assigned to local and remote
3024 transports, and removes any duplicates. This means that more work will be
3025 done, as duplicates will always all be routed, but duplicates are
3026 presumably rare, so I don't expect this is of any significance.
3028 For deliveries to pipes, files, and autoreplies, the duplicate checking
3029 still happens during the routing process, since they are not going to be
3032 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3033 It corrects a timeout issue with spamd. This is Ian's comment: "The
3034 background is that sometimes spamd either never reads data from a
3035 connection it has accepted, or it never writes response data. The exiscan
3036 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3037 blindly assumes that writes won't block so it may never time out."
3039 PH/39 Allow G after quota size as well as K and M.
3041 PH/40 The value set for $authenticated_id in an authenticator may not contain
3042 binary zeroes or newlines because the value is written to log lines and
3043 to spool files. There was no check on this. Now the value is run through
3044 the string_printing() function so that such characters are converted to
3045 printable escape sequences.
3047 PH/41 $message_linecount is a new variable that contains the total number of
3048 lines in the message. Compare $body_linecount, which is the count for the
3051 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3052 bounce and delay warning messages, except in certain special cases, which
3055 (a) An SMTP error message from a remote host;
3056 (b) A message specified in a :fail: redirection;
3057 (c) A message specified in a "fail" command in a system filter;
3058 (d) A message specified in a FAIL return from the queryprogram router;
3059 (e) A message specified by the cannot_route_message router option.
3061 In these cases only, Exim does include the error details in bounce and
3062 warning messages. There are also a few cases where bland messages such
3063 as "unrouteable address" or "local delivery error" are given.
3065 PH/43 $value is now also set for the "else" part of a ${run expansion.
3067 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3068 being worked on, but at least Exim now implements the latest version to
3071 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3072 process to complete was treated as a delivery failure, a timeout while
3073 writing the message to the pipe was logged, but erroneously treated as a
3074 successful delivery. Such timeouts include transport filter timeouts. For
3075 consistency with the overall process timeout, these timeouts are now
3076 treated as errors, giving rise to delivery failures by default. However,
3077 there is now a new Boolean option for the pipe transport called
3078 timeout_defer, which, if set TRUE, converts the failures into defers for
3079 both kinds of timeout. A transport filter timeout is now identified in
3082 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3083 systems where "make" and "gmake" are different, calling "gmake" at top
3084 level broke things. I've arranged for the value of $(MAKE) to be passed
3085 from the Makefile to this script so that it can call the same version of
3089 A note about Exim versions 4.44 and 4.50
3090 ----------------------------------------
3092 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3093 changes of various kinds. As a consequence, a big documentation update was
3094 needed. This delayed the release for rather longer than seemed good, especially
3095 in the light of a couple of (minor) security issues. Therefore, the changes
3096 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3097 release. So 4.44 and 4.50 are in effect two different branches that both start
3100 I have left the 4.50 change log unchanged; it contains all the changes since
3101 4.43. The change log for 4.44 is below; many of its items are identical to
3102 those for 4.50. This seems to be the most sensible way to preserve the
3103 historical information.
3109 1. Minor wording change to the doc/README.SIEVE file.
3111 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3112 computation of the current number of files was incorrect.
3114 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3115 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3118 4. Give more explanation in the error message when the command for a transport
3119 filter fails to execute.
3121 5. There are several places where Exim runs a non-Exim command in a
3122 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3123 was being done only for the command run by the queryprogram router. It is
3124 now done for all such subprocesses. The other cases are: ${run, transport
3125 filters, and the commands run by the lmtp and pipe transports.
3127 6. Added CONFIGURE_GROUP build-time option.
3129 7. Some older OS have a limit of 256 on the maximum number of file
3130 descriptors. Exim was using setrlimit() to set 1000 as a large value
3131 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3132 systems. I've change it so that if it can't get 1000, it tries for 256.
3134 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3135 was an oversight, and furthermore, ever since the addition of extra
3136 controls (e.g. 4.43/32), the checks on when to allow different forms of
3137 "control" were broken. There should now be diagnostics for all cases when a
3138 control that does not make sense is encountered.
3140 9. Added the /retain_sender option to "control=submission".
3142 10. $recipients is now available in the predata ACL (oversight).
3144 11. Tidy the search cache before the fork to do a delivery from a message
3145 received from the command line. Otherwise the child will trigger a lookup
3146 failure and thereby defer the delivery if it tries to use (for example) a
3147 cached ldap connection that the parent has called unbind on.
3149 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3150 of $address_data from the recipient verification was clobbered by the
3151 sender verification.
3153 13. The value of address_data from a sender verification is now available in
3154 $sender_address_data in subsequent conditions in the ACL statement.
3156 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3158 15. Added a new option "connect=<time>" to callout options, to set a different
3161 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3162 was its contents. (It was OK if the option was not defined at all.)
3164 17. A "Completed" log line is now written for messages that are removed from
3165 the spool by the -Mrm option.
3167 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3168 information about exactly what failed.
3170 19. Added -dd to debug only the daemon process.
3172 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3173 handles timeouts, both on the server side and network timeouts. Renamed the
3174 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3176 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3177 It is now set to "smtps".
3179 22. $host_address is now set to the target address during the checking of
3180 ignore_target_hosts.
3182 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3183 being passed; this would have caused $sender_host_name to have been used if
3184 matching the list had actually called for a host name (not very likely,
3185 since this list is usually IP addresses). A host name is now passed as
3188 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3189 code that specifies a non-restarting handler (typically sigaction() in
3190 modern systems) in an attempt to fix a rare and obscure crash bug.
3192 25. Narrowed the window for a race in the daemon that could cause it to ignore
3193 SIGCHLD signals. This is not a major problem, because they are used only to
3194 wake it up if nothing else does.
3196 26. A malformed maildirsize file could cause Exim to calculate negative values
3197 for the mailbox size or file count. Odd effects could occur as a result.
3198 The maildirsize information is now recalculated if the size or filecount
3201 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3202 support for a long time. Removed HAVE_SYS_VFS_H.
3204 28. Installed the latest version of exipick from John Jetmore.
3206 29. In an address list, if the pattern was not a regular expression, an empty
3207 subject address (from a bounce message) matched only if the pattern was an
3208 empty string. Non-empty patterns were not even tested. This was the wrong
3209 because it is perfectly reasonable to use an empty address as part of a
3210 database query. An empty address is now tested by patterns that are
3211 lookups. However, all the other forms of pattern expect the subject to
3212 contain a local part and a domain, and therefore, for them, an empty
3213 address still always fails if the pattern is not itself empty.
3215 30. Exim went into a mad DNS loop when attempting to do a callout where the
3216 host was specified on an smtp transport, and looking it up yielded more
3217 than one IP address.
3219 31. Re-factored the code for checking spool and log partition space into a
3220 function that finds that data and another that does the check. The former
3221 is then used to implement four new variables: $spool_space, $log_space,
3222 $spool_inodes, and $log_inodes.
3224 32. The RFC2047 encoding function was originally intended for short strings
3225 such as real names; it was not keeping to the 75-character limit for
3226 encoded words that the RFC imposes. It now respects the limit, and
3227 generates multiple encoded words if necessary. To be on the safe side, I
3228 have increased the buffer size for the ${rfc2047: expansion operator from
3231 33. It is now permitted to omit both strings after an "if" condition; if the
3232 condition is true, the result is "true". As before, when the second string
3233 is omitted, a false condition yields an empty string. This makes it less
3234 cumbersome to write custom ACL and router conditions.
3236 34. Failure to deliver a bounce message always caused it to be frozen, even if
3237 there was an errors_to setting on the router. The errors_to setting is now
3240 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3241 canonical form (fully expanded) before being placed in
3242 $sender_host_address.
3244 36. The table in the code that translates DNS record types into text (T_A to
3245 "A" for instance) was missing entries for NS and CNAME. It is just possible
3246 that this could have caused confusion if both these types were looked up
3247 for the same domain, because the text type is used as part of Exim's
3248 per-process caching. But the chance of anyone hitting this buglet seems
3251 37. The dnsdb lookup has been extended in a number of ways.
3253 (1) There is a new type, "zns", which walks up the domain tree until it
3254 finds some nameserver records. It should be used with care.
3256 (2) There is a new type, "mxh", which is like "mx" except that it returns
3257 just the host names, not the priorities.
3259 (3) It is now possible to give a list of domains (or IP addresses) to be
3260 looked up. The behaviour when one of the lookups defers can be
3261 controlled by a keyword.
3263 (4) It is now possible to specify the separator character for use when
3264 multiple records are returned.
3266 38. The dnslists ACL condition has been extended: it is now possible to supply
3267 a list of IP addresses and/or domains to be looked up in a particular DNS
3270 39. Added log_selector=+queue_time_overall.
3272 40. When running the queue in the test harness, wait just a tad after forking a
3273 delivery process, to get repeatability of debugging output.
3275 41. Include certificate and key file names in error message when GnuTLS fails
3276 to set them up, because the GnuTLS error message doesn't include the name
3277 of the failing file when there is a problem reading it.
3279 42. Allow both -bf and -bF in the same test run.
3281 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3283 44. The "Exiscan patch" is now merged into the mainline Exim source.
3285 45. Sometimes the final signoff response after QUIT could fail to get
3286 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3287 < 0 before doing a fflush(). This bug looks as though it goes back to the
3288 introduction of TLS in release 3.20, but "sometimes" must have been rare
3289 because the tests only now provoked it.
3291 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3292 (this can affect the format of dates).
3294 47. exim_tidydb, when checking for the continued existence of a message for
3295 which it has found a message-specific retry record, was not finding
3296 messages that were in split spool directories. Consequently, it was
3297 deleting retry records that should have stayed in existence.
3299 48. Steve fixed some bugs in eximstats.
3301 49. The SPA authentication driver was not abandoning authentication and moving
3302 on to the next authenticator when an expansion was forced to fail,
3303 contradicting the general specification for all authenticators. Instead it
3304 was generating a temporary error. It now behaves as specified.
3306 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3307 (the order specifies the preference for clients). The order is now AES256,
3308 AES128, 3DES, ARCFOUR128.
3310 51. Small patch to Sieve code - explicitly set From: when generating an
3313 52. Exim crashed if a remote delivery caused a very long error message to be
3314 recorded - for instance if somebody sent an entire SpamAssassin report back
3315 as a large number of 550 error lines. This bug was coincidentally fixed by
3316 increasing the size of one of Exim's internal buffers (big_buffer) that
3317 happened as part of the Exiscan merge. However, to be on the safe side, I
3318 have made the code more robust (and fixed the comments that describe what
3321 53. Now that there can be additional text after "Completed" in log lines (if
3322 the queue_time_overall log selector is set), a one-byte patch to exigrep
3323 was needed to allow it to recognize "Completed" as not the last thing in
3326 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3327 patch that reportedly fixes this has been added. I am not expert enough to
3328 create a test for it. This is what the patch creator wrote:
3330 "I found a little strange behaviour of ldap code when working with
3331 Windows 2003 AD Domain, where users was placed in more than one
3332 Organization Units. When I tried to give exim partial DN, the exit code
3333 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3334 But simultaneously result of request was absolutely normal ldap result,
3335 so I produce this patch..."
3337 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3338 REFERENCE, so I have modified the code to exclude the patch when that macro
3341 55. Some experimental protocols are using DNS PTR records for new purposes. The
3342 keys for these records are domain names, not reversed IP addresses. The
3343 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3344 leaves it alone. Component reversal etc. now happens only for IP addresses.
3347 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3349 57. Double the size of the debug message buffer (to 2048) so that more of very
3350 long debug lines gets shown.
3352 58. The exicyclog utility now does better if the number of log files to keep
3353 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3355 59. Two changes related to the smtp_active_hostname option:
3357 (1) $smtp_active_hostname is now available as a variable.
3358 (2) The default for smtp_banner uses $smtp_active_hostname instead
3359 of $primary_hostname.
3361 60. The host_aton() function is supposed to be passed a string that is known
3362 to be a valid IP address. However, in the case of IPv6 addresses, it was
3363 not checking this. This is a hostage to fortune. Exim now panics and dies
3364 if the condition is not met. A case was found where this could be provoked
3365 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3366 components; fortuitously, this particular loophole had already been fixed
3367 by change 4.50/55 above.
3369 If there are any other similar loopholes, the new check in host_aton()
3370 itself should stop them being exploited. The report I received stated that
3371 data on the command line could provoke the exploit when Exim was running as
3372 exim, but did not say which command line option was involved. All I could
3373 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3374 running as the user.
3377 61. There was a buffer overflow vulnerability in the SPA authentication code
3378 (which came originally from the Samba project). I have added a test to the
3379 spa_base64_to_bits() function which I hope fixes it.
3382 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3383 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3385 63. The daemon start-up calls getloadavg() while still root for those OS that
3386 need the first call to be done as root, but it missed one case: when
3387 deliver_queue_load_max is set with deliver_drop_privilege. This is
3388 necessary for the benefit of the queue runner, because there is no re-exec
3389 when deliver_drop_privilege is set.
3391 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3392 This has been fixed.
3394 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3395 was in use, was not putting the data itself into the right store pool;
3396 consequently, it could be overwritten for a subsequent message in the same
3397 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3400 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3402 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3403 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3404 the function were treating the return as a boolean value, which happened to
3405 work because 0=false and not-0=true, but is not correct code.
3407 68. The host_aton() function was not handling scoped IPv6 addresses (those
3408 with, for example, "%eth0" on the end) correctly.
3410 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3411 negated items (that is, ~something) in unsigned ints. Some compilers
3412 apparently mutter when there is no cast.
3414 70. If an address verification called from an ACL failed, and did not produce a
3415 user-specific message (i.e. there was only a "system" message), nothing was
3416 put in $acl_verify_message. In this situation, it now puts the system
3419 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3420 session; change 4.31/43 added the unwanted input to the log line - except
3421 that it did not do this in the start of session case. It now does.
3423 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3424 This is wrong and can cause the other end to generate a synchronization
3425 error if it is another Exim or anything else that does the synchronization
3426 check. A QUIT command is no longer sent after a timeout.
3428 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3429 during host lookups.
3431 74. The defer_ok option of callout verification was not working if it was used
3432 when verifying addresses in header lines, that is, for this case:
3434 verify = header_sender/callout=defer_ok
3436 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3437 those file descriptors could be used for SMTP connections. If anything
3438 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3439 could be sent to the SMTP client, causing chaos. The daemon now opens
3440 stdin, stdout, and stderr to /dev/null when it puts itself into the
3443 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3444 log by default. The user can override this with suitable Perl magic.
3446 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3447 the log message when discard triggers, was not working for the DATA ACL or
3448 for the non-SMTP ACL.
3450 78. Error message wording change in sieve.c.
3452 79. If smtp_accept_max_per_host was set, the number of connections could be
3453 restricted to fewer than expected, because the daemon was trying to set up
3454 a new connection before checking whether the processes handling previous
3455 connections had finished. The check for completed processes is now done
3456 earlier. On busy systems, this bug wouldn't be noticed because something
3457 else would have woken the daemon, and it would have reaped the completed
3460 80. If a message was submitted locally by a user whose login name contained one
3461 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3462 It caused a spool format error. I have fixed the spool reading code. A
3463 related problem was that the "from" clause in the Received: line became
3464 illegal because of the space(s). It is now covered by ${quote_local_part.
3466 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3469 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3471 83. Patch from Sieve maintainer to fix unterminated string problem in
3472 "vacation" handling.
3474 84. Some minor changes to the Linux configuration files to help with other
3475 OS variants using glibc.
3477 85. One more patch for Sieve to update vacation handling to latest spec.
3480 ----------------------------------------------------
3481 See the note above about the 4.44 and 4.50 releases.
3482 ----------------------------------------------------
3488 1. Change 4.43/35 introduced a bug that caused file counts to be
3489 incorrectly computed when quota_filecount was set in an appendfile
3492 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3493 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3496 3. Give more explanation in the error message when the command for a transport
3497 filter fails to execute.
3499 4. There are several places where Exim runs a non-Exim command in a
3500 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3501 was being done only for the command run by the queryprogram router. It is
3502 now done for all such subprocesses. The other cases are: ${run, transport
3503 filters, and the commands run by the lmtp and pipe transports.
3505 5. Some older OS have a limit of 256 on the maximum number of file
3506 descriptors. Exim was using setrlimit() to set 1000 as a large value
3507 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3508 systems. I've change it so that if it can't get 1000, it tries for 256.
3510 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3511 was an oversight, and furthermore, ever since the addition of extra
3512 controls (e.g. 4.43/32), the checks on when to allow different forms of
3513 "control" were broken. There should now be diagnostics for all cases when a
3514 control that does not make sense is encountered.
3516 7. $recipients is now available in the predata ACL (oversight).
3518 8. Tidy the search cache before the fork to do a delivery from a message
3519 received from the command line. Otherwise the child will trigger a lookup
3520 failure and thereby defer the delivery if it tries to use (for example) a
3521 cached ldap connection that the parent has called unbind on.
3523 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3524 of $address_data from the recipient verification was clobbered by the
3525 sender verification.
3527 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3528 was its contents. (It was OK if the option was not defined at all.)
3530 11. A "Completed" log line is now written for messages that are removed from
3531 the spool by the -Mrm option.
3533 12. $host_address is now set to the target address during the checking of
3534 ignore_target_hosts.
3536 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3537 being passed; this would have caused $sender_host_name to have been used if
3538 matching the list had actually called for a host name (not very likely,
3539 since this list is usually IP addresses). A host name is now passed as
3542 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3543 code that specifies a non-restarting handler (typically sigaction() in
3544 modern systems) in an attempt to fix a rare and obscure crash bug.
3546 15. Narrowed the window for a race in the daemon that could cause it to ignore
3547 SIGCHLD signals. This is not a major problem, because they are used only to
3548 wake it up if nothing else does.
3550 16. A malformed maildirsize file could cause Exim to calculate negative values
3551 for the mailbox size or file count. Odd effects could occur as a result.
3552 The maildirsize information is now recalculated if the size or filecount
3555 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3556 support for a long time. Removed HAVE_SYS_VFS_H.
3558 18. Updated exipick to current release from John Jetmore.
3560 19. Allow an empty sender to be matched against a lookup in an address list.
3561 Previously the only cases considered were a regular expression, or an
3564 20. Exim went into a mad DNS lookup loop when doing a callout where the
3565 host was specified on the transport, if the DNS lookup yielded more than
3568 21. The RFC2047 encoding function was originally intended for short strings
3569 such as real names; it was not keeping to the 75-character limit for
3570 encoded words that the RFC imposes. It now respects the limit, and
3571 generates multiple encoded words if necessary. To be on the safe side, I
3572 have increased the buffer size for the ${rfc2047: expansion operator from
3575 22. Failure to deliver a bounce message always caused it to be frozen, even if
3576 there was an errors_to setting on the router. The errors_to setting is now
3579 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3580 canonical form (fully expanded) before being placed in
3581 $sender_host_address.
3583 24. Updated eximstats to version 1.33
3585 25. Include certificate and key file names in error message when GnuTLS fails
3586 to set them up, because the GnuTLS error message doesn't include the name
3587 of the failing file when there is a problem reading it.
3589 26. Expand error message when OpenSSL has problems setting up cert/key files.
3592 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3593 (this can affect the format of dates).
3595 28. exim_tidydb, when checking for the continued existence of a message for
3596 which it has found a message-specific retry record, was not finding
3597 messages that were in split spool directories. Consequently, it was
3598 deleting retry records that should have stayed in existence.
3600 29. eximstats updated to version 1.35
3601 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3602 1.35 - bugfix such that pie charts by volume are generated correctly
3604 30. The SPA authentication driver was not abandoning authentication and moving
3605 on to the next authenticator when an expansion was forced to fail,
3606 contradicting the general specification for all authenticators. Instead it
3607 was generating a temporary error. It now behaves as specified.
3609 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3610 (the order specifies the preference for clients). The order is now AES256,
3611 AES128, 3DES, ARCFOUR128.
3613 31. Small patch to Sieve code - explicitly set From: when generating an
3616 32. Exim crashed if a remote delivery caused a very long error message to be
3617 recorded - for instance if somebody sent an entire SpamAssassin report back
3618 as a large number of 550 error lines. This bug was coincidentally fixed by
3619 increasing the size of one of Exim's internal buffers (big_buffer) that
3620 happened as part of the Exiscan merge. However, to be on the safe side, I
3621 have made the code more robust (and fixed the comments that describe what
3624 33. Some experimental protocols are using DNS PTR records for new purposes. The
3625 keys for these records are domain names, not reversed IP addresses. The
3626 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3627 leaves it alone. Component reversal etc. now happens only for IP addresses.
3630 34. The host_aton() function is supposed to be passed a string that is known
3631 to be a valid IP address. However, in the case of IPv6 addresses, it was
3632 not checking this. This is a hostage to fortune. Exim now panics and dies
3633 if the condition is not met. A case was found where this could be provoked
3634 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3635 components; fortuitously, this particular loophole had already been fixed
3636 by change 4.50/55 or 4.44/33 above.
3638 If there are any other similar loopholes, the new check in host_aton()
3639 itself should stop them being exploited. The report I received stated that
3640 data on the command line could provoke the exploit when Exim was running as
3641 exim, but did not say which command line option was involved. All I could
3642 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3643 running as the user.
3646 35. There was a buffer overflow vulnerability in the SPA authentication code
3647 (which came originally from the Samba project). I have added a test to the
3648 spa_base64_to_bits() function which I hope fixes it.
3651 36. The daemon start-up calls getloadavg() while still root for those OS that
3652 need the first call to be done as root, but it missed one case: when
3653 deliver_queue_load_max is set with deliver_drop_privilege. This is
3654 necessary for the benefit of the queue runner, because there is no re-exec
3655 when deliver_drop_privilege is set.
3657 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3658 was in use, was not putting the data itself into the right store pool;
3659 consequently, it could be overwritten for a subsequent message in the same
3660 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3663 38. Sometimes the final signoff response after QUIT could fail to get
3664 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3665 < 0 before doing a fflush(). This bug looks as though it goes back to the
3666 introduction of TLS in release 3.20, but "sometimes" must have been rare
3667 because the tests only now provoked it.
3673 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3674 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3675 Now it returns an integer. A number of calls were still expecting a T/F
3676 return. Fortuitously, in all cases, the tests worked in OK situations,
3677 which is the norm. However, things would have gone wrong on any write
3678 failures on the smtp file descriptor. This function is used when sending
3679 messages over SMTP and also when doing verify callouts.
3681 2. When Exim is called to do synchronous delivery of a locally submitted
3682 message (the -odf or -odi options), it no longer closes stderr before doing
3685 3. Implemented the mua_wrapper option.
3687 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3689 5. Implemented the functions header_remove(), header_testname(),
3690 header_add_at_position(), and receive_remove_recipient(), and exported them
3693 6. If an ACL "warn" statement specified the addition of headers, Exim already
3694 inserted X-ACL-Warn: at the start if there was no header name. However, it
3695 was not making this test for the second and subsequent header lines if
3696 there were newlines in the string. This meant that an invalid header could
3697 be inserted if Exim was badly configured.
3699 7. Allow an ACL "warn" statement to add header lines at the start or after all
3700 the Received: headers, as well as at the end.
3702 8. Added the rcpt_4xx retry error code.
3704 9. Added postmaster_mailfrom=xxx to callout verification option.
3706 10. Added mailfrom=xxxx to the callout verification option, for verify=
3709 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3710 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3712 12. Inserted some casts to stop certain compilers warning when using pointer
3713 differences as field lengths or precisions in printf-type calls (mostly
3714 affecting debugging statements).
3716 13. Added optional readline() support for -be (dynamically loaded).
3718 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3719 same clock tick as a message's arrival, so that its received time was the
3720 same as the "first fail" time on the retry record, and that message
3721 remained on the queue past the ultimate address timeout, every queue runner
3722 would try a delivery (because it was past the ultimate address timeout) but
3723 after another failure, the ultimate address timeout, which should have then
3724 bounced the address, did not kick in. This was a "< instead of <=" error;
3725 in most cases the first failure would have been in the next clock tick
3726 after the received time, and all would be well.
3728 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3729 being recognized when the domain list was tested by the match_domain
3730 condition in an expansion string.
3732 16. Added the ${str2b64: operator.
3734 17. Exim was always calling setrlimit() to set a large limit for the number of
3735 processes, without checking whether the existing limit was already
3736 adequate. (It did check for the limit on file descriptors.) Furthermore,
3737 errors from getrlimit() and setrlimit() were being ignored. Now they are
3738 logged to the main and panic logs, but Exim does carry on, to try to do its
3739 job under whatever limits there are.
3741 18. Imported PCRE 5.0.
3743 19. Trivial typo in log message " temporarily refused connection" (the leading
3746 20. If the log selector return_path_on_delivery was set and an address was
3747 redirected to /dev/null, the delivery process crashed because it assumed
3748 that a return path would always be set for a "successful" delivery. In this
3749 case, the whole delivery is bypassed as an optimization, and therefore no
3752 21. Internal re-arrangement: the function for sending a challenge and reading
3753 a response while authentication was assuming a zero-terminated challenge
3754 string. It's now changed to take a pointer and a length, to allow for
3755 binary data in such strings.
3757 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3759 23. Exim was not respecting finduser_retries when seeking the login of the
3760 uid under which it was called; it was always trying 10 times. (The default
3761 setting of finduser_retries is zero.) Also, it was sleeping after the final
3762 failure, which is pointless.
3764 24. Implemented tls_on_connect_ports.
3766 25. Implemented acl_smtp_predata.
3768 26. If the domain in control=submission is set empty, Exim assumes that the
3769 authenticated id is a complete email address when it generates From: or
3770 Sender: header lines.
3772 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3773 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3774 chown and chgrp in /bin and hostname in /usr/bin.
3776 28. Exim was keeping the "process log" file open after each use, just as it
3777 does for the main log. This opens the possibility of it remaining open for
3778 long periods when the USR1 signal hits a daemon. Occasional processlog
3779 errors were reported, that could have been caused by this. Anyway, it seems
3780 much more sensible not to leave this file open at all, so that is what now
3783 29. The long-running daemon process does not normally write to the log once it
3784 has entered its main loop, and it closes the log before doing so. This is
3785 so that log files can straightforwardly be renamed and moved. However,
3786 there are a couple of unusual error situations where the daemon does write
3787 log entries, and I had neglected to close the log afterwards.
3789 30. The text of an SMTP error response that was received during a remote
3790 delivery was being truncated at 512 bytes. This is too short for some of
3791 the long messages that one sometimes sees. I've increased the limit to
3794 31. It is now possible to make retry rules that apply only when a message has a
3795 specific sender, in particular, an empty sender.
3797 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3798 it possible to be selective about when SMTP synchronization is enforced.
3800 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3802 32. Implemented hosts_connection_nolog.
3804 33. Added an ACL for QUIT.
3806 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3809 35. Added mailbox_size and mailbox_filecount to appendfile.
3811 36. Added control = no_multiline_responses to ACLs.
3813 37. There was a bug in the logic of the code that waits for the clock to tick
3814 in the case where the clock went backwards by a substantial amount such
3815 that the microsecond fraction of "now" was more than the microsecond
3816 fraction of "then" (but the whole seconds number was less).
3818 38. Added support for the libradius Radius client library this is found on
3819 FreeBSD (previously only the radiusclient library was supported).
3825 1. When certain lookups returned multiple values in the form name=value, the
3826 quoting of the values was not always being done properly. Specifically:
3827 (a) If the value started with a double quote, but contained no whitespace,
3829 (b) If the value contained whitespace other than a space character (i.e.
3830 tabs or newlines or carriage returns) it was not quoted.
3831 This fix has been applied to the mysql and pgsql lookups by writing a
3832 separate quoting function and calling it from the lookup code. The fix
3833 should probably also be applied to nisplus, ibase and oracle lookups, but
3834 since I cannot test any of those, I have not disturbed their existing code.
3836 2. A hit in the callout cache for a specific address caused a log line with no
3837 reason for rejecting RCPT. Now it says "Previous (cached) callout
3838 verification failure".
3840 3. There was an off-by-one bug in the queryprogram router. An over-long
3841 return line was truncated at 256 instead of 255 characters, thereby
3842 overflowing its buffer with the terminating zero. As well as fixing this, I
3843 have increased the buffer size to 1024 (and made a note to document this).
3845 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3846 when Exim is waiting for an SMTP response from a remote server, Exim
3847 restarts its select() call on the socket, thereby resetting its timeout.
3848 This is not a problem when such interrupts are rare. Somebody set up a cron
3849 job to run exiwhat every 2 minutes, which is less than the normal select()
3850 timeout (5 or 10 minutes). This meant that the select() timeout never
3851 kicked in because it was always reset. I have fixed this by comparing the
3852 time when an interrupt arrives with the time at the start of the first call
3853 to select(). If more time than the timeout has elapsed, the interrupt is
3854 treated as a timeout.
3856 5. Some internal re-factoring in preparation for the addition of Sieve
3857 extensions (by MH). In particular, the "personal" test is moved to a
3858 separate function, and given an option for scanning Cc: and Bcc: (which is
3859 not set for Exim filters).
3861 6. When Exim created an email address using the login of the caller as the
3862 local part (e.g. when creating a From: or Sender: header line), it was not
3863 quoting the local part when it contained special characters such as @.
3865 7. Installed new OpenBSD configuration files.
3867 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3868 try to make them clearer.
3870 9. Callout options, other than the timeout value, were being ignored when
3871 verifying sender addresses in header lines. For example, when using
3873 verify = header_sender/callout=no_cache
3875 the cache was (incorrectly) being used.
3877 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3878 only the Cygwin environment.
3880 11. When return_path_on_delivery was set as a log selector, if different remote
3881 addresses in the same message used different return paths and parallel
3882 remote delivery occurred, the wrong values would sometimes be logged.
3883 (Whenever a remote delivery process finished, the return path value from
3884 the most recently started remote delivery process was logged.)
3886 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3887 lines when AUTH and/or TLS are in use. This is the "received protocol"
3888 field. Exim used to use "asmtp" for authenticated SMTP, without any
3889 indication (in the protocol name) for TLS use. Now it follows the RFC and
3890 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3891 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3892 names appear in log lines as well as in Received: header lines.
3894 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3895 extensions, and comparison tests, and to fix some bugs.
3897 14. Changes to the "personal" filter test:
3899 (1) The test was buggy in that it was just doing the equivalent of
3900 "contains" tests on header lines. For example, if a user's address was
3901 anne@some.where, the "personal" test would incorrectly be true for
3903 To: susanne@some.where
3905 This test is now done by extracting each address from the header in turn,
3906 and checking the entire address. Other tests that are part of "personal"
3907 are now done using regular expressions (for example, to check local parts
3908 of addresses in From: header lines).
3910 (2) The list of non-personal local parts in From: addresses has been
3911 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3912 taken from the Sieve specification recommendations.
3914 (3) If the message contains any header line starting with "List-" it is
3915 treated as non-personal.
3917 (4) The test for "circular" in the Subject: header line has been removed
3918 because it now seems ill-conceived.
3920 15. Minor typos in src/EDITME comments corrected.
3922 16. Installed latest exipick from John Jetmore.
3924 17. If headers_add on a router specified a text string that was too long for
3925 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3926 of string_sprintf() is now avoided.
3928 18. $message_body_size was not set (it was always zero) when running the DATA
3929 ACL and the local_scan() function.
3931 19. For the "mail" command in an Exim filter, no default was being set for
3932 the once_repeat time, causing a random time value to be used if "once" was
3933 specified. (If the value happened to be <= 0, no repeat happened.) The
3934 default is now 0s, meaning "never repeat". The "vacation" command was OK
3935 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3936 (I found it when inspecting the code).
3938 20. There is now an overall timeout for performing a callout verification. It
3939 defaults to 4 times the callout timeout, which applies to individual SMTP
3940 commands during the callout. The overall timeout applies when there is more
3941 than one host that can be tried. The timeout is checked before trying the
3942 next host. This prevents very long delays if there are a large number of
3943 hosts and all are timing out (e.g. when the network connections are timing
3944 out). The value of the overall timeout can be changed by specifying an
3945 additional sub-option for "callout", called "maxwait". For example:
3947 verify = sender/callout=5s,maxwait=20s
3949 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3950 to the end before writing, but this should make it even safer).
3952 22. Exim was forgetting that it had advertised PIPELINING for the second and
3953 subsequent messages on an SMTP connection. It was also not resetting its
3954 memory on STARTTLS and an internal HELO.
3956 23. When Exim logs an SMTP synchronization error within a session, it now
3957 records whether PIPELINING has been advertised or not.
3959 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3960 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3961 rather than long int.
3963 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3965 26. Added the never_mail option to autoreply.
3971 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3972 crash if the getsockname() call failed; this can happen if a connection is
3973 closed very soon after it is established. The problem was simply in the
3974 order in which certain operations were done, causing Exim to try to write
3975 to the SMTP stream before it had set up the file descriptor. The bug has
3976 been fixed by making things happen in the correct order.
3982 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3983 before the connection was closed, thus losing the rejection response.
3985 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3986 some early Solaris releases, but causes trouble in current releases where
3987 socklen_t is defined.
3989 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3992 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3995 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3996 that releasing the top of it at the end releases what was used for sub-
3997 expansions (unless the block got too big). However, discard this block if
3998 the first thing is a variable or header, so that we can use its block when
3999 it is dynamic (useful for very large $message_headers, for example).
4001 6. Lookups now cache *every* query, not just the most recent. A new, separate
4002 store pool is used for this. It can be recovered when all lookup caches are
4003 flushed. Lookups now release memory at the end of their result strings.
4004 This has involved some general refactoring of the lookup sources.
4006 7. Some code has been added to the store_xxx() functions to reduce the amount
4007 of flapping under certain conditions.
4009 8. log_incoming_interface used to affect only the <= reception log lines. Now
4010 it causes the local interface and port to be added to several more SMTP log
4011 lines, for example "SMTP connection from", and rejection lines.
4013 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4015 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4017 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4018 could be overwritten at the end of the current message (or the start of a
4019 new message if it was set in a HELO ACL). The value is now preserved for
4020 the duration of the SMTP connection.
4022 12. If a transport had a headers_rewrite setting, and a matching header line
4023 contained an unqualified address, that address was qualified, even if it
4024 did not match any rewriting rules. The underlying bug was that the values
4025 of the flags that permit the existence of unqualified sender and recipient
4026 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4027 non-local messages, and by -bnq for local messages) were not being
4028 preserved with the message after it was received.
4030 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4031 "next input=" as part of the text comprising the host identity instead of
4032 the correct text. The code was using the same buffer for two different
4033 strings. However, depending on which order the printing function evaluated
4034 its arguments, the bug did not always show up. Under Linux, for example, my
4035 test suite worked just fine.
4037 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4038 doesn't work with some older versions of Perl. It has been changed to "my",
4039 which in any case is probably the better facility to use.
4041 15. A really picky compiler found some instances of statements for creating
4042 error messages that either had too many or two few arguments for the format
4045 16. The size of the buffer for calls to the DNS resolver has been increased
4046 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4047 for addresses that have a lot of PTR records. This alleviates a problem; it
4048 does not fully solve it.
4050 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4051 buffer now truncates the list and logs the incident, which is the same
4052 action as happens when Exim is looking up a host name and its aliases.
4053 Previously in this situation something unpredictable would happen;
4054 sometimes it was "internal error: store_reset failed".
4056 18. If a server dropped the connection unexpectedly when an Exim client was
4057 using GnuTLS and trying to read a response, the client delivery process
4058 crashed while trying to generate an error log message.
4060 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4061 string, for example:
4063 warn message = H1: something\nH2: something
4065 the text was added as a single header line from Exim's point of view
4066 though it ended up OK in the delivered message. However, searching for the
4067 second and subsequent header lines using $h_h2: did not work. This has been
4068 fixed. Similarly, if a system filter added multiple headers in this way,
4069 the routers could not see them.
4071 20. Expanded the error message when iplsearch is called with an invalid key to
4072 suggest using net-iplsearch in a host list.
4074 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4075 ACLs are no longer actually imposed (and a message to that effect is
4078 22. If a "gecos" field in a passwd entry contained escaped characters, in
4079 particular, if it contained a \" sequence, Exim got it wrong when building
4080 a From: or a Sender: header from that name. A second bug also caused
4081 incorrect handling when an unquoted " was present following a character
4082 that needed quoting.
4084 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4085 was not being matched caselessly.
4087 24. Arranged for all hyphens in the exim.8 source to be escaped with
4090 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4091 properly. Recipient callout cache records were still being keyed to include
4092 the sender, even when use_sender was set false. This led to far more
4093 callouts that were necessary. The sender is no longer included in the key
4094 when use_sender is false.
4096 26. Added "control = submission" modifier to ACLs.
4098 27. Added the ${base62d: operator to decode base 62 numbers.
4100 28. dnsdb lookups can now access SRV records.
4102 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4103 the configuration file.
4105 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4106 (-v) mode. This makes the output for a verbose queue run more intelligible.
4108 31. Added a use_postmaster feature to recipient callouts.
4110 32. Added the $body_zerocount variable, containing the number of binary zero
4111 bytes in the message body.
4113 33. The time of last modification of the "new" subdirectory is now used as the
4114 "mailbox time last read" when there is a quota error for a maildir
4117 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4119 35. Added +ignore_unknown as a special item in host lists.
4121 36. Code for decoding IPv6 addresses in host lists is now included, even if
4122 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4123 address was recognized as an IP address, but was then not correctly decoded
4124 into binary, causing unexpected and incorrect effects when compared with
4131 1. Very minor rewording of debugging text in manualroute to say "list of
4132 hosts" instead of "hostlist".
4134 2. If verify=header_syntax was set, and a header line with an unqualified
4135 address (no domain) and a large number of spaces between the end of the
4136 name and the colon was received, the reception process suffered a buffer
4137 overflow, and (when I tested it) crashed. This was caused by some obsolete
4138 code that should have been removed. The fix is to remove it!
4140 3. When running in the test harness, delay a bit after writing a bounce
4141 message to get a bit more predictability in the log output.
4143 4. Added a call to search_tidyup() just before forking a reception process. In
4144 theory, someone could use a lookup in the expansion of smtp_accept_max_
4145 per_host which, without the tidyup, could leave open a database connection.
4147 5. Added the variables $recipient_data and $sender_data which get set from a
4148 lookup success in an ACL "recipients" or "senders" condition, or a router
4149 "senders" option, similar to $domain_data and $local_part_data.
4151 6. Moved the writing of debug_print from before to after the "senders" test
4154 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4155 problems for message scanning, either using a data ACL, or using
4156 local_scan() because the Received: header was not generated till after they
4157 were called (in order to set the time as the time of reception completion).
4158 I have revised the way this works. The header is now generated after the
4159 body is received, but before the ACL or local_scan() are called. After they
4160 are run, the timestamp in the header is updated.
4166 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4167 before starting a queue runner without re-exec. This happened only when
4168 deliver_drop_privilege was set or when the Exim user was set to root. The
4169 effect of the bug was that timeouts during subsequent deliveries caused
4170 crashes instead of being properly handled. The handler is now left at its
4171 default (and expected) setting.
4173 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4174 message, again when deliver_drop_privilege is set or Exim is run as root.
4175 The bug described in (1) was not present in this case, but the tidying up
4176 of the other signals was missing. I have made the two cases consistent.
4178 3. The ignore_target_hosts setting on a manualroute router was being ignored
4179 for hosts that were looked up using the /MX notation.
4181 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4184 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4185 operated on the sender address. After changing the $sender_address to <>
4186 for the sender address verify, Exim was re-instated it as the original
4187 (before rewriting) address, but remembering that it had rewritten it, so it
4188 wasn't rewriting it again. This bug also had the effect of breaking the
4189 sender address verification caching when the sender address was rewritten.
4191 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4192 This has been changed so that if the ip literal address matches
4193 ignore_target_hosts, the router declines.
4195 7. Added expansion conditions match_domain, match_address, and match_local_
4196 part (NOT match_host).
4198 8. The placeholder for the Received: header didn't have a length field set.
4200 9. Added code to Exim itself and to exim_lock to test for a specific race
4201 condition that could lead to file corruption when using MBX delivery. The
4202 issue is with the lockfile that is created in /tmp. If this file is removed
4203 after a process has opened it but before that process has acquired a lock,
4204 there is the potential for a second process to recreate the file and also
4205 acquire a lock. This could lead to two Exim processes writing to the file
4206 at the same time. The added code performs the same test as UW imapd; it
4207 checks after acquiring the lock that its file descriptor still refers to
4208 the same named file.
4210 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4211 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4214 11. Added the smtp_active_hostname option. If used, this will typically be made
4215 to depend on the incoming interface address. Because $interface_address is
4216 not set up until the daemon has forked a reception process, error responses
4217 that can happen earlier (such as "too many connections") no longer contain
4220 12. If an expansion in a condition on a "warn" statement fails because a lookup
4221 defers, the "warn" statement is abandoned, and the next ACL statement is
4222 processed. Previously this caused the whole ACL to be aborted.
4224 13. Added the iplsearch lookup type.
4226 14. Added ident_timeout as a log selector.
4228 15. Added tls_certificate_verified as a log selector.
4230 16. Added a global option tls_require_ciphers (compare the smtp transport
4231 option of the same name). This controls incoming TLS connections.
4233 17. I finally figured out how to make tls_require_ciphers do a similar thing
4234 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4235 before starting the TLS session.
4237 18. Tabs are now shown as \t in -bP output.
4239 19. If the log selector return_path_on_delivery was set, Exim crashed when
4240 bouncing a message because it had too many Received: header lines.
4242 20. If two routers both had headers_remove settings, and the first one included
4243 a superfluous trailing colon, the final name in the first list and the
4244 first name in the second list were incorrectly joined into one item (with a
4245 colon in the middle).
4251 1. Added -C and -D options to the exinext utility, mainly to make it easier
4252 to include in the automated testing, but these could be helpful when
4253 multiple configurations are in use.
4255 2. The exinext utility was not formatting the output nicely when there was
4256 an alternate port involved in the retry record key, nor when there was a
4257 message id as well (for retries that were specific to a specific message
4258 and a specific host). It was also confused by IPv6 addresses, because of
4259 the additional colons they contain. I have fixed the IPv4 problem, and
4260 patched it up to do a reasonable job for IPv6.
4262 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4263 delivery, the log line now contains "pipelined" if PIPELINING was used.
4265 4. An SMTP transport process used to panic and die if the bind() call to set
4266 an explicit outgoing interface failed. This has been changed; it is now
4267 treated in the same way as a connect() failure.
4269 5. A reference to $sender_host_name in the part of a conditional expansion
4270 that was being skipped was still causing a DNS lookup. This no longer
4273 6. The def: expansion condition was not recognizing references to header lines
4274 that used bh_ and bheader_.
4276 7. Added the _cache feature to named lists.
4278 8. The code for checking quota_filecount in the appendfile transport was
4279 allowing one more file than it should have been.
4281 9. For compatibility with Sendmail, the command line option
4289 and sets the incoming protocol and host name (for trusted callers). The
4290 host name and its colon can be omitted when only the protocol is to be set.
4291 Note the Exim already has two private options, -pd and -ps, that refer to
4292 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4293 "s", but I don't think that's a major issue.
4295 10. A number of refactoring changes to the code, none of which should affect
4298 (a) The number of logging options was getting close to filling up the
4299 32-bit word that was used as a bit map. I have split them into two classes:
4300 those that are passed in the argument to log_write(), and those that are
4301 only ever tested independently outside of that function. These are now in
4302 separate 32-bit words, so there is plenty of room for expansion again.
4303 There is no change in the user interface or the logging behaviour.
4305 (b) When building, for example, log lines, the code previously used a
4306 macro that called string_cat() twice, in order to add two strings. This is
4307 not really sufficiently general. Furthermore, there was one instance where
4308 it was actually wrong because one of the argument was used twice, and in
4309 one call a function was used. (As it happened, calling the function twice
4310 did not affect the overall behaviour.) The macro has been replaced by a
4311 function that can join an arbitrary number of extra strings onto a growing
4314 (c) The code for expansion conditions now uses a table and a binary chop
4315 instead of a serial search (which was left over from when there were very
4316 few conditions). Also, it now recognizes conditions like "pam" even when
4317 the relevant support is not compiled in: a suitably worded error message is
4318 given if an attempt is made to use such a condition.
4320 11. Added ${time_interval:xxxxx}.
4322 12. A bug was causing one of the ddress fields not to be passed back correctly
4323 from remote delivery subprocesses. The field in question was not being
4324 subsequently used, so this caused to problems in practice.
4326 13. Added new log selectors queue_time and deliver_time.
4328 14. Might have fixed a bug in maildirsizefile handling that threw up
4329 "unexpected character" debug warnings, and recalculated the data
4330 unnecessarily. In any case, I expanded the warning message to give more
4333 15. Added the message "Restricted characters in address" to the statements in
4334 the default ACL that block characters like @ and % in local parts.
4336 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4337 Three changes have been made:
4339 (a) There was a serious bug; a negative response to MAIL caused the whole
4340 recipient domain to be cached as invalid, thereby blocking all messages
4341 to all local parts at the same domain, from all senders. This bug has
4342 been fixed. The domain is no longer cached after a negative response to
4343 MAIL if the sender used is not empty.
4345 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4348 (c) A new callout option, "use_sender" has been added for people who want
4349 the modified behaviour.
4355 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4358 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4359 indeed breaks things for older releases.
4361 3. Added additional logging to the case where there is a problem reading data
4362 from a filter that is running in a subprocess using a pipe, in order to
4363 try to track down a specific problem.
4365 4. Testing facility fudge: when running in the test harness and attempting
4366 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4367 getting "No route to host". Convert this to a timeout.
4369 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4372 6. Some OS don't have socklen_t but use size_t instead. This affects the
4373 fifth argument of getsockopt() amongst other things. This is now
4374 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4375 can be set for individual OS. I have set it for SunOS5, OSF1, and
4376 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4377 some earlier ones do not.
4379 7. Change 4.30/15 was not doing the test caselessly.
4381 8. The standard form for an IPv6 address literal was being rejected by address
4382 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4383 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4384 this, as well as the form without the "IPv6" on the front (but only when
4385 address literals are enabled, of course).
4387 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4389 10. Exim crashed if a message with an empty sender address specified by -f
4390 encountered a router with an errors_to setting. This could be provoked only
4391 by a command such as
4395 where an empty string was supplied; "<>" did not hit this bug.
4397 11. Installed PCRE release 4.5.
4399 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4400 remained set. It is now erased.
4402 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4403 times from message ids (which are base 36 rather than the normal 62).
4405 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4406 were being counted as actual protocol errors, and logged if the log
4407 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4408 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4409 and DATA following a set of rejected RCPTs do not count as protocol errors.
4410 In other words, Exim assumes they were pipelined, though this may not
4411 actually be the case. Of course, in all cases the client gets an
4412 appropriate error code.
4414 15. If a lookup fails in an ACL condition, a message about the failure may
4415 be available; it is used if testing the ACL cannot continue, because most
4416 such messages specify what the cause of the deferral is. However, some
4417 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4418 that caused an old message to be retained and used if a later statement
4419 caused a defer, replacing the real cause of the deferral.
4421 16. If an IP address had so many PTR records that the DNS lookup buffer
4422 was not large enough to hold them, Exim could crash while trying to process
4423 the truncated data. It now detects and logs this case.
4425 17. Further to 4.21/58, another change has been made: if (and only if) the
4426 first line of a message (the first header line) ends with CRLF, a bare LF
4427 in a subsequent header line has a space inserted after it, so as not to
4428 terminate the header.
4430 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4431 unnecessarily, used atoi() instead of strtol(), and didn't check the
4432 termination when getting file sizes from file names by regex.
4434 19. Completely re-implemented the support for maildirsize files, in the light
4435 of a number of problems with the previous contributed implementation
4436 (4.30/29). In particular:
4438 . If the quota is zero, the maildirsize file is maintained, but no quota is
4441 . If the maildir directory does not exist, it is created before any attempt
4442 to write a maildirsize file.
4444 . The quota value in the file is just a cache; if the quota is changed in
4445 the transport, the new value overrides.
4447 . A regular expression is available for excluding directories from the
4450 20. The autoreply transport checks the characters in options that define the
4451 message's headers; it allows continued headers, but it was checking with
4452 isspace() after an embedded newline instead of explicitly looking for a
4455 21. If all the "regular" hosts to which an address was routed had passed their
4456 expiry times, and had not reached their retry times, the address was
4457 bounced, even if fallback hosts were defined. Now Exim should go on to try
4460 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4461 equivalent code in the SMTP transport. Some hosts send humungous responses
4462 to HELO/EHLO, more than 1024 it seems.
4464 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4465 gives compiler warnings in some environments. I've now done it "properly",
4468 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4469 (because of problems with the built-in one) was declared to return uschar *
4470 instead of char *, causing compiler failure.
4472 25. Fixed a file descriptor leak when processing alias/forward files.
4474 26. Fixed a minor format string issue in dbfn.c.
4476 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4478 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4479 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4482 29. When a sender address is verified, it is cached, to save repeating the test
4483 when there is more than one recipient in a message. However, when the
4484 verification involves a callout, it is possible for different callout
4485 options to be set for different recipients. It is too complicated to keep
4486 track of this in the cache, so now Exim always runs a verification when a
4487 callout is required, relying on the callout cache for the optimization.
4488 The overhead is duplication of the address routing, but this should not be
4491 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4492 to be verified with callout=postmaster, and the main callout worked but the
4493 postmaster check failed, the verification correctly failed. However, if a
4494 subsequent RCPT command asked for sender verification *without* the
4495 postmaster check, incorrect caching caused this verification also to fail,
4498 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4499 it was not caching the DNS options (qualify_single, search_parents) that
4500 were used when the lookup failed. A subsequent lookup with different
4501 options therefore always gave the same answer, though there were cases
4502 where it should not have. (Example: a "domains = !$mx_any" option on a
4503 dnslookup router: the "domains" option is always processed without any
4504 widening, but the router might have qualify_single set.) Now Exim uses the
4505 cached value only when the same options are set.
4507 32. Added John Jetmore's "exipick" utility to the distribution.
4509 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4510 than a timeout (e.g. a certificate is required, and is not provided), an
4511 Exim server now closes the connection immediately. Previously it waited for
4512 the client to close - but if the client is SSL, it seems that they each
4513 wait for each other, leading to a delay before one of them times out.
4515 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4516 maintained 0.8.x compatibility because I don't think many are using it, and
4517 it is clearly obsolete.
4519 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4522 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4523 client certificate was expired. A simple patch fixes this, though I don't
4524 understand the full logic of why the verify callback is called multiple
4527 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4528 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4529 which causes problems with some clients (such as the Certicom SSL Plus
4530 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4531 disables the coutermeasure allowing Eudora to connect."
4533 38. Exim was not checking that a write() to a log file succeeded. This could
4534 lead to Bad Things if a log got too big, in particular if it hit a file
4535 size limit. Exim now panics and dies if it cannot write to a log file, just
4536 as it does if it cannot open a log file.
4538 39. Modified OS/Makefile-Linux so that it now contains
4540 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4542 The two -D definitions ensure that Exim is compiled with large file
4543 support, which makes it possible to handle log files that are bigger than
4546 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4547 instance) a domain was checked against a named list that involved a lookup,
4548 causing $domain_data to be set, then another domain was checked against the
4549 same list, then the first domain was re-checked, the value of $domain_data
4550 after the final check could be wrong. In particular, if the second check
4551 failed, it could be set empty. This bug probably also applied to
4554 41. The strip_trailing_dot option was not being applied to the address given
4555 with the -f command-line option.
4557 42. The code for reading a message's header from the spool was incrementing
4558 $received_count, but never initializing it. This meant that the value was
4559 incorrect (doubled) while delivering a message in the same process in which
4560 it was received. In the most common configuration of Exim, this never
4561 happens - a fresh exec is done - but it can happen when
4562 deliver_drop_privilege is set.
4564 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4565 it now includes up to 150 characters of the unexpected data in the log
4568 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4569 and building data strings. The size of both of these buffers was 10 000
4570 bytes - far larger than anybody would *ever* want, thought I. Needless to
4571 say, somebody hit the limit. I have increased the maximum line length to
4572 20 000 and the maximum data length of concatenated lines to 100 000. I have
4573 also fixed two bugs, because there was no checking on these buffers. Tsk,
4574 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4575 buffer is too small.
4577 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4578 lsearch lookups. Now it does.
4580 46. When parsing a route_list item in a manualroute router, a fixed-length
4581 buffer was used for the list of hosts. I made this 1024 bytes long,
4582 thinking that nobody would ever have a list of hosts that long. Wrong.
4583 Somebody had a whole pile of complicated expansion conditions, and the
4584 string was silently truncated, leading to an expansion error. It turns out
4585 that it is easier to change to an unlimited length (owing to other changes
4586 that have happened since this code was originally written) than to build
4587 structure for giving a limitation error. The length of the item that
4588 expands into the list of hosts is now unlimited.
4590 47. The lsearch lookup could not handle data where the length of text line was
4591 more than 4095 characters. Such lines were truncated, leading to shortened
4592 data being returned. It should now handle lines of any length.
4594 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4595 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4598 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4599 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4600 no longer happens in comment lines. A list of replacements is now placed
4601 at the head of all of the source files, except those whose only change is
4602 to replace PERL_COMMAND in the very first #! line.
4604 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4605 messages on the queue, with a bottom-up merge sort, using code contributed
4606 by Michael Haardt. This should make operations like -bp somewhat faster on
4607 large queues. It won't affect queue runners, except when queue_run_in_order
4610 51. Installed eximstats 1.31 in the distribution.
4612 52. Added support for SRV lookups to the dnslookup router.
4614 53. If an ACL referred to $message_body or $message_body_end, the value was not
4615 reset for any messages that followed in the same SMTP session.
4617 54. The store-handling optimization for building very long strings was not
4618 differentiating between the different store pools. I don't think this
4619 actually made any difference in practice, but I've tidied it.
4621 55. While running the routers to verify a sender address, $sender_address
4622 was still set to the sender address. This is wrong, because when routing to
4623 send a bounce to the sender, it would be empty. Therefore, I have changed
4624 it so that, while verifying a sender address, $sender_address is set to <>.
4625 (There is no change to what happens when verifying a recipient address.)
4627 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4628 target A or AAAA records (if not already returned) without resetting the
4629 qualify_single or search_parents options of the DNS resolver. These are
4630 inappropriate in this case because the targets of MX and SRV records must
4631 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4632 target that, when qualified, matched something in the local domain. These
4633 two options are now turned off when doing these lookups.
4635 57. It seems that at least some releases of Reiserfs (which does not have the
4636 concept of a fixed number of inodes) returns zero and not -1 for the
4637 number of available inodes. This interacted badly with check_spool_inodes,
4638 which assumed that -1 was the "no such thing" setting. What I have done is
4639 to check that the total number of inodes is greater than zero before doing
4640 the test of how many are available.
4642 58. When a "warn" ACL statement has a log_message modifier, the message is
4643 remembered, and not repeated. This is to avoid a lot of repetition when a
4644 message has many recipients that cause the same warning to be written.
4645 Howewer, Exim was preserving the list of already written lines for an
4646 entire SMTP session, which doesn't seem right. The memory is now reset if a
4647 new message is started.
4649 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4650 debugging output unless log_rewrite was also set.
4652 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4653 of (int)(handle) when we know that handle contains (void *)(-1).
4655 61. The Exim daemon panic-logs an error return when it closes the incoming
4656 connection. However "connection reset by peer" seems to be common, and
4657 isn't really an error worthy of noting specially, so that particular error
4660 62. When Exim is trying to find all the local interfaces, it used to panic and
4661 die if the ioctl to get the interface flags failed. However, it seems that
4662 on at least one OS (Solaris 9) it is possible to have an interface that is
4663 included in the list of interfaces, but for which you get a failure error
4664 for this call. This happens when the interface is not "plumbed" into a
4665 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4666 failure of the "get flags" call assumes that the interface is down.
4668 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4669 makes life easier for people who are doing arithmetic on fields extracted
4670 from dates, where you often get leading zeros that should not be
4671 interpreted as octal.
4673 64. Added qualify_domain to the redirect router, to override the global
4676 65. If a pathologically long header line contained very many addresses (the
4677 report of this problem mentioned 10 000) and each of them was rewritten,
4678 Exim could use up a very large amount of memory. (It kept on making new
4679 copies of the header line as it rewrote, and never released the old ones.)
4680 At the expense of a bit more processing, the header rewriting function has
4681 been changed so that it no longer eats memory in this way.
4683 66. The generation of the Received: header has been moved from the time that a
4684 message starts to be received, to the time that it finishes. The timestamp
4685 in the Received: header should now be very close to that of the <= log
4686 line. There are two side-effects of this change:
4688 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4689 logged header lines no longer include the local Received: line, because
4690 it has not yet been created. The same applies to a copy of the message
4691 that is returned to a non-SMTP sender when a message is rejected.
4693 (b) When a filter file is tested using -bf, no additional Received: header
4694 is added to the test message. After some thought, I decided that this
4697 This change does not affect the value of $received_for. It is still set
4698 after address rewriting, but before local_scan() is called.
4700 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4702 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4703 gave an unhelpful panic error message, and a defer error. I have managed to
4704 change this behaviour so that it now rejects any supplied certificate,
4705 which seems right, as the list of acceptable certificates is empty.
4707 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4708 gave an unhelpful defer error. I have not managed to make this reject any
4709 supplied certificates, but the error message it gives is "no certificate
4710 supplied", which is not helpful.
4712 70. exigrep's output now also includes lines that are not associated with any
4713 message, but which match the given pattern. Implemented by a patch from
4714 Martin Sluka, which also tidied up the Perl a bit.
4716 71. Recipient callout verification, like sender verification, was using <> in
4717 the MAIL FROM command. This isn't really the right thing, since the actual
4718 sender may affect whether the remote host accepts the recipient or not. I
4719 have changed it to use the actual sender in the callout; this means that
4720 the cache record is now keyed on a recipient/sender pair, not just the
4721 recipient address. There doesn't seem to be a real danger of callout loops,
4722 since a callout by the remote host to check the sender would use <>.
4723 [SEE ABOVE: changed after hitting problems.]
4725 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4726 temporary errors. However, in the case of such a code being given after
4727 the end of a data transmission (i.e. after ".") Exim was failing to write
4728 a retry record for the message. (Yes, there was some broken host that was
4729 actually sending 8xx at this point.)
4731 73. An unknown lookup type in a host list could cause Exim to panic-die when
4732 the list was checked. (An example that provoked this was putting <; in the
4733 middle of a list instead of at the start.) If this happened during a DATA
4734 ACL check, a -D file could be left lying around. This kind of configuration
4735 error no longer causes Exim to die; instead it causes a defer errror. The
4736 incident is still logged to the main and panic logs.
4738 74. Buglet left over from Exim 3 conversion. The message "too many messages
4739 in one connection" was written to the rejectlog but not the mainlog, except
4740 when address rewriting (yes!) was being logged.
4742 75. Added write_rejectlog option.
4744 76. When a system filter was run not as root (that is, when system_filter_user
4745 was set), the values of the $n variables were not being returned to the
4746 main process; thus, they were not subsequently available in the $sn
4749 77. Added +return_path_on_delivery log selector.
4751 78. A connection timeout was being treated differently from recipients deferred
4752 when testing hosts_max_try with a message that was older than the host's
4753 retry timeout. (The host should not be counted, thus allowing all hosts to
4754 be tried at least once before bouncing.) This may have been the cause of an
4755 occasionally reported bug whereby a message would remain on the queue
4756 longer than the retry timeout, but would be bounced if a delivery was
4757 forced. I say "may" because I never totally pinned down the problem;
4758 setting up timeout/retry tests is difficult. See also the next item.
4760 79. The ultimate address timeout was not being applied to errors that involved
4761 a combination of host plus message (for example, a timeout on a MAIL
4762 command). When an address resolved to a number of possible hosts, and they
4763 were not all tried for each delivery (e.g. because of hosts_max_try), a
4764 message could remain on the queue longer than the retry timeout.
4766 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4769 81. Fixed an obscure SMTP outgoing bug which required at least the following
4770 conditions: (a) there was another message waiting for the same server;
4771 (b) the server returned 5xx to all RCPT commands in the first message so
4772 that the message was not completed; (c) the server dropped the connection
4773 or gave a negative response to the RSET that Exim sends to abort the
4774 transaction. The observed case was a dropped connection after DATA that had
4775 been sent in pipelining mode. That is, the server had advertised PIPELINING
4776 but was not implementing it correctly. The effect of the bug was incorrect
4777 behaviour, such as trying another host, and this could lead to a crash.
4783 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4784 and daemon.c were passed as pointers to ints; they should have been
4785 pointers to socklen_t variables (which are typically unsigned ints).
4787 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4790 3. Fixed a really odd bug that affected only the testing scheme; patching a
4791 certain fixed string in the binary changed the value of another string that
4792 happened to be identical to the end of the original first string.
4794 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4795 name", it returns that address as the IP address. On some operating
4796 systems (e.g. Solaris), it also passes back the IP address string as the
4797 "host name". However, on others (e.g. Linux), it passes back an empty
4798 string. Exim wasn't checking for this, and was changing the host name to an
4799 empty string, assuming it had been canonicized.
4801 5. Although rare, it is permitted to have more than one PTR record for a given
4802 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4803 all the names associated with an address, because they do in Solaris.
4804 However, it seems that they do not in Linux for data that comes from the
4805 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4806 I found this out when I moved to a new Linux workstation and tried to run
4807 the Exim test suite.
4809 To get round this problem I have changed the code so that it now does its
4810 own call to the DNS to look up PTR records when searching for a host name.
4811 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4812 addresses that are only in /etc/hosts are still found.
4814 This behaviour is, however, controlled by an option called host_lookup_
4815 order, which defaults to "bydns:byaddr". If people want to use the other
4816 order, or indeed, just use one or the other means of lookup, they can
4817 specify it in this variable.
4819 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4820 some operating systems, this comes back from gethostbyaddr() as an empty
4821 string, and this is what Exim used to test for. However, it seems that in
4822 other systems, "." is yielded. Exim now tests for this case too.
4824 7. The values of check_spool_space and check_log_space are now held internally
4825 as a number of kilobytes instead of an absolute number of bytes. If a
4826 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4827 kilobyte. This means that much larger values can be stored.
4829 8. Exim monitor: an attempt to get the action menu when not actually pointing
4830 at a message produces an empty menu entitled "No message selected". This
4831 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4832 no entries in it ("Shell widget menu has zero width and/or height"). So I
4833 have added a single, blank menu entry in this case.
4835 9. Added ${quote_local_part.
4837 10. MIME decoding is now applied to the contents of Subject: header lines when
4840 11. Now that a reference to $sender_host_address automatically causes a reverse
4841 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4842 host lookup before query-style lookups in lists that might use this
4843 variable. This has therefore been abolished, and the "net-" prefix is no
4844 longer necessary for query-style lookups.
4846 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4847 have been a typo for LFLAGS, so it has been changed.
4849 13. The install script calls Exim with "-C /dev/null" in order to find the
4850 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4851 to be output. Howeve, since Exim outputs its version number before the
4852 error, it didn't break the script. It just looked ugly. I fixed this by
4853 always allowing "-C /dev/null" if the caller is root.
4855 14. Ignore overlarge ACL variable number when reading spool file - insurance
4856 against a later release with more variables having written the file.
4858 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4859 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4860 this, as well as the form without the "IPv6" on the front.
4862 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4863 OS/Makefile-Darwin file.
4865 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4866 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4868 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4869 variable. However, some OS use ints for inodes. I've added cast to long int
4870 to get rid of the compiler warning.
4872 19. I had forgotten to lock out "/../" in configuration file names when
4873 ALT_CONFIG_PREFIX was set.
4875 20. Routers used for verification do not need to specify transports. However,
4876 if such a router generated a host list, and callout was configured, Exim
4877 crashed, because it could not find a port number from the (non-existent)
4878 transport. It now assumes port 25 in this circumstance.
4880 21. Added the -t option to exigrep.
4882 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4883 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4884 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4885 them is set, LOOKUP_LSEARCH is forced.
4887 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4889 24. Added sender and host information to the "rejected by local_scan()" log
4890 line; previously there was no indication of these.
4892 25. Added .include_if_exists.
4894 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4895 Linux. It turns out that not all file systems support this. Apparently some
4896 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4897 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4898 not supported on the file descriptor, is now ignored when Exim is trying to
4899 sync a directory. This applies only to Linux.
4901 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4903 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4904 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4907 29. Added support for maildirsize files from supplied patch (modified a bit).
4909 30. The use of :fail: followed by an empty string could lead Exim to respond to
4910 sender verification failures with (e.g.):
4912 550 Verification failed for <xxx>
4913 550 Sender verify failed
4915 where the first response line was missing the '-' that indicates it is not
4916 the final line of the response.
4918 31. The loop for finding the name of the user that called Exim had a hardwired
4919 limit of 10; it now uses the value of finduser_retries, which is used for
4920 all other user lookups.
4922 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4925 33. Exim was neglecting to zero errno before one call of strtol() when
4926 expanding a string and expecting an integer value. On some systems this
4927 resulted in spurious "integer overflow" errors. Also, it was casting the
4928 result into an int without checking.
4930 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4931 did not work. The code looks as if it has *never* worked, though it appears
4932 to have been documented since at least releast 1.62. I have made it work.
4934 35. The "timeout_DNS" error in retry rules, also documented since at least
4935 1.62, also never worked. As it isn't clear exactly what this means, and
4936 clearly it isn't a major issue, I have abolished the feature by treating it
4937 as "timeout", and writing a warning to the main and panic logs.
4939 36. The display of retry rules for -brt wasn't always showing the error code
4942 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4943 timeout_connect_A, timeout_connect_MX.
4945 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4946 to the empty sender.
4948 39. The daemon was not analysing the content of -oX till after it had closed
4949 stderr and disconnected from the controlling terminal. This meant that any
4950 syntax errors were only noted on the panic log, and the return code from
4951 the command was 0. By re-arranging the code a little, I've made the
4952 decoding happen first, so such errors now appear on stderr, and the return
4953 code is 1. However, the actual setting up of the sockets still happens in
4954 the disconnected process, so errors there are still only recorded on the
4957 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4958 connections (as happens on some IP stacks) was logged at start up time as
4959 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4960 it from "IPv6 and IPv4", which means that two separate sockets are being
4963 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4964 says whether AF_INET or AF_INET6 was passed as an argument.
4966 42. Exiwhat output was messed up when time zones were included in log
4969 43. Exiwhat now gives more information about the daemon's listening ports,
4970 and whether -tls-on-connect was used.
4972 44. The "port" option of the smtp transport is now expanded.
4974 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4975 silently ignored. Now an error message is written to the main and panic
4978 46. There's a new ACL modifier called "logwrite" which writes to a log file
4979 as soon as it is encountered.
4981 47. Added $local_user_uid and $local_user_gid at routing time.
4983 48. Exim crashed when trying to verify a sender address that was being
4986 49. Exim was recognizing only a space character after ".include". It now also
4987 recognizes a tab character.
4989 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4990 extracting the relevant information from the specification. The man page no
4991 longer contains scrambled data for the -d option, and I've added a section
4992 at the front about calling Exim under different names.
4994 51. Added "extra_headers" argument to the "mail" command in filter files.
4996 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4999 53. Installed eximstats 1.29.
5001 54. Added transport_filter_timeout as a generic transport option.
5003 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5004 Cc: header lines. This was required by RFC 822, but it not required by RFC
5007 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5008 incoming messages that did not have them. Now it does so only if the
5009 message originates locally, that is, if there is no associated remote host
5010 address. When Resent- header lines are present, this applies to the Resent-
5011 lines rather than the non-Resent- lines.
5013 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5014 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5016 58. Messages for configuration errors now include the name of the main
5017 configuration file - useful now that there may be more than one file in a
5018 list (.included file names were always shown).
5020 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5021 for those rare installations that do not start the daemon as root or run it
5022 setuid root. I've cut out the call to initgroups() if the daemon is not
5025 60. The Exim user and group can now be bound into the binary as text strings
5026 that are looked up at the start of Exim's processing.
5028 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5030 62. Added $mailstore_basename variable.
5032 63. Installed patch to sieve.c from Michael Haardt.
5034 64. When Exim failed to open the panic log after failing to open the main log,
5035 the original message it was trying to log was written to stderr and debug
5036 output, but if they were not available (the usual case in production), it
5037 was lost. Now it is written to syslog before the two lines that record the
5038 failures to open the logs.
5040 65. Users' Exim filters run in subprocesses under the user's uid. It is
5041 possible for a "deliver" command or an alias in a "personal" command to
5042 provoke an address rewrite. If logging of address rewriting is configured,
5043 this fails because the process is not running as root or exim. There may be
5044 a better way of dealing with this, but for the moment (because 4.30 needs
5045 to be released), I have disabled address rewrite logging when running a
5046 filter in a non-root, non-exim process.
5052 1. The buildconfig auxiliary program wasn't quoting the value set for
5053 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5054 not defined. This bug was masked in 4.22 by the effect that was fixed in
5057 2. Some messages that were rejected after a message id was allocated were
5058 shown as "incomplete" by exigrep. It no longer does this for messages that
5059 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5061 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5062 have allow_domain_literals set, the ID did not get logged in the <= line.
5063 Domain literals are now always recognized in Message-ID: header lines.
5065 4. The first argument for a ${extract expansion item is the key name or field
5066 number. Leading and trailing spaces in this item were not being ignored,
5067 causing some misleading effects.
5069 5. When deliver_drop_privilege was set, single queue runner processes started
5070 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5071 same command in the process it spins off) were not dropping privilege.
5073 6. When the daemon running as "exim" started a queue runner, it always
5074 re-executed Exim in the spun-off process. This is a waste of effort when
5075 deliver_drop_privilege is set. The new process now just calls the
5076 queue-runner function directly.
5082 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5085 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5086 ignored. Though the use of -oP was forcing the writing of a pid file, it
5087 was always written to the default place.
5089 3. If the message "no IP address found for host xxxx" is generated during
5090 incoming verification, it is now followed by identification of the incoming
5091 connection (so you can more easily find what provoked it).
5093 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5095 5. Added some features to "harden" Exim a bit more against certain attacks:
5097 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5098 be put in Local/Makefile. This is like the never_users runtime option,
5099 but it cannot be overridden. The default setting is "root".
5101 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5102 prefix string with which any file named in a -C command line option
5105 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5106 is retained for -C and -D only if the caller of Exim is root. Without
5107 it, the exim user may also use -C and -D and retain privilege.
5109 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5110 command line option is disabled.
5112 6. Macro names set by the -D option must start with an upper case letter, just
5113 like macro names defined in the configuration file.
5115 7. Added "dereference=" facility to LDAP.
5117 8. Two instances of the typo "uknown" in the source files are fixed.
5119 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5120 the Configure-Makefile script screwed up while processing it.
5122 10. Incorporated PCRE 4.4.
5124 11. The SMTP synchronization check was not operating right at the start of an
5125 SMTP session. For example, it could not catch a HELO sent before the client
5126 waited for the greeting. There is now a check for outstanding input at the
5127 point when the greeting is written. Because of the duplex, asynchronous
5128 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5129 way, but not yet received, when the check is performed.
5131 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5132 on TCP/IP sockets, because this apparently causes some broken clients to
5135 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5136 unchanged) from the Cygwin maintainer.
5138 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5139 was supported instead of testing for mbx. Effectively a typo.
5141 15. The spa authenticator server code was not checking that the input it
5142 received was valid base64.
5144 16. The debug output line for the "set" modifier in ACLs was not showing the
5145 name of the variable that was being set.
5147 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5149 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5150 lookup on its own. Something else was needed to trigger the lookup. For
5151 example, a match in host_lookup or the need for a host name in a host list.
5152 Now, if $sender_host_name is referenced and the host name has not yet been
5153 looked up, a lookup is performed. If the lookup fails, the variable remains
5154 empty, and $host_lookup_failed is set to "1".
5156 19. Added "eqi" as a case-independent comparison operator.
5158 20. The saslauthd authentication condition could segfault if neither service
5159 nor realm was specified.
5161 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5162 error message that was logged was misleading, and incoming SMTP
5163 connections were dropped. The message is now more accurate, and temporary
5164 errors are given to SMTP connections.
5166 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5167 (except RSET) with a 421 error, until QUIT is received. However, it was
5168 failing to send a response to QUIT.
5170 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5171 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5172 ACL is now run after the helo_try_verify_hosts code.
5174 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5175 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5176 case-independent, so other case variants are also recognized). Apparently
5177 some systems use these upper case variants.
5179 25. If more than two messages were waiting for the same host, and a transport
5180 filter was specified for the transport, Exim sent two messages over the
5181 same TCP/IP connection, and then failed with "socket operation on non-
5182 socket" when it tried to send the third.
5184 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5186 27. The extern definition of crypt16() in expand.c was not being excluded when
5187 the OS had its own crypt16() function.
5189 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5190 as a preferred synonym for return_size_limit, both as an option and as an
5193 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5195 30. Changed the default configuration ACL to relax the local part checking rule
5196 for addresses that are not in any local domains. For these addresses,
5197 slashes and pipe symbols are allowed within local parts, but the sequence
5198 /../ is explicitly forbidden.
5200 31. SPA server authentication was not clearing the challenge buffer before
5203 32. log_message in a "warn" ACL statement was writing to the reject log as
5204 well as to the main log, which contradicts the documentation and doesn't
5205 seem right (because no rejection is happening). So I have stopped it.
5207 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5208 However, I am unable to do any testing of this.
5210 34. Fixed an infelicity in the appendfile transport. When checking directories
5211 for a mailbox, to see if any needed to be created, it was accidentally
5212 using path names with one or more superfluous leading slashes; tracing
5213 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5215 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5216 contents are added to the log line that is written for every discarded
5217 recipient. (Previously a log_message setting was ignored.)
5219 36. The ${quote: operator now quotes the string if it is empty.
5221 37. The install script runs exim in order to find its version number. If for
5222 some reason other than non-existence or emptiness, which it checks, it
5223 could not run './exim', it was installing it with an empty version number,
5224 i.e. as "exim-". This error state is now caught, and the installation is
5227 38. An argument was missing from the function that creates an error message
5228 when Exim fails to connect to the socket for saslauthd authentication.
5229 This could cause Exim to crash, or give a corrupted message.
5231 39. Added isip, isip4, and isip6 to ${if conditions.
5233 40. The ACL variables $acl_xx are now saved with the message, and can be
5234 accessed later in routers, transports, and filters.
5236 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5237 strings in the file are not string-expanded.
5239 42. If a MAIL command specified a SIZE value that was too large to fit into an
5240 int variable, the check against message_size_limit failed. Such values are
5241 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5242 day this will have to be increased, but I don't think I want to be around
5243 when emails are that large.
5250 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5251 iconv() is not standard in FreeBSD.
5253 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5254 IPv6 enabled. The observed symptom was a segmentation fault on return from
5255 the function os_common_find_running_interfaces() in src/os.c.
5257 3. In the check_special_case() function in daemon.c I had used "errno" as an
5258 argument name, which causes warnings on some systems. This was basically a
5259 typo, since it was named "eno" in the comments!
5261 4. The code that waits for the clock to tick (at a resolution of some fraction
5262 of a second) so as to ensure message-id uniqueness was always waiting for
5263 at least one whole tick, when it could have waited for less. [This is
5264 almost certainly not relevant at current processor speeds, where it is
5265 unlikely to ever wait at all. But we try to future-proof.]
5267 5. The function that sleeps for a time interval that includes fractions of a
5268 second contained a race. It did not block SIGALRM between setting the
5269 timer, and suspending (a couple of lines later). If the interval was short
5270 and the sigsuspend() was delayed until after it had expired, the suspension
5271 never ended. On busy systems this could lead to processes getting stuck for
5274 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5275 process, before it forks any delivery processes. The open lookup caching
5276 mechanism meant that the open file or database connection was passed into
5277 the delivery process. The problem was that delivery processes always tidy
5278 up cached lookup data. This could cause a problem for the next delivery
5279 process started by the queue runner, because the external queue runner
5280 process does not know about the closure. So the next delivery process
5281 still has data in the lookup cache. In the case of a file lookup, there was
5282 no problem because closing a file descriptor in a subprocess doesn't affect
5283 the parent. However, if the lookup was caching a connection to a database,
5284 the connection was closed, and the second delivery process was likely to
5285 see errors such as "PGSQL: query failed: server closed the connection
5286 unexpectedly". The problem has been fixed by closing all cached lookups
5287 in a queue runner before running a delivery process.
5289 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5290 seem to have the "const" qualifier which it has on other OS. I've
5293 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5294 *of the same type* (client or server) with the same public name that an
5295 error should be diagnosed.
5297 9. When Exim looked up a host name for an IP address, but failed to find the
5298 original IP address when looking up the host name (a safety check), it
5299 output the message "<ip address> does not match any IP for NULL", which was
5300 confusing, to say the least. The bug was that the host name should have
5301 appeared instead of "NULL".
5303 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5304 user that is built into the binary, and the -C or -D options is used, root
5305 privilege is dropped before the configuration file is read. In addition,
5306 logging is switched to stderr instead of the normal log files. If the
5307 configuration then re-defines the Exim user, the unprivileged environment
5308 is probably not what is expected, so Exim logs a panic warning message (but
5311 However, if deliver_drop_privilege is set, the unprivileged state may well
5312 be exactly what is intended, so the warning has been cut out in that case,
5313 and Exim is allowed to try to write to its normal log files.
5319 1. smtp_return_error_details was not giving details for temporary sender
5320 or receiver verification errors.
5322 2. Diagnose a configuration error if two authenticators have the same public
5325 3. Exim used not to create the message log file for a message until the first
5326 delivery attempt. This could be confusing when incoming messages were held
5327 for policy or load reasons. The message log file is now created at the time
5328 the message is received, and an initial "Received" line is written to it.
5330 4. The automatically generated man page for command line options had a minor
5331 bug that caused no ill effects; however, a more serious problem was that
5332 the procedure for building the man page automatically didn't always
5333 operate. Consequently, release 4.20 contains an out-of-date version. This
5334 shouldn't happen again.
5336 5. When building Exim with embedded Perl support, the script that builds the
5337 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5338 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5340 6. The freeze_tell option was not being used for messages that were frozen on
5341 arrival, either by an ACL or by local_scan().
5343 7. Added the smtp_incomplete_transaction log selector.
5345 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5346 was accepting AUTH without a new EHLO.
5348 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5349 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5352 10. Logging of TCP/IP connections (when configured) now happens in the main
5353 daemon process instead of the child process, so that the TCP/IP connection
5354 count is more accurate (but it can never be perfect).
5356 11. The use of "drop" in a nested ACL was not being handled correctly in the
5357 outer ACL. Now, if condition failure induced by the nested "drop" causes
5358 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5359 or "require"), the connection is dropped.
5361 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5362 that yield "discard" can now be used with an "accept" or a "discard" verb,
5363 but an error is generated for any others (because I can't see a useful way
5364 to define what should happen).
5366 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5367 are now processed in the same way as lines in the Exim configuration file.
5368 In particular, continuation lines are supported.
5370 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5372 15. Added -ti meaning -t -i.
5374 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5375 domains, and warn by logging if others are found.
5377 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5378 returned from GIFCONF when Exim is trying to find the list of interfaces on
5379 a host. The code in os.c has been modified to copy each ifreq to an aligned
5380 structure in all cases.
5382 Also, in some cases, the returned ifreq's were being copied to a 'struct
5383 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5384 means the last couple of bytes of an IPv6 address could be chopped if the
5385 ifreq contained only a normal sockaddr (14 bytes storage).
5387 18. Named domain lists were not supported in the hosts_treat_as_local option.
5388 An entry such as +xxxx was not recognized, and was treated as a literal
5391 19. Ensure that header lines added by a DATA ACL are included in the reject log
5392 if the ACL subsequently rejects the message.
5394 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5395 MD5 (which is deprecated).
5397 21. When testing a filter file using -bf, Exim was writing a message when it
5398 took the sender from a "From " line in the message, but it was not doing so
5399 when it took $return_path from a Return-Path: header line. It now does.
5401 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5402 with a valid header line field name (a series of printing characters
5403 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5405 23. Changed "disc" in the source to "disk" to conform to the documentation and
5406 the book and for uniformity.
5408 24. Ignore Sendmail's -Ooption=value command line item.
5410 25. When execve() failed while trying to run a command in a pipe transport,
5411 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5412 could be confused with a return value of 69 from the command itself. This
5413 has been changed to 127, the value the shell returns if it is asked to run
5414 a non-existent command. The wording for the related log line suggests a
5415 non-existent command as the problem.
5417 26. If received_header_text expands to an empty string, do not add a Received:
5418 header line to the message. (Well, it adds a token one on the spool, but
5419 marks it "old" so that it doesn't get used or transmitted.)
5421 27. Installed eximstats 1.28 (addition of -nt option).
5423 28. There was no check for failure on the call to getsockname() in the daemon
5424 code. This can fail if there is a shortage of resources on the system, with
5425 ENOMEM, for example. A temporary error is now given on failure.
5427 29. Contrary to the C standard, it seems that in some environments, the
5428 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5429 program. Exim now does this explicitly; it affects the formatting of
5430 timestamps using strftime().
5432 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5433 complaints. I've now initialized all the variables, to avoid this.
5435 32. Header lines added by a system filter were not being "seen" during
5436 transport-time rewrites.
5438 33. The info_callback() function passed to OpenSSL is set up with type void
5439 (*)(SSL *, int, int), as described somewhere. However, when calling the
5440 function (actually a macro) that sets it up, the type void(*)() is
5441 expected. I've put in a cast to prevent warnings from picky compilers.
5443 34. If a DNS black list lookup found a CNAME record, but there were no A
5444 records associated with the domain it pointed at, Exim crashed.
5446 35. If a DNS black list lookup returned more than one A record, Exim ignored
5447 all but the first. It now scans all returned addresses if a particular IP
5448 value is being sought. In this situation, the contents of the
5449 $dnslist_value variable are a list of all the addresses, separated by a
5452 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5453 to accept a host name and all its aliases if the forward lookup for any of
5454 them yielded the IP address of the incoming connection. Now it accepts only
5455 those names whose forward lookup yields the correct IP address. Any other
5456 names are discarded. This closes a loophole whereby a rogue DNS
5457 administrator could create reverse DNS records to break through a
5458 wildcarded host restriction in an ACL.
5460 37. If a user filter or a system filter that ran in a subprocess used any of
5461 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5462 the wrong values were passed to the pipe command ($thisaddress had the
5463 value of $0, $0 had the value of $1, etc). This bug was introduced by
5464 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5466 38. Improved the line breaking for long SMTP error messages from ACLs.
5467 Previously, if there was no break point between 40 and 75 characters, Exim
5468 left the rest of the message alone. Two changes have been made: (a) I've
5469 reduced the minimum length to 35 characters; (b) if it can't find a break
5470 point between 35 and 75 characters, it looks ahead and uses the first one
5471 that it finds. This may give the occasional overlong line, but at least the
5472 remaining text gets split now.
5474 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5475 file descriptors might be low, and that setting 1000 would always raise it.
5476 It turns out that in some environments, the limit is already over 1000 and
5477 that lowering it causes trouble. So now Exim takes care not to decrease it.
5479 40. When delivering a message, the value of $return_path is set to $sender_
5480 address at the start of routing (routers may change the value). By an
5481 oversight, this default was not being set up when an address was tested by
5482 -bt or -bv, which affected the outcome if any router or filter referred to
5485 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5486 log only when "message" or "log_message" are set. However, if one of the
5487 conditions was an address verification, or a call to a nested ACL, the
5488 messages generated by the underlying test were being passed through. This
5489 no longer happens. The underlying message is available in $acl_verify_
5490 message for both "message" and "log_message" expansions, so it can be
5491 passed through if needed.
5493 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5494 new expansion $bh_ to give the encoded byte string without charset
5495 translation. Translation happens only if iconv() is available; HAVE_ICONV
5496 indicates this at build time. HEADERS_CHARSET gives the charset to
5497 translate to; headers_charset can change it in the configuration, and
5498 "headers charset" can change it in an individual filter file.
5500 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5501 that creates RFC 2047 encoded "words" labels them as that charset instead
5502 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5503 expansion operator; (ii) when Exim creates a From: line for a local
5504 message; (iii) when a header line is rewritten to include a "phrase" part.
5506 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5507 buggy, causing it to skip the first lines of messages whose message ID
5508 ended in 'D'. This would not have bitten before Exim release 4.14, because
5509 message IDs were unlikely to end in 'D' before then. The effect was to have
5510 incorrect size information for certain domains.
5512 45. #include "config.h" was missing at the start of the crypt16.c module. This
5513 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5516 46. If there was a timeout during a "random" callout check, Exim treated it as
5517 a failure of the random address, and carried on sending RSET and the real
5518 address. If the delay was just some slowness somewhere, the response to the
5519 original RCPT would be taken as a response to RSET and so on, causing
5520 mayhem of various kinds.
5522 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5523 when I implemented it. It didn't allow for the fact that some option values
5524 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5525 the right test for positive values.
5527 48. Domain names in DNS records are case-independent. Exim always looks them up
5528 in lower case. Some resolvers return domain names in exactly the case they
5529 appear in the zone file, that is, they may contain uppercase letters. Not
5530 all resolvers do this - some return always lower case. Exim was treating a
5531 change of case by a resolver as a change of domain, similar to a widening
5532 of a domain abbreviation. This triggered its re-routing code and so it was
5533 trying to route what was effectively the same domain again. This normally
5534 caused routing to fail (because the router wouldn't handle the domain
5535 twice). Now Exim checks for this case specially, and just changes the
5536 casing of the domain that it ultimately uses when it transmits the message
5539 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5542 50. If a filter generated a file delivery with a non-absolute name (possible if
5543 no home directory exists for the router), the forbid_file option was not
5546 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5547 the existing equality matching.
5549 52. Exim was using ints instead of ino_t variables in some places where it was
5550 dealing with inode numbers.
5552 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5553 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5554 called TMPDIR, and if it finds it is different, it changes its value.
5556 54. The smtp_printf() function is now made available to local_scan() so
5557 additional output lines can be written before returning. There is also an
5558 smtp_fflush() function to enable the detection of a dropped connection.
5559 The variables smtp_input and smtp_batched_input are exported to
5562 55. Changed the default runtime configuration: the message "Unknown user"
5563 has been removed from the ACL, and instead placed on the localuser router,
5564 using the cannot_route_message feature. This means that any verification
5565 failures that generate their own messages won't get overridden. Similarly,
5566 the "Unrouteable address" message that was in the ACL for unverifiable
5567 relay addresses has also been removed.
5569 56. Added hosts_avoid_esmtp to the smtp transport.
5571 57. The exicyclog script was not checking for the esoteric option
5572 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5573 will work only if exicyclog is run under the appropriate euid.
5575 58. Following a discussion on the list, the rules by which Exim recognises line
5576 endings on incoming messages have been changed. The -dropcr and drop_cr
5577 options are now no-ops, retained only for backwards compatibility. The
5578 following line terminators are recognized: LF CRLF CR. However, special
5579 processing applies to CR:
5581 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5582 nor a local message in the state where . is a terminator.
5584 (ii) If a bare CR is encountered in a header line, an extra space is added
5585 after the line terminator so as not to end the header. The reasoning
5586 behind this is that bare CRs in header lines are most likely either
5587 to be mistakes, or people trying to play silly games.
5589 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5590 was being incorrectly given as 18 bytes larger than it should have been.
5591 This is a VOB (very old bug).
5593 60. This may never have affected anything current, but just in case it has:
5594 When the local host is found other than at the start of a list of hosts,
5595 the local host, those with the same MX, and any that follow, are discarded.
5596 When the list in question was part of a longer list of hosts, the following
5597 hosts (not currently being processed) were also being discarded. This no
5598 longer happens. I'm not sure if this situation could ever has previously
5601 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5604 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5605 header. This is something that is recommended in a new Internet Draft, and
5606 is something that is documented as being done by Sendmail. There are two
5607 possible values. For messages generated by the autoreply transport, Exim
5610 Auto-Submitted: auto-replied
5612 whereas for all other generated messages (e.g. bounces) it adds
5614 Auto-Submitted: auto-generated
5616 63. The "personal" condition in filters now includes a test for the
5617 Auto-Submitted: header. If it contains the string "auto-" the message it
5618 not considered personal.
5620 64. Added rcpt_include_affixes as a generic transport option.
5622 65. Added queue_only_override (default true).
5624 66. Added the syslog_duplication option.
5626 67. If what should have been the first header line of a message consisted of
5627 a space followed by a colon, Exim was mis-interpreting it as a header line.
5628 It isn't of course - it is syntactically invalid and should therefore be
5629 treated as the start of the message body. The misbehaviour could have
5630 caused a number of strange effects, including loss of data in subsequent
5631 header lines, and spool format errors.
5633 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5634 client host had authenticated. This control can now be exercised by an ACL
5635 for more flexibility.
5637 69. By default, callouts do not happen when testing with -bh. There is now a
5638 variant, -bhc, which does actually run the callout code, including
5639 consulting and updating the callout cache.
5641 70. Added support for saslauthd authentication, courtesy of Alexander
5644 71. If statvfs() failed on the spool or log directories while checking their
5645 size for availability, Exim confusingly gave the error "space shortage".
5646 Furthermore, in debugging mode it crashed with a floating point exception.
5647 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5648 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5649 serious problem, Exim now writes to the main and panic logs when this
5650 happens, with details of the failure. It then refuses to accept the
5651 incoming message, giving the message "spool directory problem" or "log
5652 directory problem" with a 421 code for SMTP messages.
5654 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5655 0, 1, and 2 exist, because some OS complain for execs without them (see
5656 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5657 descriptors. However, the code omitted to check that the open succeeded,
5658 causing mysterious errors if for some reason the permissions on /dev/null
5659 got screwed. Now Exim writes a message to the main and panic logs, and
5660 bombs out if it can't open /dev/null.
5662 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5663 interact so that it is all more flexible. It is supposed to remain
5664 backwards compatible. Also added extra_local_interfaces.
5666 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5667 to bomb out with an assertion failure - to the client this appears as a
5668 connection drop. This problem occurs in the part of the code that was taken
5669 from the Samba project. Fortunately, the assertion is in a very simple
5670 function, so I have fixed this by reproducing the function inline in the
5671 one place where it is called, and arranging for authentication to fail
5672 instead of killing the process with assert().
5674 75. The SPA client code was not working when the server requested OEM rather
5675 than Unicode encoding.
5677 76. Added code to make require_files with a specific uid setting more usable in
5678 the case where statting the file as root fails - usually a non-root-mounted
5679 NFS file system. When this happens and the failure is EACCES, Exim now
5680 forks a subprocess and does the per-uid checking as the relevant uid.
5682 77. Added process_log_path.
5684 78. If log_file_path was not explicitly set, a setting of check_log_space or
5685 check_log_inodes was ignored.
5687 79. If a space check for the spool or log partitions fails, the incident is now
5688 logged. Of course, in the latter case the data may get lost...
5690 80. Added the %p formatting code to string_format() so that it can be used to
5691 print addresses in debug_print(). Adjusted all the address printing in the
5692 debugging in store.c to use %p rather than %d.
5694 81. There was a concern that a line of code in smtp_in.c could overflow a
5695 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5696 initially expressed, the concern was not well-founded, because trailing
5697 spaces are removed early. However, if the trailing spaces were followed by
5698 a NULL, they did not get removed, so the overflow was possible. Two fixes
5701 (a) I re-wrote the offending code in a cleaner fashion.
5702 (b) If an incoming SMTP command contains a NULL character, it is rejected
5705 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5706 runs initgroups(), so that if the Exim user is in any additional groups,
5707 they will be used during message reception.
5713 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob