1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Handle short writes when writing local log-files.
8 In practice, only affects FreeBSD (8 onwards).
9 Bugzilla 1053, with thanks to Dmitry Isaikin.
11 NM/01 Bugzilla 949 - Documentation tweak
13 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
16 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
18 PP/02 Implemented gsasl authenticator.
20 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
22 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
23 `pkg-config foo` for cflags/libs.
25 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
26 with rest of GSASL and with heimdal_gssapi.
28 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
29 `pkg-config foo` for cflags/libs for the TLS implementation.
31 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
32 properties get this fed in as external SSF. A number of robustness
33 and debugging improvements to the cyrus_sasl authenticator.
35 PP/08 cyrus_sasl server now expands the server_realm option.
37 PP/09 Bugzilla 1214 - Log authentication information in reject log.
38 Patch by Jeremy Harris.
40 PP/10 Added dbmjz lookup type.
42 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
44 PP/12 MAIL args handles TAB as well as SP, for better interop with
45 non-compliant senders.
46 Analysis and variant patch by Todd Lyons.
48 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
49 Bug report from Lars Müller <lars@samba.org> (via SUSE),
50 Patch from Dirk Mueller <dmueller@suse.com>
52 PP/13 tls_peerdn now print-escaped for spool files.
53 Observed some $tls_peerdn in wild which contained \n, which resulted
54 in spool file corruption.
56 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
57 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
58 or write after TLS renegotiation, which otherwise led to messages
61 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
62 as a tracking header (ie: a signed header comes before the signature).
63 Patch from Wolfgang Breyha.
65 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
66 comma-sep list; embedded commas doubled.
68 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
70 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
72 Report and patch from Dmitry Banschikov.
74 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
75 Removed SSL_clear() after SSL_new() which led to protocol negotiation
76 failures. We appear to now support TLS1.1+ with Exim.
78 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
79 lets Exim select keys and certificates based upon TLS SNI from client.
80 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
81 before an outbound SMTP session. New log_selector, +tls_sni.
83 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
84 NULL dereference. Report and patch from Alun Jones.
86 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
87 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
88 Not seeing resolver debug output on NetBSD, but suspect this is a
89 resolver implementation change.
91 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
92 Left warnings. Added "eximon gdb" invocation mode.
94 PP/21 Defaulting "accept_8bitmime" to true, not false.
96 PP/22 Added -bw for inetd wait mode support.
98 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
99 locate the relevant includes and libraries. Made this the default.
101 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
102 Bugzilla 1246, report and most of solution from Tomasz Kusy.
104 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
105 This may cause build issues on older platforms.
107 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
108 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
109 gnutls_require_mac & gnutls_require_protocols (no longer supported).
110 Added SNI support via GnuTLS too.
111 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
113 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
115 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
116 Applied second patch from Janne, implementing suggestion to default
117 multiple-strings-in-record handling to match SPF spec.
119 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
121 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
122 read-only, out of scope).
123 Patch from Wolfgang Breyha, report from Stuart Northfield.
125 PP/29 Fix three issues highlighted by clang analyser static analysis.
126 Only crash-plausible issue would require the Cambridge-specific
127 iplookup router and a misconfiguration.
128 Report from Marcin Mirosław.
130 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
132 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
133 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
134 As part of this, removing so much warning spew let me fix some minor
135 real issues in debug logging.
137 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
138 assignment on my part. Fixed.
140 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
141 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
142 Janne Snabb (who went above and beyond: thank you).
144 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
145 string otherwise requires a connection and a bunch more work and it's
146 relatively easy to get wrong. Should also expose TLS library linkage
149 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
150 64-bit ${eval} (JH/03).
152 PP/36 Define _ISOC99_SOURCE in exim.h; it's needed for some releases of
153 GNU libc to support some of the 64-bit stuff, should not lead to
154 conflicts. Defined before os.h is pulled in, so if a given platform
155 needs to override this, it can.
161 PP/01 Solaris build fix for Oracle's LDAP libraries.
162 Bugzilla 1109, patch from Stephen Usher.
164 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
166 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
169 TF/02 Fix a couple more cases where we did not log the error message
170 when unlink() failed. See also change 4.74-TF/03.
172 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
173 lock up or crash if it happened to be inside a call to libc when it
174 got a SIGUSR1 from exiwhat.
176 The SIGUSR1 handler appends the current process status to the process
177 log which is later printed by exiwhat. It used to use the general
178 purpose logging code to do this, but several functions it calls are
179 not safe for signals.
181 The new output code in the SIGUSR1 handler is specific to the process
182 log, and simple enough that it's easy to inspect for signal safety.
183 Removing some special cases also simplifies the general logging code.
184 Removing the spurious timestamps from the process log simplifies
187 TF/04 Improved ratelimit ACL condition.
189 The /noupdate option has been deprecated in favour of /readonly which
190 has clearer semantics. The /leaky, /strict, and /readonly update modes
191 are mutually exclusive. The update mode is no longer included in the
192 database key; it just determines when the database is updated. (This
193 means that when you upgrde Exim will forget old rate measurements.)
195 Exim now checks that the per_* options are used with an update mode that
196 makes sense for the current ACL. For example, when Exim is processing a
197 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
198 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
199 must specify per_mail/readonly. If you omit the update mode it defaults to
200 /leaky where that makes sense (as before) or /readonly where required.
202 The /noupdate option is now undocumented but still supported for
203 backwards compatibility. It is equivalent to /readonly except that in
204 ACLs where /readonly is required you may specify /leaky/noupdate or
205 /strict/noupdate which are treated the same as /readonly.
207 A useful new feature is the /count= option. This is a generalization
208 of the per_byte option, so that you can measure the throughput of other
209 aggregate values. For example, the per_byte option is now equivalent
210 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
212 The per_rcpt option has been generalized using the /count= mechanism
213 (though it's more complicated than the per_byte equivalence). When it is
214 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
215 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
216 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
217 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
218 Note that using per_rcpt with a non-readonly update mode in more than
219 one ACL will cause the recipients to be double-counted. (The per_mail
220 and per_byte options don't have this problem.)
222 The handling of very low rates has changed slightly. If the computed rate
223 is less than the event's count (usually one) then this event is the first
224 after a long gap. In this case the rate is set to the same as this event's
225 count, so that the first message of a spam run is counted properly.
227 The major new feature is a mechanism for counting the rate of unique
228 events. The new per_addr option counts the number of different
229 recipients that someone has sent messages to in the last time period. It
230 behaves like per_rcpt if all the recipient addresses are different, but
231 duplicate recipient addresses do not increase the measured rate. Like
232 the /count= option this is a general mechanism, so the per_addr option
233 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
234 example, measure the rate that a client uses different sender addresses
235 with the options per_mail/unique=$sender_address. There are further
236 details in the main documentation.
238 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
240 TF/06 Removed a few PCRE remnants.
242 TF/07 Automatically extract Exim's version number from tags in the git
243 repository when doing development or release builds.
245 PP/02 Raise smtp_cmd_buffer_size to 16kB.
246 Bugzilla 879. Patch from Paul Fisher.
248 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
249 Heavily based on revision 40f9a89a from Simon Arlott's tree.
252 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
254 PP/05 Variable $av_failed, true if the AV scanner deferred.
255 Bugzilla 1078. Patch from John Horne.
257 PP/06 Stop make process more reliably on build failure.
258 Bugzilla 1087. Patch from Heiko Schlittermann.
260 PP/07 Make maildir_use_size_file an _expandable_ boolean.
261 Bugzilla 1089. Patch from Heiko Schlittermann.
263 PP/08 Handle ${run} returning more data than OS pipe buffer size.
264 Bugzilla 1131. Patch from Holger Weiß.
266 PP/09 Handle IPv6 addresses with SPF.
267 Bugzilla 860. Patch from Wolfgang Breyha.
269 PP/10 GnuTLS: support TLS 1.2 & 1.1.
271 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
274 PP/11 match_* no longer expand right-hand-side by default.
275 New compile-time build option, EXPAND_LISTMATCH_RHS.
276 New expansion conditions, "inlist", "inlisti".
278 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
280 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
282 PP/14 fix log_write() format string regression from TF/03.
283 Bugzilla 1152. Patch from Dmitry Isaikin.
289 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
291 PP/02 Harmonised TLS library version reporting; only show if debugging.
292 Layout now matches that introduced for other libraries in 4.74 PP/03.
294 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
296 PP/04 New "dns_use_edns0" global option.
298 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
301 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
302 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
304 TK/01 Updated PolarSSL code to 0.14.2.
305 Bugzilla 1097. Patch from Andreas Metzler.
307 PP/07 Catch divide-by-zero in ${eval:...}.
310 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
313 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
314 format-string attack -- SECURITY: remote arbitrary code execution.
316 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
317 time unintentionally subject to list matching rules, letting the header
318 cause arbitrary Exim lookups (of items which can occur in lists, *not*
319 arbitrary string expansion). This allowed for information disclosure.
321 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
322 INT_MIN/-1 -- value coerced to INT_MAX.
328 NM/01 Workround for PCRE version dependency in version reporting
331 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
332 This fixes portability to compilers other than gcc, notably
333 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
335 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
336 makefiles for portability to HP-UX and POSIX correctness.
338 PP/01 Permit LOOKUP_foo enabling on the make command-line.
339 Also via indented variable definition in the Makefile.
340 (Debugging by Oliver Heesakkers).
342 PP/02 Restore caching of spamd results with expanded spamd_address.
343 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
345 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
346 Improves build reliability. Fix from: Frank Elsner
348 NM/02 Fix wide character breakage in the rfc2047 coding
349 Fixes bug 1064. Patch from Andrey N. Oktyabrski
351 NM/03 Allow underscore in dnslist lookups
352 Fixes bug 1026. Patch from Graeme Fowler
354 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
355 Code patches from Adam Ciarcinski of NetBSD.
357 NM/04 Fixed exiqgrep to cope with mailq missing size issue
360 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
361 is logged, to avoid truncation. Patch from John Horne.
363 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
364 Patch from Jakob Hirsch.
366 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
367 SQL string expansion failure details.
368 Patch from Andrey Oktyabrski.
370 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
371 Patch from Simon Arlott.
373 PP/09 New lookups functionality failed to compile on old gcc which rejects
374 extern declarations in function scope.
375 Patch from Oliver Fleischmann
377 PP/10 Use sig_atomic_t for flags set from signal handlers.
378 Check getgroups() return and improve debugging.
379 Fixed developed for diagnosis in bug 927 (which turned out to be
382 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
383 Patch from Mark Zealey.
385 PP/12 Bugzilla 1056: Improved spamd server selection.
386 Patch from Mark Zealey.
388 PP/13 Bugzilla 1086: Deal with maildir quota file races.
389 Based on patch from Heiko Schlittermann.
391 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
392 Patch from Uwe Doering, sign-off by Michael Haardt.
394 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
395 variable declaration deep within a block. Bug and patch from
398 PP/15 lookups-Makefile IRIX compatibilty coercion.
400 PP/16 Make DISABLE_DKIM build knob functional.
402 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
403 Patch by Simon Arlott
405 TF/03 Fix valgrind.h portability to C89 compilers that do not support
406 variable argument macros. Our copy now differs from upstream.
412 TF/01 Failure to get a lock on a hints database can have serious
413 consequences so log it to the panic log.
415 TF/02 Log LMTP confirmation messages in the same way as SMTP,
416 controlled using the smtp_confirmation log selector.
418 TF/03 Include the error message when we fail to unlink a spool file.
420 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
421 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
422 for maintaining out-of-tree patches for some time.
424 PP/01 Bugzilla 139: Documentation and portability issues.
425 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
426 Handle per-OS dynamic-module compilation flags.
428 PP/02 Let /dev/null have normal permissions.
429 The 4.73 fixes were a little too stringent and complained about the
430 permissions on /dev/null. Exempt it from some checks.
431 Reported by Andreas M. Kirchwitz.
433 PP/03 Report version information for many libraries, including
434 Exim version information for dynamically loaded libraries. Created
435 version.h, now support a version extension string for distributors
436 who patch heavily. Dynamic module ABI change.
438 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
439 privilege escalation vulnerability whereby the Exim run-time user
440 can cause root to append content of the attacker's choosing to
443 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
446 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
447 If dropping privileges for untrusted macros, we disabled normal logging
448 on the basis that it would fail; for the Exim run-time user, this is not
449 the case, and it resulted in successful deliveries going unlogged.
450 Fixed. Reported by Andreas Metzler.
456 PP/01 Date: & Message-Id: revert to normally being appended to a message,
457 only prepend for the Resent-* case. Fixes regression introduced in
458 Exim 4.70 by NM/22 for Bugzilla 607.
460 PP/02 Include check_rfc2047_length in configure.default because we're seeing
461 increasing numbers of administrators be bitten by this.
463 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
465 PP/03 Bugzilla 994: added openssl_options main configuration option.
467 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
469 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
471 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
473 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
474 without a peer certificate, leading to a segfault because of an
475 assumption that peers always have certificates. Be a little more
476 paranoid. Problem reported by Martin Tscholak.
478 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
479 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
480 NB: ClamAV planning to remove STREAM in "middle of 2010".
481 CL also introduces -bmalware, various -d+acl logging additions and
482 more caution in buffer sizes.
484 PP/09 Implemented reverse_ip expansion operator.
486 PP/10 Bugzilla 937: provide a "debug" ACL control.
488 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
490 PP/12 Bugzilla 973: Implement --version.
492 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
494 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
496 PP/15 Bugzilla 816: support multiple condition rules on Routers.
498 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
499 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
500 ignore trailing whitespace.
502 JJ/02 prevent non-panic DKIM error from being sent to paniclog
504 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
507 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
508 Notification from Dr Andrew Aitchison.
510 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
511 ExtendedDetectionInfo response format.
512 Notification from John Horne.
514 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
517 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
518 XSL and documented dependency on system catalogs, with examples of how
521 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
524 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
525 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
526 configuration file which is writeable by the Exim user or group.
528 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
529 of configuration files to cover files specified with the -C option if
530 they are going to be used with root privileges, not just the default
533 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
534 option (effectively making it always true).
536 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
537 files to be used while preserving root privileges.
539 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
540 that rogue child processes cannot use them.
542 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
543 run-time user, instead of root.
545 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
546 Exim run-time user without dropping privileges.
548 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
549 result string, instead of calling string_vformat() twice with the same
552 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
553 for other users. Others should always drop root privileges if they use
554 -C on the command line, even for a whitelisted configure file.
556 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
558 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
564 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
565 $data_path, and $header_path variables; fixed documentation bugs and
568 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
569 exipick to access non-standard spools, including the "frozen" queue
572 NM/01 Bugzilla 965: Support mysql stored procedures.
573 Patch from Alain Williams
575 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
577 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
578 Patch from Andreas Metzler
580 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
581 Patch from Kirill Miazine
583 NM/05 Bugzilla 671: Added umask to procmail example.
585 JJ/03 installed exipick 20100323.0, fixing doc bug
587 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
588 directory. Notification and patch from Dan Rosenberg.
590 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
592 TK/02 Improve log output when DKIM signing operation fails.
594 MH/01 Treat the transport option dkim_domain as a colon separated
595 list, not as a single string, and sign the message with each element,
596 omitting multiple occurences of the same signer.
598 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
599 Bugzilla 985, 986. Patch by Simon Arlott
601 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
602 Patch by Simon Arlott
604 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
605 MBX locking. Notification from Dan Rosenberg.
611 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
613 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
615 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
617 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
619 NM/04 Bugzilla 915: Flag broken perl installation during build.
625 TK/01 Added patch by Johannes Berg that expands the main option
626 "spamd_address" if it starts with a dollar sign.
628 TK/02 Write list of recipients to X-Envelope-Sender header when building
629 the mbox-format spool file for content scanning (suggested by Jakob
632 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
633 (http://www.dcc-servers.net/) support via dccifd. Activated by
634 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
636 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
637 by Mark Daniel Reidel <mr@df.eu>.
639 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
640 When building exim an external PCRE library is now needed -
641 PCRE is a system library on the majority of modern systems.
642 See entry on PCRE_LIBS in EDITME file.
644 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
645 conversation. Added nologin parameter to request.
646 Patch contributed by Kirill Miazine.
648 TF/01 Do not log submission mode rewrites if they do not change the address.
650 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
652 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
653 log files in place. Contributed by Roberto Lima.
655 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
657 TF/03 Bugzilla 615: When checking the local_parts router precondition
658 after a local_part_suffix or local_part_prefix option, Exim now
659 does not use the address's named list lookup cache, since this
660 contains cached lookups for the whole local part.
662 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
663 Robert Millan. Documentation is in experimental-spec.txt.
665 TF/04 Bugzilla 668: Fix parallel build (make -j).
667 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
669 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
670 Patch provided by Jan Srzednicki.
672 TF/05 Leading white space used to be stripped from $spam_report which
673 wrecked the formatting. Now it is preserved.
675 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
676 that they are available at delivery time.
678 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
680 TF/08 TLS error reporting now respects the incoming_interface and
681 incoming_port log selectors.
683 TF/09 Produce a more useful error message if an SMTP transport's hosts
684 setting expands to an empty string.
686 NM/06 Bugzilla 744: EXPN did not work under TLS.
687 Patch provided by Phil Pennock.
689 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
690 Patch provided by Richard Godbee.
692 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
693 acl_smtp_notquit, added index entry.
695 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
696 Patch provided by Eugene Bujak.
698 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
699 accept(). Patch provided by Maxim Dounin.
701 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
702 Patch provided by Phil Pennock.
704 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
706 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
707 Patch provided by Brad "anomie" Jorsch.
709 NM/14 Bugzilla 622: Added timeout setting to transport filter.
710 Patch provided by Dean Brooks.
712 TK/05 Add native DKIM support (does not depend on external libraries).
714 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
715 Patch provided by Graeme Fowler.
717 NM/16 Bugzilla 851: Documentation example syntax fix.
719 NM/17 Changed NOTICE file to remove references to embedded PCRE.
721 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
724 NM/19 Bugzilla 745: TLS version reporting.
725 Patch provided by Phil Pennock.
727 NM/20 Bugzilla 167: bool: condition support.
728 Patch provided by Phil Pennock.
730 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
731 clients. Patch provided by Phil Pennock.
733 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
734 Patch provided by Brad "anomie" Jorsch.
736 NM/23 Bugzilla 687: Fix misparses in eximstats.
737 Patch provided by Heiko Schlittermann.
739 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
740 Patch provided by Heiko Schlittermann.
742 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
743 plus update to original patch.
745 NM/26 Bugzilla 799: Documentation correction for ratelimit.
747 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
748 Patch provided by David Brownlee.
750 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
752 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
754 NM/30 Bugzilla 888: TLS documentation bugfixes.
756 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
758 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
759 Unlike the original bugzilla I have changed all shell scripts in src tree.
761 NM/33 Bugzilla 898: Transport filter timeout fix.
762 Patch by Todd Rinaldo.
764 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
765 Patch by Serge Demonchaux.
767 NM/35 Bugzilla 39: Base64 decode bug fixes.
768 Patch by Jakob Hirsch.
770 NM/36 Bugzilla 909: Correct connect() call in dcc code.
772 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
774 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
776 NM/39 Bugzilla 911: Fixed MakeLinks build script.
782 TK/01 Add preliminary DKIM support. Currently requires a forked version of
783 ALT-N's libdkim that I have put here:
784 http://duncanthrax.net/exim-experimental/
786 Note to Michael Haardt: I had to rename some vars in sieve.c. They
787 were called 'true' and it seems that C99 defines that as a reserved
788 keyword to be used with 'bool' variable types. That means you could
789 not include C99-style headers which use bools without triggering
790 build errors in sieve.c.
792 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
793 as mailq or other aliases. Changed the --help handling significantly
794 to do whats expected. exim_usage() emits usage/help information.
796 SC/01 Added the -bylocaldomain option to eximstats.
798 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
800 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
802 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
808 PH/01 Another patch from the Sieve maintainer.
810 PH/02 When an IPv6 address is converted to a string for single-key lookup
811 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
812 dots are used instead of colons so that keys in lsearch files need not
813 contain colons. This was done some time before quoting was made available
814 in lsearch files. However, iplsearch files do require colons in IPv6 keys
815 (notated using the quote facility) so as to distinguish them from IPv4
816 keys. This meant that lookups for IP addresses in host lists did not work
817 for iplsearch lookups.
819 This has been fixed by arranging for IPv6 addresses to be expressed with
820 colons if the lookup type is iplsearch. This is not incompatible, because
821 previously such lookups could never work.
823 The situation is now rather anomolous, since one *can* have colons in
824 ordinary lsearch keys. However, making the change in all cases is
825 incompatible and would probably break a number of configurations.
827 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
830 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
831 conversion specification without a maximum field width, thereby enabling
832 a rogue spamd server to cause a buffer overflow. While nobody in their
833 right mind would setup Exim to query an untrusted spamd server, an
834 attacker that gains access to a server running spamd could potentially
835 exploit this vulnerability to run arbitrary code as the Exim user.
837 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
838 $primary_hostname instead of what libspf2 thinks the hosts name is.
840 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
841 a directory entry by the name of the lookup key. Previously, if a
842 symlink pointed to a non-existing file or a file in a directory that
843 Exim lacked permissions to read, a lookup for a key matching that
844 symlink would fail. Now it is enough that a matching directory entry
845 exists, symlink or not. (Bugzilla 503.)
847 PH/03 The body_linecount and body_zerocount variables are now exported in the
850 PH/04 Added the $dnslist_matched variable.
852 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
853 This means they are set thereafter only if the connection becomes
856 PH/06 Added the client_condition to authenticators so that some can be skipped
857 by clients under certain conditions.
859 PH/07 The error message for a badly-placed control=no_multiline_responses left
860 "_responses" off the end of the name.
862 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
864 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
865 (without spaces) instead of just copying the configuration text.
867 PH/10 Added the /noupdate option to the ratelimit ACL condition.
869 PH/11 Added $max_received_linelength.
871 PH/12 Added +ignore_defer and +include_defer to host lists.
873 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
874 way in which PCRE > 7.0 is built.
876 PH/14 Implemented queue_only_load_latch.
878 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
879 MAIL command. The effect was to mangle the value on 64-bit systems.
881 PH/16 Another patch from the Sieve maintainer.
883 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
885 PH/18 If a system quota error occurred while trying to create the file for
886 a maildir delivery, the message "Mailbox is full" was not appended to the
887 bounce if the delivery eventually timed out. Change 4.67/27 below applied
888 only to a quota excession during the actual writing of the file.
890 PH/19 It seems that peer DN values may contain newlines (and other non-printing
891 characters?) which causes problems in log lines. The DN values are now
892 passed through string_printing() before being added to log lines.
894 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
895 and InterBase are left for another time.)
897 PH/21 Added message_body_newlines option.
899 PH/22 Guard against possible overflow in moan_check_errorcopy().
901 PH/23 POSIX allows open() to be a macro; guard against that.
903 PH/24 If the recipient of an error message contained an @ in the local part
904 (suitably quoted, of course), incorrect values were put in $domain and
905 $local_part during the evaluation of errors_copy.
911 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
912 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
915 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
916 issue a MAIL command.
918 PH/02 In an ACL statement such as
920 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
922 if a client was not listed at all, or was listed with a value other than
923 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
924 the condition was not true (as it should be), so access was not denied.
925 The bug was that the ! inversion was incorrectly passed on to the second
926 item. This has been fixed.
928 PH/03 Added additional dnslists conditions == and =& which are different from
929 = and & when the dns lookup returns more than one IP address.
931 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
932 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
934 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
935 FSYNC, which compiles an option called disable_fsync that allows for
936 bypassing fsync(). The documentation is heavily laced with warnings.
938 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
940 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
941 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
942 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
943 including adding "make clean"; (3) Added -fPIC when compiling the test
944 dynamically loaded module, to get rid of a warning.
946 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
947 message fails, move_frozen_messages = true and ignore_bounce_errors_after
948 = 0s. The bug is otherwise harmless.
950 PH/07 There was a bug in the dovecot authenticator such that the value of
951 $auth1 could be overwritten, and so not correctly preserved, after a
952 successful authentication. This usually meant that the value preserved by
953 the server_setid option was incorrect.
955 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
957 PH/09 Installed PCRE release 7.0.
959 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
960 run for batched SMTP input. It is now run at the start of every message
961 in the batch. While fixing this I discovered that the process information
962 (output by running exiwhat) was not always getting set for -bs and -bS
963 input. This is fixed, and it now also says "batched" for BSMTP.
965 PH/11 Added control=no_pipelining.
967 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
968 patch, slightly modified), and move the expansion of helo_data till after
969 the connection is made in the smtp transport (so it can use these
972 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
974 PH/14 Added log_selector = +pid.
976 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
978 PH/16 Add ${if forany and ${if forall.
980 PH/17 Added dsn_from option to vary the From: line in DSNs.
982 PH/18 Flush SMTP output before performing a callout, unless control =
983 no_callout_flush is set.
985 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
986 was true (the default) a successful delivery failed to delete the retry
987 item, thus causing premature timeout of the address. The bug is now
990 PH/20 Added hosts_avoid_pipelining to the smtp transport.
992 PH/21 Long custom messages for fakedefer and fakereject are now split up
993 into multiline reponses in the same way that messages for "deny" and
994 other ACL rejections are.
996 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
997 with slight modification.
999 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1000 draft, changing the syntax and factoring some duplicate code".
1002 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1003 for deliveries of the second and subsequent messages over the same SMTP
1006 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1007 ${reduce, with only minor "tidies".
1009 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1011 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1012 expansion side effects.
1014 PH/27 When a message times out after an over-quota error from an Exim-imposed
1015 quota, the bounce message says "mailbox is full". This message was not
1016 being given when it was a system quota that was exceeded. It now should
1019 MH/03 Made $recipients available in local_scan(). local_scan() already has
1020 better access to the recipient list through recipients_list[], but
1021 $recipients can be useful in postmaster-provided expansion strings.
1023 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1024 in the case of a MAIL command with additional options following the
1025 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1026 were accidentally chopped off.
1028 PH/29 SMTP synchronization checks are implemented when a command is read -
1029 there is a check that no more input is waiting when there shouldn't be
1030 any. However, for some commands, a delay in an ACL can mean that it is
1031 some time before the response is written. In this time, more input might
1032 arrive, invalidly. So now there are extra checks after an ACL has run for
1033 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1034 pipelining has not been advertised.
1036 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1038 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1039 returned in the SMTP response when smtp_return_error_details was false.
1040 This has been fixed.
1042 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1043 instead of the C I/O that was originally supplied, because problems were
1044 reported on Solaris.
1046 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1047 Exim which did not show up earlier: it was assuming that a call to
1048 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1049 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1050 was a macro that became an assignment, so it seemed to work. This has
1051 changed to a proper function call with a void return, hence the compile
1052 error. Exim's code has been fixed.
1054 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1057 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1059 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1061 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1062 because it turns out to be too big or there is a timeout) there is no
1063 "Completed" line in the log. When some messages of this type were
1064 selected by exigrep, they were listed as "not completed". Others were
1065 picked up by some special patterns. I have improved the selection
1066 criteria to be more general.
1068 PH/38 The host_find_failed option in the manualroute router can now be set
1069 to "ignore", to completely ignore a host whose IP address cannot be
1070 found. If all hosts are ignored, the behaviour is controlled by the new
1071 host_all_ignored option.
1073 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1074 homing or because of multiple MX records with /mx) generated more than
1075 one IP address, and the following item turned out to be the local host,
1076 all the secondary addresses of the first item were incorrectly removed
1077 from the list, along with the local host and any following hosts (which
1078 is what is supposed to happen).
1080 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1081 whoever called Exim into the -H file. In the case of the daemon it was
1082 behaving confusingly. When first started, it used values for whoever
1083 started the daemon, but after a SIGHUP it used the Exim user (because it
1084 calls itself on a restart). I have changed the code so that it now always
1087 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1088 message are rejected with the same error (e.g. no authentication or bad
1089 sender address), and a DATA command is nevertheless sent (as can happen
1090 with PIPELINING or a stupid MUA), the error message that was given to the
1091 RCPT commands is included in the rejection of the DATA command. This is
1092 intended to be helpful for MUAs that show only the final error to their
1095 PH/42 Another patch from the Sieve maintainer.
1097 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1098 Eximstats - Fixed some broken HTML links and added missing column headers
1100 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1101 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1103 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1105 PH/43 Yet another patch from the Sieve maintainer.
1107 PH/44 I found a way to check for a TCP/IP connection going away before sending
1108 the response to the final '.' that terminates a message, but only in the
1109 case where the client has not sent further data following the '.'
1110 (unfortunately, this is allowed). However, in many cases there won't be
1111 any further data because there won't be any more messages to send. A call
1112 to select() can be used: if it shows that the input is "ready", there is
1113 either input waiting, or the socket has been closed. An attempt to read
1114 the next input character can distinguish the two cases. Previously, Exim
1115 would have sent an OK response which the client would never have see.
1116 This could lead to message repetition. This fix should cure that, at
1117 least in a lot of common cases.
1119 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1120 advertised in response to EHLO.
1126 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1127 fixed by 4.65/MH/01 (is this a record?) are fixed:
1129 (i) An empty string was always treated as zero by the numeric comparison
1130 operators. This behaviour has been restored.
1132 (ii) It is documented that the numeric comparison operators always treat
1133 their arguments as decimal numbers. This was broken in that numbers
1134 starting with 0 were being interpreted as octal.
1136 While fixing these problems I realized that there was another issue that
1137 hadn't been noticed. Values of message_size_limit (both the global option
1138 and the transport option) were treated as octal if they started with 0.
1139 The documentation was vague. These values are now always treated as
1140 decimal, and I will make that clear in the documentation.
1146 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1147 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1150 MH/01 Don't check that the operands of numeric comparison operators are
1151 integers when their expansion is in "skipping" mode (fixes bug
1152 introduced by 4.64-PH/07).
1154 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1155 child addresses, Exim now panics and dies. Previously, because the count
1156 is held in a short int, deliveries were likely to be lost. As such a
1157 large number of recipients for a single message is ridiculous
1158 (performance will be very, very poor), I have chosen to impose a limit
1159 rather than extend the field.
1165 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1166 leftover -K file (the existence of which was triggered by #402).
1167 While we were at it, introduced process PID as part of the -K
1168 filename. This should rule out race conditions when creating
1171 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1172 processing considerably. Previous code took too long for large mails,
1173 triggering a timeout which in turn triggers #401.
1175 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1176 in the DK code in transports.c. sendfile() is not really portable,
1177 hence the _LINUX specificness.
1179 TF/01 In the add_headers option to the mail command in an Exim filter,
1180 there was a bug that Exim would claim a syntax error in any
1181 header after the first one which had an odd number of characters
1184 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1185 callout verification, Exim cached a "reject" for the entire domain. This
1186 is correct for most verifications, but it is not correct for a recipient
1187 verification with use_sender or use_postmaster set, because in that case
1188 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1189 case of MAIL FROM:<> rejection from other early rejections (e.g.
1190 rejection of HELO). When verifying a recipient using a non-null MAIL
1191 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1192 Whatever the result of the callout, the value of the domain cache is
1193 left unchanged (for any other kind of callout, getting as far as trying
1194 RCPT means that the domain itself is ok).
1196 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1199 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1200 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1201 session, the daemon ignores ECONNECT errors and logs others; it now
1202 ignores EPIPE as well.
1204 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1205 (quoted-printable decoding).
1207 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1208 later the small subsequent patch to fix an introduced bug.
1210 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1212 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1214 PH/08 An error is now given if message_size_limit is specified negative.
1216 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1217 to be given (somewhat) arbitrary names.
1219 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1222 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1223 miscellaneous code fixes
1225 PH/10 Added the log_reject_target ACL modifier to specify where to log
1228 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1229 hostname. This is wrong, because it relates to the incoming message (and
1230 probably the interface on which it is arriving) and not to the outgoing
1231 callout (which could be using a different interface). This has been
1232 changed to use the value of the helo_data option from the smtp transport
1233 instead - this is what is used when a message is actually being sent. If
1234 there is no remote transport (possible with a router that sets up host
1235 addresses), $smtp_active_hostname is used.
1237 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1238 tweaks were necessary in order to get it to work (see also 21 below):
1239 (a) The code assumed that strncpy() returns a negative number on buffer
1240 overflow, which isn't the case. Replaced with Exim's string_format()
1242 (b) There were several signed/unsigned issues. I just did the minimum
1243 hacking in of casts. There is scope for a larger refactoring.
1244 (c) The code used strcasecmp() which is not a standard C function.
1245 Replaced with Exim's strcmpic() function.
1246 (d) The code set only $1; it now sets $auth1 as well.
1247 (e) A simple test gave the error "authentication client didn't specify
1248 service in request". It would seem that Dovecot has changed its
1249 interface. Fortunately there's a specification; I followed it and
1250 changed what the client sends and it appears to be working now.
1252 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1255 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1256 address A is aliased to B and C, where B exists and C does not. Without
1257 -v the output is "A verified" because verification stops after a
1258 successful redirection if more than one address is generated. However,
1259 with -v the child addresses are also verified. Exim was outputting "A
1260 failed to verify" and then showing the successful verification for C,
1261 with its parentage. It now outputs "B failed to verify", showing B's
1262 parentage before showing the successful verification of C.
1264 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1265 look up a TXT record in a specific list after matching in a combined
1268 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1269 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1270 they consult the DNS. I had assumed they would set it the way they
1271 wanted; and indeed my experiments on Linux seem to show that in some
1272 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1273 To be on the safe side, however, I have now made the interface to
1274 host_find_byname() similar to host_find_bydns(), with an argument
1275 containing the DNS resolver options. The host_find_byname() function now
1276 sets these options at its start, just as host_find_bydns() does. The smtp
1277 transport options dns_qualify_single and dns_search_parents are passed to
1278 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1279 of host_find_byname() use the default settings of RES_DEFNAMES
1280 (qualify_single) but not RES_DNSRCH (search_parents).
1282 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1283 spool_read_header() do less string testing, by means of a preliminary
1284 switch on the second character of optional "-foo" lines. (This is
1285 overdue, caused by the large number of possibilities that now exist.
1286 Originally there were few.) While I was there, I also converted the
1287 str(n)cmp tests so they don't re-test the leading "-" and the first
1288 character, in the hope this might squeeze out yet more improvement.
1290 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1291 flag allowing group syntax was set by the header_syntax check but not
1292 turned off, possible causing trouble later; (2) The flag was not being
1293 set at all for the header_verify test, causing "group"-style headers to
1294 be rejected. I have now set it in this case, and also caused header_
1295 verify to ignore an empty address taken from a group. While doing this, I
1296 came across some other cases where the code for allowing group syntax
1297 while scanning a header line wasn't quite right (mostly, not resetting
1298 the flag correctly in the right place). These bugs could have caused
1299 trouble for malformed header lines. I hope it is now all correct.
1301 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1302 with the "reply" argument non-NULL. The code, however (which originally
1303 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1304 but it didn't always do it. This confused somebody who was copying the
1305 code for some other use. I have removed all the tests.
1307 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1308 feature that was used to support insecure browsers during the U.S. crypto
1309 embargo. It requires special client support, and Exim is probably the
1310 only MTA that supported it -- and would never use it because real RSA is
1311 always available. This code has been removed, because it had the bad
1312 effect of slowing Exim down by computing (never used) parameters for the
1313 RSA_EXPORT functionality.
1315 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1316 authenticator to fail if there's a tab character in the incoming data
1317 (there should never be unless someone is messing about, as it's supposed
1318 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1319 if the connection is using TLS or if the remote IP is the same as the
1320 local IP, and the "valid-client-cert option" if a client certificate has
1323 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1324 authenticators. This can be used for authorization after authentication
1325 succeeds. (In the case of plaintext, it servers for both authentication
1328 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1329 if any retry times were supplied.
1331 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1332 connection, where there is no client IP address to check. In this
1333 situation, the verify now always succeeds.
1335 PH/25 Applied John Jetmore's -Mset patch.
1337 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1339 PH/27 In a string expansion for a processed (not raw) header when multiple
1340 headers of the same name were present, leading whitespace was being
1341 removed from all of them, but trailing whitespace was being removed only
1342 from the last one. Now trailing whitespace is removed from each header
1343 before concatenation. Completely empty headers in a concatenation (as
1344 before) are ignored.
1346 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1347 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1349 PH/29 [Removed. This was a change that I later backed out, and forgot to
1350 correct the ChangeLog entry (that I had efficiently created) before
1351 committing the later change.]
1353 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1354 address errors (4xx response to RCPT) over the same connection as other
1355 messages routed to the same hosts. Such deliveries are always "forced",
1356 so retry times are not inspected. This resulted in far too many retries
1357 for the affected addresses. The effect occurred only when there were more
1358 hosts than the hosts_max_try setting in the smtp transport when it had
1359 the 4xx errors. Those hosts that it had tried were not added to the list
1360 of hosts for which the message was waiting, so if all were tried, there
1361 was no problem. Two fixes have been applied:
1363 (i) If there are any address or message errors in an SMTP delivery, none
1364 of the hosts (tried or untried) are now added to the list of hosts
1365 for which the message is waiting, so the message should not be a
1366 candidate for sending over the same connection that was used for a
1367 successful delivery of some other message. This seems entirely
1368 reasonable: after all the message is NOT "waiting for some host".
1369 This is so "obvious" that I'm not sure why it wasn't done
1370 previously. Hope I haven't missed anything, but it can't do any
1371 harm, as the worst effect is to miss an optimization.
1373 (ii) If, despite (i), such a delivery is accidentally attempted, the
1374 routing retry time is respected, so at least it doesn't keep
1375 hammering the server.
1377 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1378 in ${readsocket because some servers need this prod.
1380 PH/32 Added some extra debug output when updating a wait-xxx database.
1382 PH/33 The hint "could be header name not terminated by colon", which has been
1383 given for certain expansion errors for a long time, was not being given
1384 for the ${if def:h_colon_omitted{... case.
1386 PH/34 The spec says: "With one important exception, whenever a domain list is
1387 being scanned, $domain contains the subject domain." There was at least
1388 one case where this was not true.
1390 PH/35 The error "getsockname() failed: connection reset by peer" was being
1391 written to the panic log as well as the main log, but it isn't really
1392 panic-worthy as it just means the connection died rather early on. I have
1393 removed the panic log writing for the ECONNRESET error when getsockname()
1396 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1397 runs only) independently of the message's sender address. This meant
1398 that, if the 4xx error was in fact related to the sender, a different
1399 message to the same recipient with a different sender could confuse
1400 things. In particualar, this can happen when sending to a greylisting
1401 server, but other circumstances could also provoke similar problems.
1402 I have changed the default so that the retry time for these errors is now
1403 based a combination of the sender and recipient addresses. This change
1404 can be overridden by setting address_retry_include_sender=false in the
1407 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1408 remote server are returned as part of bounce messages. This was not
1409 happening for LMTP over a pipe (the lmtp transport), but now it is the
1410 same for both kinds of LMTP.
1412 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1413 in header lines that were in fact CNAMEs. This is no longer the case.
1415 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1416 and queue runs started by the daemon processed all messages. This has
1417 been fixed so that -R and -S can now usefully be given with -q<time>.
1419 PH/40 Import PCRE release 6.7 (fixes some bugs).
1421 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1423 PH/42 Give an error if -q is specified more than once.
1425 PH/43 Renamed the variables $interface_address and $interface_port as
1426 $received_ip_address and $received_port, to make it clear that these
1427 values apply to message reception, and not to the outgoing interface when
1428 a message is delivered. (The old names remain recognized, of course.)
1430 PH/44 There was no timeout on the connect() call when using a Unix domain
1431 socket in the ${readsocket expansion. There now is.
1433 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1434 be meaningful with "accept".
1436 SC/01 Eximstats V1.43
1437 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1439 SC/02 Eximstats V1.44
1440 Use a glob alias rather than an array ref in the generated
1441 parser. This improves both readability and performance.
1443 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1444 Collect SpamAssassin and rejection statistics.
1445 Don't display local sender or destination tables unless
1446 there is data to show.
1447 Added average volumes into the top table text output.
1449 SC/04 Eximstats V1.46
1450 Collect data on the number of addresses (recipients)
1451 as well as the number of messages.
1453 SC/05 Eximstats V1.47
1454 Added 'Message too big' to the list of mail rejection
1455 reasons (thanks to Marco Gaiarin).
1457 SC/06 Eximstats V1.48
1458 Mainlog lines which have GMT offsets and are too short to
1459 have a flag are now skipped.
1461 SC/07 Eximstats V1.49 (Alain Williams)
1462 Added the -emptyok flag.
1464 SC/08 Eximstats V1.50
1465 Fixes for obtaining the IP address from reject messages.
1467 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1468 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1469 whitesspace changes from 4.64-PH/27
1471 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1474 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1475 are found, allow negative numbers in numeric criteria)
1477 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1479 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1480 to match changes made in 4.64-PH/43
1482 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1484 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1485 "domain" is actually an IP address, and force a failure. This locks out
1486 those revolvers/nameservers that support "A-for-A" lookups, in
1487 contravention of the specifications.
1489 PH/48 When a host name was looked up from an IP address, and the subsequent
1490 forward lookup of the name timed out, the host name was left in
1491 $sender_host_name, contrary to the specification.
1493 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1494 restricted to single-key lookups, Exim was not diagnosing an error if
1495 * or *@ was used with a query-style lookup.
1497 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1499 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1500 long ago, but noone interested enough thought of it. Let's just say that
1501 the "1.1" means that there are some new functions that weren't there at
1502 some point in the past.
1504 PH/51 Error processing for expansion failure of helo_data from an smtp
1505 transport during callout processing was broken.
1507 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1508 tested/used via the -bh/-bhc/-bs options.
1510 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1511 bug, fixed in subsequent PCRE releases).
1513 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1514 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1516 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1522 SC/01 Use a glob alias rather than an array ref in eximstats generated
1523 parser. This improves both readability and performance.
1525 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1526 Don't display local sender or destination tables in eximstats unless
1527 there is data to show.
1528 Added average volumes into the eximstats top table text output.
1530 SC/03 Collect data on the number of addresses (recipients) as well
1531 as the number of messages in eximstats.
1533 TF/01 Correct an error in the documentation for the redirect router. Exim
1534 does (usually) call initgroups() when daemonizing.
1536 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1537 with consistent privilege compared to when running as a daemon.
1539 TF/03 Note in the spec that $authenticated_id is not set for local
1540 submissions from trusted users.
1542 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1543 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1545 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1546 by adding some example configuration directives to the default
1547 configuration file. A little bit of work is required to uncomment the
1548 directives and define how usernames and passwords are checked, but
1549 there is now a framework to start from.
1551 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1552 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1553 without this. I don't know how relevant this is to other LDAP libraries.
1555 PH/02 Add the verb name to the "unknown ACL verb" error.
1557 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1559 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1561 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1562 directory not expanded when it should be if an expanded home directory
1563 was set for the address (which is overridden by the transport).
1565 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1568 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1569 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1570 because it is too late at that time, and has no effect.
1572 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1573 security issue with \' (bugzilla #107). I could not use the
1574 PQescapeStringConn() function, because it needs a PGconn value as one of
1577 PH/08 When testing addresses using -bt, indicate those final addresses that
1578 are duplicates that would not cause an additional delivery. At least one
1579 person was confused, thinking that -bt output corresponded to deliveries.
1580 (Suppressing duplicates isn't a good idea as you lose the information
1581 about possibly different redirections that led to the duplicates.)
1583 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1584 systems where poll() doesn't work, in particular OS X.
1586 PH/10 Added more information to debugging output for retry time not reached.
1588 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1589 operations in malware.c.
1591 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1594 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1595 syslog_duplication set false, log lines that would normally be written
1596 both the the main log and to the reject log were not written to syslog at
1599 PH/14 In the default configuration, change the use of "message" in ACL warn
1600 statements to "add_header".
1602 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1603 not followed by a command (e.g. "seen endif").
1605 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1606 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1609 PH/17 Added extra conditions to the default value of delay_warning_condition
1613 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1614 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1615 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1618 The Auto-Submitted: and various List- headers are standardised, whereas I
1619 don't think Precedence: ever was.
1621 PH/18 Refactored debugging code in route_finduser() to show more information,
1622 in particular, the error code if getpwnam() issues one.
1624 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1625 This is apparently needed in addition to the PH/07 change above to avoid
1626 any possible encoding problems.
1628 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1629 but not after initializing Perl.
1631 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1632 output them only if debugging. By default they are written stderr,
1633 apparently, which is not desirable.
1635 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1638 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1641 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1643 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1644 authenticated or an ident call has been made. Suppress the default
1645 values for $authenticated_id and $authenticated_sender (but permit -oMai
1646 and -oMas) when testing with -bh.
1648 PH/24 Re-jigged the order of the tests in the default configuration so that the
1649 tests for valid domains and recipients precede the DNS black list and CSA
1650 tests, on the grounds that those ones are more expensive.
1652 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1653 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1654 This bug exists in every version of Exim that I still have, right back to
1657 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1658 However, an attempt to turn on case-sensitivity in a regex key by
1659 including (?-i) didn't work because the subject string was already
1660 lowercased, and the effects were non-intuitive. It turns out that a
1661 one-line patch can be used to allow (?-i) to work as expected.
1667 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1668 other effects) broke the use of negated acl sub-conditions.
1670 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1673 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1674 "Deny" causes Exim to reject the incoming connection with a 554 error.
1675 Unfortunately, if there is a major crisis, such as a disk failure,
1676 tcp-wrappers gives "deny", whereas what one would like would be some
1677 kind of temporary error. A kludge has been added to help with this.
1678 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1679 554 error is used if errno is still zero or contains ENOENT (which occurs
1680 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1683 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1685 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1686 errors. Otherwise a message that provokes a temporary error (when other
1687 messages do not) can cause a whole host to time out.
1689 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1690 addresses were routed directly to files or pipes from a redirect router.
1691 File deliveries just didn't batch; pipe deliveries might have suffered
1694 PH/06 A failure to get a lock for a hints database would erroneously always say
1695 "Failed to get write lock", even when it was really a read lock.
1697 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1698 of 0600. This has been changed to use the value of the lockfile_mode
1699 option (which defaults to 0600).
1701 PH/08 Applied small patch from the Sieve maintainer.
1703 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1704 folder from quota calculations, a direct delivery into this folder messed
1705 up the contents of the maildirsize file. This was because the regex was
1706 used only to exclude .Trash (or whatever) when the size of the mailbox
1707 was calculated. There was no check that a delivery was happening into an
1708 excluded directory. This bug has been fixed by ignoring all quota
1709 processing for deliveries into excluded directories.
1711 PH/10 Added the maildirfolder_create_regex option to appendfile.
1717 PH/01 The code for finding all the local interface addresses on a FreeBSD
1718 system running IPv6 was broken. This may well have applied to all BSD
1719 systems, as well as to others that have similar system calls. The broken
1720 code found IPv4 interfaces correctly, but gave incorrect values for the
1721 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1722 that it would not match correctly against @[] and not recognize the IPv6
1725 PH/02 The ipliteral router was not recognizing addresses of the form user@
1726 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1728 PH/03 Added disable_ipv6.
1730 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1731 decoded form, because it is most often used to construct To: headers
1732 lines in autoreplies, and the decoded form may well be syntactically
1733 invalid. However, $reply_address has leading white space removed, and all
1734 newlines turned into spaces so that the autoreply transport does not
1737 PH/05 If group was specified without a user on a router, and no group or user
1738 was specified on a transport, the group from the router was ignored.
1740 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1741 for visible compile-time settings that can be used to change these
1742 numbers, for those that want even more. Backwards compatibility with old
1743 spool files has been maintained. However, going back to a previous Exim
1744 release will lost any variables that are in spool files.
1746 PH/07 Two small changes when running in the test harness: increase delay when
1747 passing a TCP/IP connection to a new process, in case the original
1748 process has to generate a bounce, and remove special handling of
1749 127.0.0.2 (sic), which is no longer necessary.
1751 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1752 be the same on different OS.
1754 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1757 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1758 whether --show-vars was specified or not
1760 JJ/02 exipick: Added support for new ACL variable spool format introduced
1763 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1764 syntactically invalid From: or Reply-to: line, and a filter used this to
1765 generate an autoreply, and therefore failed to obtain an address for the
1766 autoreply, Exim could try to deliver to a non-existent relative file
1767 name, causing unrelated and misleading errors. What now happens is that
1768 it logs this as a hard delivery error, but does not attempt to create a
1771 PH/11 The exinext utility has a -C option for testing purposes, but although
1772 the given file was scanned by exinext itself; it wasn't being passed on
1773 when Exim was called.
1775 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1776 an end-of-file indication when reading a command response.
1778 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1779 compiled. In many other places in Exim, IPv6 addresses are always
1780 recognized, so I have changed this. It also means that IPv4 domain
1781 literals of the form [IPV4:n.n.n.n] are now always recognized.
1783 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1784 used if the router is not running as root, for example, when verifying at
1785 ACL time, or when using -bh. The debugging output from this situation was
1786 non-existent - all you got was a failure to exec. I have made two
1789 (a) Failures to set uid/gid, the current directory, or a process leader
1790 in a subprocess such as that created by queryprogram now generate
1791 suitable debugging ouput when -d is set.
1793 (b) The queryprogram router detects when it is not running as root,
1794 outputs suitable debugging information if -d is set, and then runs
1795 the subprocess without attempting to change uid/gid.
1797 PH/15 Minor change to Makefile for building test_host (undocumented testing
1800 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1801 additional section of a DNS packet that returns MX or SRV records.
1802 Instead, it always explicitly searches for A/AAAA records. This avoids
1803 major problems that occur when a DNS server includes only records of one
1804 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1805 fixed another bug: if SRV records were looked up and the corresponding
1806 address records were *not* found in the additional section, the port
1807 values from the SRV records were lost.
1809 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1810 using the correct key (the original address) when searching the retry
1811 rules in order to find which one to use for generating the retry hint.
1813 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1814 adding the default one. Similarly, if it contains a Reply-To: header, the
1815 errors_reply_to option, if set, is not used.
1817 PH/19 When calculating a retry time, Exim used to measure the "time since
1818 failure" by looking at the "first failed" field in the retry record. Now
1819 it does not use this if it is later than than the arrival time of the
1820 message. Instead it uses the arrival time. This makes for better
1821 behaviour in cases where some deliveries succeed, thus re-setting the
1822 "first failed" field. An example is a quota failure for a huge message
1823 when small messages continue to be delivered. Without this change, the
1824 "time since failure" will always be short, possible causing more frequent
1825 delivery attempts for the huge message than are intended.
1826 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1828 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1829 $1, $2, $3) because the numerical variables can be reset during some
1830 expansion items (e.g. "match"), thereby losing the authentication data.
1832 PH/21 Make -bV show the size of off_t variables so that the test suite can
1833 decide whether to run tests for quotas > 2G.
1835 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1836 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1837 filecount value is greater than 2G or if a quota value is greater than 2G
1838 on a system where the size of off_t is not greater than 4, a panic error
1841 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1842 never match. The debug and -bh output now contains an explicit error
1843 message indicating a malformed IPv4 address or mask.
1845 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1846 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1847 PH/23 above applies.
1849 PH/25 Do not write to syslog when running in the test harness. The only
1850 occasion when this arises is a failure to open the main or panic logs
1851 (for which there is an explicit test).
1853 PH/26 Added the /no_tell option to "control=freeze".
1855 PH/27 If a host name lookup failed very early in a connection, for example, if
1856 the IP address matched host_lookup and the reverse lookup yielded a name
1857 that did not have a forward lookup, an error message of the form "no IP
1858 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1859 could be logged. Now it outputs the IP address instead of "NULL".
1861 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1862 allows the sender and the authenticated sender to be set when
1863 submitting a message from within Exim. Since child_open_exim() is
1864 documented for local_scan(), the new function should be too.
1866 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1867 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1868 results in an empty string is now treated as unset.
1870 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1872 PH/31 Added sender_verify_fail logging option.
1874 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1875 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1876 tidied the source and removed it altogether.
1878 PH/33 When a queue run was abandoned because the load average was too high, a
1879 log line was always written; now it is written only if the queue_run log
1880 selector is set. In addition, the log line for abandonment now contains
1881 information about the queue run such as the pid. This is always present
1882 in "start" and "stop" lines but was omitted from the "abandon" line.
1884 PH/34 Omit spaces between a header name and the colon in the error message that
1885 is given when verify = headers_syntax fails (if there are lots of them,
1886 the message gets confusing).
1888 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1889 names, as there are now some PTR records that contain slashes. This check
1890 is only to protect against broken name servers that fall over on strange
1891 characters, so the fact that it applies to all lookups doesn't matter.
1893 PH/36 Now that the new test suite is complete, we can remove some of the
1894 special code in Exim that was needed for the old test suite. For example,
1895 sorting DNS records because real resolvers return them in an arbitrary
1896 order. The new test suite's fake resolver always returns records in the
1899 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1900 bounces) except when queue_only is set, to avoid logging races between
1901 the different processes.
1903 PH/38 Panic-die if .include specifies a non-absolute path.
1905 PH/39 A tweak to the "H" retry rule from its user.
1907 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1908 a label. They prevented compilation on older perls.
1910 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1911 a warning to be raised on newish perls.
1913 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1914 on queue. Changes to match documented behaviour of showing count of
1915 messages matching specified criteria.
1917 PH/40 Changed the default ident timeout from 30s to 5s.
1919 PH/41 Added support for the use of login_cap features, on those BSD systems
1920 that have them, for controlling the resources used by pipe deliveries.
1922 PH/42 The content-scanning code uses fopen() to create files in which to put
1923 message data. Previously it was not paying any attention to the mode of
1924 the files. Exim runs with umask(0) because the rest of the code creates
1925 files with open(), and sets the required mode explicitly. Thus, these
1926 files were ending up world-writeable. This was not a big issue, because,
1927 being within the spool directory, they were not world-accessible. I have
1928 created a function called modefopen, which takes an additional mode
1929 argument. It sets umask(777), creates the file, chmods it to the required
1930 mode, then resets the umask. All the relevant calls to fopen() in the
1931 content scanning code have been changed to use this function.
1933 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1934 to 24 hours. This avoids potential overflow problems when processing G
1935 and H retry rules. I suspect nobody ever tinkers with this value.
1937 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1939 PH/45 When the plaintext authenticator is running as a client, the server's
1940 challenges are checked to ensure they are valid base64 strings. By
1941 default, the authentication attempt is cancelled if an invalid string is
1942 received. Setting client_ignore_invalid_base64 true ignores these errors.
1943 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1944 they are received. Thus, the responses can be made to depend on the
1945 challenges. If an invalid string is ignored, an empty string is placed in
1948 PH/46 Messages that are created by the autoreply transport now contains a
1949 References: header, in accordance with RFCs 2822 and 3834.
1951 PH/47 Added authenticated_sender_force to the smtp transport.
1953 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1955 PH/49 Installed latest patch from the Sieve maintainer.
1957 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1958 was also set, the appendfile transport was unnecessarily scanning a
1959 directory of message files (e.g. for maildir delivery) to find the count
1960 of files (along with the size), even though it did not need this
1961 information. It now does the scan only if it needs to find either the
1962 size of the count of files.
1964 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1966 PH/52 Two bugs concerned with error handling when the smtp transport is
1969 (i) Exim was not creating retry information for temporary errors given
1970 for individual recipients after the DATA command when the smtp transport
1971 was used in LMTP mode. This meant that they could be retried too
1972 frequently, and not timed out correctly.
1974 (ii) Exim was setting the flag that allows error details to be returned
1975 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1976 recipients that were returned after the DATA command.
1978 PH/53 This is related to PH/52, but is more general: for any failing address,
1979 when detailed error information was permitted to be returned to the
1980 sender, but the error was temporary, then after the final timeout, only
1981 "retry timeout exceeded" was returned. Now it returns the full error as
1982 well as "retry timeout exceeded".
1984 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1985 do this, and (what is worse) MTAs that accept it.
1987 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1988 will now be deprecated.
1990 PH/56 New os.c-cygwin from the Cygwin maintainer.
1992 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1993 formats (previously only available in exim formats via -bpr, -bpru,
1994 and -bpra. Now also available in native and exiqgrep formats)
1996 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1997 with very large, slow to parse queues
1999 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2001 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2003 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2004 responses to authentication challenges, though it was showing the
2005 challenges; (ii) I've removed the CR characters from the debug output for
2008 PH/58 Allow for the insertion of a newline as well as a space when a string
2009 is turned into more than one encoded-word during RFC 2047 encoding. The
2010 Sieve code now uses this.
2012 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2013 data_4xx, lost_connection, tls_required.
2015 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2016 message was being sent as an SMTP response.
2018 PH/61 Add -l and -k options to exicyclog.
2020 PH/62 When verifying, if an address was redirected to one new address, so that
2021 verification continued, and the new address failed or deferred after
2022 having set something in $address_data, the value of $address_data was not
2023 passed back to the ACL. This was different to the case when no
2024 redirection occurred. The value is now passed back in both cases.
2026 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2027 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2028 use login_cap.h, so on its own it isn't the distinguishing feature. The
2029 new name refers directly to the setclassresources() function.
2031 PH/65 Added configuration files for NetBSD3.
2033 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2035 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2036 is preferred over IPv4.
2038 PH/68 The bounce_return_message and bounce_return_body options were not being
2039 honoured for bounces generated during the reception of non-SMTP messages.
2040 In particular, this applied to messages rejected by the ACL. This bug has
2041 been fixed. However, if bounce_return_message is true and bounce_return_
2042 body is false, the headers that are returned for a non-SMTP message
2043 include only those that have been read before the error was detected.
2044 (In the case of an ACL rejection, they have all been read.)
2046 PH/69 The HTML version of the specification is now built in a directory called
2047 spec_html instead of spec.html, because the latter looks like a path with
2048 a MIME-type, and this confuses some software.
2050 PH/70 Catch two compiler warnings in sieve.c.
2052 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2053 function verify_get_ident() calls ip_connect() to connect a socket, but
2054 if the "connect()" function timed out, ip_connect() used to close the
2055 socket. However, verify_get_ident() also closes the socket later, and in
2056 between Exim writes to the log, which may get opened at this point. When
2057 the socket was closed in ip_connect(), the log could get the same file
2058 descriptor number as the socket. This naturally causes chaos. The fix is
2059 not to close the socket in ip_connect(); the socket should be closed by
2060 the function that creates it. There was only one place in the code where
2061 this was missing, in the iplookup router, which I don't think anybody now
2062 uses, but I've fixed it anyway.
2064 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2065 well as to direct DNS lookups. Otherwise the handling of names in host
2066 lists is inconsistent and therefore confusing.
2072 PH/01 Two changes to the default runtime configuration:
2074 (1) Move the checks for relay_from_hosts and authenticated clients from
2075 after to before the (commented out) DNS black list checks.
2077 (2) Add control=submission to the relay_from_hosts and authenticated
2078 clients checks, on the grounds that messages accepted by these
2079 statements are most likely to be submissions.
2081 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2083 (1) Generate an error if the third argument for the ${prvs expansion is
2086 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2089 (3) Reset the variables that are obtained from the first argument of
2090 ${prvscheck and used in the second argument before leaving the code,
2091 because their memory is reclaimed, so using them afterwards may do
2094 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2095 one (it's much easier than Tom thought :-).
2097 (5) Because of (4), we can now allow for the use of $prvscheck_result
2098 inside the third argument.
2100 PH/03 For some reason, the default setting of PATH when running a command from
2101 a pipe transport was just "/usr/bin". I have changed it to
2104 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2105 anything to be listed in the output from -bV.
2107 PH/05 When a filter generated an autoreply, the entire To: header line was
2108 quoted in the delivery log line, like this:
2110 => >A.N.Other <ano@some.domain> <original@ddress> ...
2112 This has been changed so that it extracts the operative address. There
2113 may be more than one such address. If so, they are comma-separated, like
2116 => >ano@some.domain,ona@other.domain <original@ddress> ...
2118 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2119 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2120 not being looked up in the rDNS to get a host name, Exim was showing the
2121 IP address twice in Received: lines, even though the IP addresses were
2122 identical. For example:
2124 Received: from [1.2.3.4] (helo=[1.2.3.4])
2126 However, if the real host name was known, it was omitting the HELO data
2127 if it matched the actual IP address. This has been tidied up so that it
2128 doesn't show the same IP address twice.
2130 PH/07 When both +timestamp and +memory debugging was on, the value given by
2131 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2132 called by the debug printing, thereby overwriting the timestamp buffer.
2133 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2135 PH/08 When the original message was included in an autoreply transport, it
2136 always said "this is a copy of the message, including all the headers",
2137 even if body_only or headers_only was set. It now gives an appropriate
2140 PH/09 Applied a patch from the Sieve maintainer which:
2142 o fixes some comments
2143 o adds the (disabled) notify extension core
2144 o adds some debug output for the result of if/elsif tests
2145 o points to the current vacation draft in the documentation
2146 and documents the missing references header update
2150 o fixes a bug in processing the envelope test (when testing
2151 multiple envelope elements, the last element determinted the
2154 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2155 Electronic Mail") by including:
2157 Auto-submitted: auto-generated
2159 in the messages that it generates (bounce messages and others, such as
2160 warnings). In the case of bounce messages for non-SMTP mesages, there was
2161 also a typo: it was using "Auto_submitted" (underscore instead of
2162 hyphen). Since every message generated by Exim is necessarily in response
2163 to another message, thes have all been changed to:
2165 Auto-Submitted: auto-replied
2167 in accordance with these statements in the RFC:
2169 The auto-replied keyword:
2171 - SHOULD be used on messages sent in direct response to another
2172 message by an automatic process,
2174 - MUST NOT be used on manually-generated messages,
2176 - MAY be used on Delivery Status Notifications (DSNs) and Message
2177 Disposition Notifications (MDNs),
2179 - MUST NOT be used on messages generated by automatic or periodic
2180 processes, except for messages which are automatic responses to
2183 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2184 to the default Received: header definition.
2186 PH/12 Added log selector acl_warn_skipped (default on).
2188 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2189 variables because (a) they are in the wrong storage pool and (b) even if
2190 they were copied, it wouldn't work properly because of the caching.
2192 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2193 checking when decoding. Apparently there are clients that generate
2194 overlong encoded strings. Why am I not surprised?
2196 PH/15 If the first argument of "${if match_address" was not empty, but did not
2197 contain an "@" character, Exim crashed. Now it writes a panic log message
2198 and treats the condition as false.
2200 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2202 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2203 extension "envelope-auth". The code is finished and in agreement with
2204 other implementations, but there is no documentation so far and in fact,
2205 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2206 not changing the active code.
2208 Print executed "if" and "elsif" statements when debugging is used. This
2209 helps a great deal to understand what a filter does.
2211 Document more things not specified clearly in RFC3028. I had all this
2212 sorted out, when out of a sudden new issues came to my mind. Oops."
2214 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2217 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2218 canonical form (as documented). However, after a host name lookup from
2219 the IP address, check_host() was doing a simple string comparison with
2220 addresses acquired from the DNS when checking that the found name did
2221 have the original IP as one of its addresses. Since any found IPv6
2222 addresses are likely to be in abbreviated form, the comparison could
2223 fail. Luckily, there already exists a function for doing the comparison
2224 by converting both addresses to binary, so now that is used instead of
2225 the text comparison.
2227 PH/20 There was another similar case to PH/19, when a complete host name was
2228 given in a host list; looking up its IP address could give an abbreviated
2229 form, whereas the current host's name might or might not be abbreviated.
2230 The same fix has been applied.
2236 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2237 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2240 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2241 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2243 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2245 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2246 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2247 Also, for "Auto-Submitted", treat anything other than "no" as
2248 non-personal, in accordance with RFC 3834. (Previously it treated
2249 anything starting "auto-" as non-personal.)
2251 TF/01 The control=submission/name=... option had a problem with syntax
2252 errors if the name included a slash character. The /name= option
2253 now slurps the rest of the string, so it can include any characters
2254 but it must come last in the list of options (after /sender_retain
2257 PH/05 Some modifications to the interface to the fake nameserver for the new
2265 TK/01 Added the "success_on_redirect" address verification option. See
2266 NewStuff for rationale and an example.
2268 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2270 PH/02 Patch to exigrep to allow it to work on syslog lines.
2272 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2273 fread() to skip over the body file's header line, because in Cygwin the
2274 header line is locked and is inaccessible.
2276 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2277 co-exist for some time) to make it clear that it is the Exim ID that is
2278 referenced, not the Message-ID: header line.
2280 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2281 string_format() function, because snprintf() does not exist on all
2284 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2285 ${stat: expansion item.
2287 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2288 protocol synchronization error", to keep the pedants happy.
2290 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2291 well as for IRIX systems, when gcc is being used. See the host.c source
2294 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2296 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2299 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2300 transport and to the smtp transport in LMTP mode.
2302 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2304 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2305 run a filter in a subprocess. This could lead to confusion in subsequent
2306 lookups in the parent process. There should also be a search_tidyup() at
2307 the end of the subprocess.
2309 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2310 only if the host matched helo_try_verify_hosts, which caused the
2311 verification to occur when the EHLO/HELO command was issued. The ACL just
2312 tested the remembered result. Now, if a previous verification attempt has
2313 not happened, "verify = helo" does it there and then.
2315 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2317 TK/03 Fix log output including CR from clamd.
2319 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2320 exist provoked a memory error which could cause a segfault.
2322 PH/15 Installed PCRE 6.2
2324 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2326 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2327 of the problem. Specifically, suggested +O2 rather than +O1 for the
2330 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2332 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2333 record, the port was not correctly logged when the outgoing_port log
2334 selector was set (it logged the transort's default port).
2336 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2337 fallback_hosts, and "hosts" in the smtp transport.
2339 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2340 host errors such as "Connection refused".
2342 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2343 authentication with radiusclient 0.4.9:
2345 - Error returned from rc_read_config was caught wrongly
2346 - Username/password not passed on to radius server due to wrong length.
2348 The presumption is that some radiusclient API changes for 4.51/PH/17
2349 were not taken care of correctly. The code is still untested by me (my
2350 Linux distribution still has 0.3.2 of radiusclient), but it was
2351 contributed by a Radius user.
2353 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2354 expanding the "port" option of the smtp transport.
2356 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2357 while reading a MIME header. Thanks to Tom Hughes for a patch.
2359 PH/24 Include config.h inside local_scan.h so that configuration settings are
2362 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2363 that in an ACL for RCPT (for example), you can examine exactly what was
2366 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2367 commands, but it was not correctly comparing the address with the actual
2368 client host address. Thus, it would show the EHLO address in Received:
2369 header lines when this was not necessary.
2371 PH/27 Added the % operator to ${eval:}.
2373 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2374 it should be ignoring failures (because with -C, for example, it has lost
2375 privilege). It wasn't ignoring creation failures other than "already
2378 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2381 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2382 because an input line was too long, either on its own, or by virtue of
2383 too many continuations, the temporary file was not being removed, and the
2384 return code was incorrect.
2386 PH/31 Missing "BOOL" in function definition in filtertest.c.
2388 PH/32 Applied Sieve patches from the maintainer.
2390 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2392 PH/33 Added "verify = not_blind".
2394 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2395 Local/Makefile (with some defaults set). These are used in built scripts
2396 such as exicyclog, but they have never been used in the exim_install
2397 script (though there are many overriding facilities there). I have
2398 arranged that the exim_install script now takes note of these two
2401 PH/35 Installed configuration files for Dragonfly.
2403 PH/36 When a locally submitted message by a trusted user did not contain a
2404 From: header, and the sender address was obtained from -f or from an SMTP
2405 MAIL command, and the trusted user did not use -F to supply a sender
2406 name, $originator_name was incorrectly used when constructing a From:
2407 header. Furthermore, $originator_name was used for submission mode
2408 messages from external hosts without From: headers in a similar way,
2409 which is clearly wrong.
2411 PH/37 Added control=suppress_local_fixups.
2413 PH/38 When log_selector = +received_sender was set, and the addition of the
2414 sender made the log line's construction buffer exactly full, or one byte
2415 less than full, an overflow happened when the terminating "\n" was
2418 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2419 when the result of a list match is failure because a DNS lookup failed.
2421 PH/40 RM_COMMAND is now used in the building process.
2423 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2424 the "build-* directories that it finds.
2426 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2427 address in a domain literal was a prefix of an interface address.
2429 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2430 when verifying a sender address, unless rewrite_headers is false.
2432 PH/44 Wrote a long comment about why errors_to addresses are verified as
2433 recipients, not senders.
2435 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2436 the ratelimit ACL was added.
2438 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2440 PH/46 Added extra information about PostgreSQL errors to the error string.
2442 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2443 avoiding the need to install special zones in a real server. This is
2444 backwards compatible; if it can't find the fake resolver, it drops back.
2445 Thus, both old and new test suites can be run.
2447 TF/02 Added util/ratelimit.pl
2449 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2450 clock is set back in time.
2452 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2453 Candler <B.Candler@pobox.com>.
2455 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2456 OK for addresses that are the result of redirections.
2458 PH/48 A number of further additions for the benefit of the new test suite,
2459 including a fake gethostbyname() that interfaces to the fake DNS resolver
2462 TF/06 The fix for widen_domains has also been applied to qualify_single and
2463 search_parents which are the other dnslookup options that can cause
2466 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2469 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2471 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2472 service" if it encountered an MX record with an empty target hostname.
2473 The message is now "an MX or SRV record indicated no SMTP service".
2475 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2476 if the DNS of the sending site is misconfigured. This is quite a
2477 common situation. This change restores the behaviour of treating a
2478 helo verification defer as a failure.
2480 PH/51 If self=fail was set on a router, the bounce message did not include the
2481 actual error message.
2487 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2489 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2490 command itself ended in error, the underlying message about the transport
2491 filter timeout was being overwritten with the pipe command error. Now the
2492 underlying error message should be appended to the second error message.
2494 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2496 PH/02 Reduce the amount of output that "make" produces by default. Full output
2497 can still be requested.
2499 PH/03 The warning log line about a condition test deferring for a "warn" verb
2500 was being output only once per connection, rather than after each
2501 occurrence (because it was using the same function as for successful
2502 "warn" verbs). This seems wrong, so I have changed it.
2504 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2505 it should not have, which might have caused a crash in the right
2506 circumstances, but probably never did.
2508 PH/04 Installed a modified version of Tony Finch's patch to make submission
2509 mode fix the return path as well as the Sender: header line, and to
2510 add a /name= option so that you can make the user's friendly name appear
2513 TF/03 Added the control = fakedefer ACL modifier.
2515 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2516 Mark Lowes for thorough testing.
2518 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2520 TK/03 Merged latest SRS patch from Miles Wilton.
2522 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2523 with the definition in sysexits.h (which is #included earlier).
2524 Fortunately, Exim does not actually use EX_OK. The code used to try to
2525 preserve the sysexits.h value, by assumimg that macro definitions were
2526 scanned for macro replacements. I have been disabused of this notion,
2527 so now the code just undefines EX_OK before #including unistd.h.
2529 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2530 in the smtp transport. When a block could not be written in a single
2531 write() function, the timeout was being re-applied to each part-write.
2532 This seems wrong - if the receiver was accepting one byte at a time it
2533 would take for ever. The timeout is now adjusted when this happens. It
2534 doesn't have to be particularly precise.
2536 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2537 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2539 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2540 without a domain if the check to <postmaster@domain> fails.
2542 SC/01 Eximstats: added -xls and the ability to specify output files
2543 (patch written by Frank Heydlauf).
2545 SC/02 Eximstats: use FileHandles for outputing results.
2547 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2549 SC/04 Eximstats: fixed display of large numbers with -nvr option
2551 SC/05 Eximstats: fixed merging of reports with empty tables.
2553 SC/06 Eximstats: added the -include_original_destination flag
2555 SC/07 Eximstats: removed tabs and trailing whitespace.
2557 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2559 TK/06 MBOX spool code: Add real "From " MBOX separator line
2560 so the .eml file is really in mbox format (even though
2561 most programs do not really care). Patch from Alex Miller.
2563 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2564 The latter is generated from $received_to and is only set if the
2565 message has one envelope recipient. SA can use these headers,
2566 obviously out-of-the-box. Patch from Alex Miller.
2568 PH/08 The ${def test on a variable was returning false if the variable's
2569 value was "0", contrary to what the specification has always said!
2570 The result should be true unless the variable is empty.
2572 PH/09 The syntax error of a character other than { following "${if
2573 def:variable_name" (after optional whitespace) was not being diagnosed.
2574 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2575 accidental colon was present, for example, could give incorrect results.
2577 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2578 result is used (not including appendfile, where other changes are about
2581 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2582 This involved changing a lot of size variables from int to off_t. It
2583 should work with maildirs and everything.
2585 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2586 spamd dying while we are connected to it.
2588 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2591 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2592 function for BDB 4.3.
2594 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2596 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2597 the sources for PCRE have been re-arranged and more files are now
2600 PH/15 The code I had for printing potentially long long variables in PH/11
2601 above was not the best (it lost precision). The length of off_t variables
2602 is now inspected at build time, and an appropriate printing format (%ld
2603 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2604 to be "long long int" or "long int". This is needed for the internal
2605 formatting function string_vformat().
2607 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2608 the configuration file to be ":syslog", then the script "guesses" where
2609 the logs files are, rather than using the compiled in default. In our
2610 case the guess is not the same as the compiled default, so the script
2611 suddenly stopped working when I started to use syslog. The patch checks
2612 to see if log_file_path is "". If so, it attempts to read it from exim
2613 with no configuration file to get the compiled in version, before it
2614 falls back to the previous guessing code."
2616 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2617 implementing BATV in an Exim configuration. See NewStuff for the gory
2620 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2621 Makefile that are specific to HP-UX.
2623 PH/18 If the "use_postmaster" option was set for a recipient callout together
2624 with the "random" option, the postmaster address was used as the MAIL
2625 FROM address for the random test, but not for the subsequent recipient
2626 test. It is now used for both.
2628 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2629 patch removes a few documentation additions to RFC 3028, because the
2630 latest draft now contains them. It adds the new en;ascii-case comparator
2631 and a new error check for 8bit text in MIME parts. Comparator and
2632 require names are now matched exactly. I enabled the subaddress
2633 extension, but it is not well tested yet (read: it works for me)."
2635 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2636 rework some of the code of TK/09 above to avoid the hardwired use of
2637 "%lld" and "long long". Replaced the call to snprintf() with a call to
2640 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2641 records point to non-existent hosts", "retry timeout exceeded", and
2642 "retry time not reached for any host after a long failure period".
2644 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2645 experimental DomainKeys support:
2647 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2648 (2) On an error such as an illegally used "control", the wrong name for
2649 the control was given.
2651 These problems did NOT occur unless DomainKeys support was compiled.
2653 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2655 PH/24 Added ${if match_ip condition.
2657 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2658 databases so that it will be absolutely obvious if a crash occurs in the
2659 DB library. This is a regular occurrence (often caused by mis-matched
2662 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2663 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2664 picked up on a user's system that detects such things. There doesn't seem
2665 to be a gcc warning option for this - only an attribute that has to be
2666 put on the function's prototype. It seems that in Fedora Core 4 they have
2667 set this on a number of new functions. No doubt there will be more in due
2670 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2671 specify a transport. However, if an address that was verified by such a
2672 router was the subject of a callout, Exim crashed because it tried to
2673 read the rcpt_include_affixes from the non-existent transport. Now it
2674 just assumes that the setting of that option is false. This bug was
2675 introduced by 4.51/PH/31.
2677 PH/28 Changed -d+all to exclude +memory, because that information is very
2678 rarely of interest, but it makes the output a lot bigger. People tend to
2679 do -d+all out of habit.
2681 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2682 code in os-type was giving problems when libc.so lives in lib64, like on
2685 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2686 aren't the modern standard, and it seems that some systems' include files
2687 don't always have them. Exim was already checking for some of the newer
2688 ones like T_AAAA, and defining it itself. I've added checks for all the
2689 record types that Exim uses.
2691 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2692 not automatically generating a new one, as it is supposed to. This
2693 prevented TLS from working. If the file did exist, but contained invalid
2694 data, a new version was generated, as expected. It was only the case of a
2695 non-existent file that was broken.
2697 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2698 with a change in libdomainkeys > 0.64.
2700 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2701 from DNS. If the selector record carries the flag, it now has
2702 precedence over the domain-wide flag.
2704 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2706 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2707 the use of a port name that isn't defined in /etc/services) occurred, the
2708 message was deferred as in a normal delivery, and thus remained on the
2709 spool, instead of being failed because of the mua_wrapper setting. This
2710 is now fixed, and I tidied up some of the mua_wrapper messages at the
2713 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2714 the messages in a hash of arrays rather than using individual hashes.
2715 This is a bit cleaner and results in dramatic memory savings, albeit
2716 at a slight CPU cost.
2718 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2719 as requested by Marc Sherman.
2721 SC/10 Eximstats: added histograms for user specified patterns as requested
2724 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2726 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2727 fopen() in the content-scanning modules that did not already have it.
2733 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2734 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2736 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2738 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2740 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2741 to test Sieve filters that use "vacation".
2743 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2744 that changes the way the GnuTLS parameters are stored in the cache file.
2745 The new format can be generated externally. For backward compatibility,
2746 if the data in the cache doesn't make sense, Exim assumes it has read an
2747 old-format file, and it generates new data and writes a new file. This
2748 means that you can't go back to an older release without removing the
2751 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2752 work if there are any delivery delays because "one_time" forces the
2753 parent to be marked "delivered", so its unseen clone is never tried
2754 again. For this reason, Exim now forbids the simultaneous setting of
2757 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2758 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2759 entry if you want to know the details. The fix, however, neglected to
2760 consider the case when local delivery batching is involved. The test for
2761 "previously delivered" was not happening when checking to see if an
2762 address could be batched with a previous (undelivered) one; under
2763 certain circumstances this could lead to multiple deliveries to the same
2766 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2767 in its include files, and this causes problems building Exim.
2769 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2770 header_syntax) but Exim was just ignoring anything given after a slash.
2771 In particular, this caused confusion with an attempt to use "verify =
2772 reverse_host_lookup/defer_ok". An error is now given when options are
2773 supplied for verify items that do not have them. (Maybe reverse_host_
2774 lookup should have a defer_ok option, but that's a different point.)
2776 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2777 defined by RFC 821) to 2048, because there were problems with some AUTH
2778 commands, and RFC 1869 says the size should be increased for extended
2779 SMTP commands that take arguments.
2781 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2784 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2785 "unknown" error; now it says that the functionality isn't in the binary.
2787 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2788 an address' error message when a string expansion fails (syntax or
2789 whatever). Otherwise the password may appear in the log. Following change
2790 PH/42 below, there is no longer a chance of it appearing in a bounce
2793 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2795 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2796 of its addresses was ever tried. (Bugzilla bug #2.)
2798 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2799 the result incorrectly in the debug output. (It correctly added a newline
2800 to what was transported.)
2802 TF/01 Added $received_time.
2804 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2805 commented out examples of how to interface to a virus scanner and to
2806 SpamAssassin. Also added commented examples of av_scanner and
2807 spamd_address settings.
2809 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2810 and controls are allowed in which ACLs. There were a couple of minor
2811 errors. Some of the entries in the conditions table (which is a table of
2812 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2813 negation of where the condition IS allowed.
2815 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2817 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2818 header file does not have a version number, so I've had to invent a new
2819 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2820 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2821 radiusclient), but it was contributed by a Radius user.
2823 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2824 files or directories, for OpenSSL.
2826 PH/19 When an Exim process that is running as root has to create an Exim log
2827 file, it does so in a subprocess that runs as exim:exim so as to get the
2828 ownership right at creation (otherwise, other Exim processes might see
2829 the file with the wrong ownership). There was no test for failure of this
2830 fork() call, which would lead to the process getting stuck as it waited
2831 for a non-existent subprocess. Forks do occasionally fail when resources
2832 run out. I reviewed all the other calls to fork(); they all seem to check
2835 PH/20 When checking for unexpected SMTP input at connect time (before writing
2836 the banner), Exim was not dealing correctly with a non-positive return
2837 from the read() function. If the client had disconnected by this time,
2838 the result was a log entry for a synchronization error with an empty
2839 string after "input=" when read() returned zero. If read() returned -1
2840 (an event I could not check), uninitialized data bytes were printed.
2841 There were reports of junk text (parts of files, etc) appearing after
2844 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2846 PH/22 Added support for macro redefinition, and (re)definition in between
2847 driver and ACL definitions.
2849 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2850 forgetting to use the resulting value; it was using the unexpanded value.
2852 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2853 hadn't been configured. The fix is from Juergen Kreileder, who
2854 understands it better than I do:
2856 "Here's what I see happening with three configured cyrus_sasl
2857 authenticators configured (plain, login, cram-md5):
2859 On startup auth_cyrus_sasl_init() gets called for each of these.
2860 This means three calls to sasl_listmech() without a specified mech_list.
2861 => SASL tests which mechs of all available mechs actually work
2862 => three warnings about OTP not working
2863 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2865 With the patch, sasl_listmech() also gets called three times. But now
2866 SASL's mech_list option is set to the server_mech specified in the the
2867 authenticator. Or in other words, the answer from sasl_listmech()
2868 gets limited to just the mech you're testing for (which is different
2870 => the return list contains just 'plain' or 'login', 'cram-md5' or
2871 nothing depending on the value of ob->server_mech.
2873 I've just tested the patch: Authentication still works fine,
2874 unavailable mechs specified in the exim configuration are still
2875 caught, and the auth.log warnings about OTP are gone."
2877 PH/25 When debugging is enabled, the contents of the command line are added
2878 to the debugging output, even when log_selector=+arguments is not
2881 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2882 answer is "GNU", and only if the return is "GNU/something" is the answer
2885 PH/27 $acl_verify_message is now set immediately after the failure of a
2886 verification in an ACL, and so is available in subsequent modifiers. In
2887 particular, the message can be preserved by coding like this:
2889 warn !verify = sender
2890 set acl_m0 = $acl_verify_message
2892 Previously, $acl_verify_message was set only while expanding "message"
2893 and "log_message" when a very denied access.
2895 PH/28 Modified OS/os.c-Linux with
2897 -#ifndef OS_LOAD_AVERAGE
2898 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2900 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2901 nomenclature these days.)
2903 PH/29 Installed patch from the Sieve maintainer that adds the options
2904 sieve_useraddress and sieve_subaddress to the redirect router.
2906 PH/30 In these circumstances:
2907 . Two addresses routed to the same list of hosts;
2908 . First host does not offer TLS;
2909 . First host accepts first address;
2910 . First host gives temporary error to second address;
2911 . Second host offers TLS and a TLS session is established;
2912 . Second host accepts second address.
2913 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2914 and peerdn, if requested) that were in fact used only for the second
2917 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2918 attention to any local part prefix or suffix that was matched by the
2919 router that accepted the address. It now behaves in the same way as it
2920 does for delivery: the affixes are removed from the local part unless
2921 rcpt_include_affixes is set on the transport.
2923 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2924 timeout during the DATA phase of an incoming message.
2926 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2927 applied a patch sanctioned by the Sieve maintainer.
2929 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2930 the uid or gid is negative. A case of a negative gid caused this to be
2931 noticed. The fix allows for either to be negative.
2933 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2934 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2937 PH/36 The change PH/12 above was broken. Fixed it.
2939 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2940 the grounds that routing the same address twice would always produce the
2941 same answer. This might have been true once, but it is certainly no
2942 longer true now. Routing a child address may depend on the previous
2943 routing that produced that child. Some complicated redirection strategies
2944 went wrong when messages had multiple recipients, and made Exim's
2945 behaviour dependent on the order in which the addresses were given.
2947 I have moved the duplicate checking until after the routing is complete.
2948 Exim scans the addresses that are assigned to local and remote
2949 transports, and removes any duplicates. This means that more work will be
2950 done, as duplicates will always all be routed, but duplicates are
2951 presumably rare, so I don't expect this is of any significance.
2953 For deliveries to pipes, files, and autoreplies, the duplicate checking
2954 still happens during the routing process, since they are not going to be
2957 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2958 It corrects a timeout issue with spamd. This is Ian's comment: "The
2959 background is that sometimes spamd either never reads data from a
2960 connection it has accepted, or it never writes response data. The exiscan
2961 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2962 blindly assumes that writes won't block so it may never time out."
2964 PH/39 Allow G after quota size as well as K and M.
2966 PH/40 The value set for $authenticated_id in an authenticator may not contain
2967 binary zeroes or newlines because the value is written to log lines and
2968 to spool files. There was no check on this. Now the value is run through
2969 the string_printing() function so that such characters are converted to
2970 printable escape sequences.
2972 PH/41 $message_linecount is a new variable that contains the total number of
2973 lines in the message. Compare $body_linecount, which is the count for the
2976 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2977 bounce and delay warning messages, except in certain special cases, which
2980 (a) An SMTP error message from a remote host;
2981 (b) A message specified in a :fail: redirection;
2982 (c) A message specified in a "fail" command in a system filter;
2983 (d) A message specified in a FAIL return from the queryprogram router;
2984 (e) A message specified by the cannot_route_message router option.
2986 In these cases only, Exim does include the error details in bounce and
2987 warning messages. There are also a few cases where bland messages such
2988 as "unrouteable address" or "local delivery error" are given.
2990 PH/43 $value is now also set for the "else" part of a ${run expansion.
2992 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2993 being worked on, but at least Exim now implements the latest version to
2996 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2997 process to complete was treated as a delivery failure, a timeout while
2998 writing the message to the pipe was logged, but erroneously treated as a
2999 successful delivery. Such timeouts include transport filter timeouts. For
3000 consistency with the overall process timeout, these timeouts are now
3001 treated as errors, giving rise to delivery failures by default. However,
3002 there is now a new Boolean option for the pipe transport called
3003 timeout_defer, which, if set TRUE, converts the failures into defers for
3004 both kinds of timeout. A transport filter timeout is now identified in
3007 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3008 systems where "make" and "gmake" are different, calling "gmake" at top
3009 level broke things. I've arranged for the value of $(MAKE) to be passed
3010 from the Makefile to this script so that it can call the same version of
3014 A note about Exim versions 4.44 and 4.50
3015 ----------------------------------------
3017 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3018 changes of various kinds. As a consequence, a big documentation update was
3019 needed. This delayed the release for rather longer than seemed good, especially
3020 in the light of a couple of (minor) security issues. Therefore, the changes
3021 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3022 release. So 4.44 and 4.50 are in effect two different branches that both start
3025 I have left the 4.50 change log unchanged; it contains all the changes since
3026 4.43. The change log for 4.44 is below; many of its items are identical to
3027 those for 4.50. This seems to be the most sensible way to preserve the
3028 historical information.
3034 1. Minor wording change to the doc/README.SIEVE file.
3036 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3037 computation of the current number of files was incorrect.
3039 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3040 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3043 4. Give more explanation in the error message when the command for a transport
3044 filter fails to execute.
3046 5. There are several places where Exim runs a non-Exim command in a
3047 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3048 was being done only for the command run by the queryprogram router. It is
3049 now done for all such subprocesses. The other cases are: ${run, transport
3050 filters, and the commands run by the lmtp and pipe transports.
3052 6. Added CONFIGURE_GROUP build-time option.
3054 7. Some older OS have a limit of 256 on the maximum number of file
3055 descriptors. Exim was using setrlimit() to set 1000 as a large value
3056 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3057 systems. I've change it so that if it can't get 1000, it tries for 256.
3059 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3060 was an oversight, and furthermore, ever since the addition of extra
3061 controls (e.g. 4.43/32), the checks on when to allow different forms of
3062 "control" were broken. There should now be diagnostics for all cases when a
3063 control that does not make sense is encountered.
3065 9. Added the /retain_sender option to "control=submission".
3067 10. $recipients is now available in the predata ACL (oversight).
3069 11. Tidy the search cache before the fork to do a delivery from a message
3070 received from the command line. Otherwise the child will trigger a lookup
3071 failure and thereby defer the delivery if it tries to use (for example) a
3072 cached ldap connection that the parent has called unbind on.
3074 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3075 of $address_data from the recipient verification was clobbered by the
3076 sender verification.
3078 13. The value of address_data from a sender verification is now available in
3079 $sender_address_data in subsequent conditions in the ACL statement.
3081 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3083 15. Added a new option "connect=<time>" to callout options, to set a different
3086 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3087 was its contents. (It was OK if the option was not defined at all.)
3089 17. A "Completed" log line is now written for messages that are removed from
3090 the spool by the -Mrm option.
3092 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3093 information about exactly what failed.
3095 19. Added -dd to debug only the daemon process.
3097 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3098 handles timeouts, both on the server side and network timeouts. Renamed the
3099 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3101 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3102 It is now set to "smtps".
3104 22. $host_address is now set to the target address during the checking of
3105 ignore_target_hosts.
3107 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3108 being passed; this would have caused $sender_host_name to have been used if
3109 matching the list had actually called for a host name (not very likely,
3110 since this list is usually IP addresses). A host name is now passed as
3113 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3114 code that specifies a non-restarting handler (typically sigaction() in
3115 modern systems) in an attempt to fix a rare and obscure crash bug.
3117 25. Narrowed the window for a race in the daemon that could cause it to ignore
3118 SIGCHLD signals. This is not a major problem, because they are used only to
3119 wake it up if nothing else does.
3121 26. A malformed maildirsize file could cause Exim to calculate negative values
3122 for the mailbox size or file count. Odd effects could occur as a result.
3123 The maildirsize information is now recalculated if the size or filecount
3126 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3127 support for a long time. Removed HAVE_SYS_VFS_H.
3129 28. Installed the latest version of exipick from John Jetmore.
3131 29. In an address list, if the pattern was not a regular expression, an empty
3132 subject address (from a bounce message) matched only if the pattern was an
3133 empty string. Non-empty patterns were not even tested. This was the wrong
3134 because it is perfectly reasonable to use an empty address as part of a
3135 database query. An empty address is now tested by patterns that are
3136 lookups. However, all the other forms of pattern expect the subject to
3137 contain a local part and a domain, and therefore, for them, an empty
3138 address still always fails if the pattern is not itself empty.
3140 30. Exim went into a mad DNS loop when attempting to do a callout where the
3141 host was specified on an smtp transport, and looking it up yielded more
3142 than one IP address.
3144 31. Re-factored the code for checking spool and log partition space into a
3145 function that finds that data and another that does the check. The former
3146 is then used to implement four new variables: $spool_space, $log_space,
3147 $spool_inodes, and $log_inodes.
3149 32. The RFC2047 encoding function was originally intended for short strings
3150 such as real names; it was not keeping to the 75-character limit for
3151 encoded words that the RFC imposes. It now respects the limit, and
3152 generates multiple encoded words if necessary. To be on the safe side, I
3153 have increased the buffer size for the ${rfc2047: expansion operator from
3156 33. It is now permitted to omit both strings after an "if" condition; if the
3157 condition is true, the result is "true". As before, when the second string
3158 is omitted, a false condition yields an empty string. This makes it less
3159 cumbersome to write custom ACL and router conditions.
3161 34. Failure to deliver a bounce message always caused it to be frozen, even if
3162 there was an errors_to setting on the router. The errors_to setting is now
3165 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3166 canonical form (fully expanded) before being placed in
3167 $sender_host_address.
3169 36. The table in the code that translates DNS record types into text (T_A to
3170 "A" for instance) was missing entries for NS and CNAME. It is just possible
3171 that this could have caused confusion if both these types were looked up
3172 for the same domain, because the text type is used as part of Exim's
3173 per-process caching. But the chance of anyone hitting this buglet seems
3176 37. The dnsdb lookup has been extended in a number of ways.
3178 (1) There is a new type, "zns", which walks up the domain tree until it
3179 finds some nameserver records. It should be used with care.
3181 (2) There is a new type, "mxh", which is like "mx" except that it returns
3182 just the host names, not the priorities.
3184 (3) It is now possible to give a list of domains (or IP addresses) to be
3185 looked up. The behaviour when one of the lookups defers can be
3186 controlled by a keyword.
3188 (4) It is now possible to specify the separator character for use when
3189 multiple records are returned.
3191 38. The dnslists ACL condition has been extended: it is now possible to supply
3192 a list of IP addresses and/or domains to be looked up in a particular DNS
3195 39. Added log_selector=+queue_time_overall.
3197 40. When running the queue in the test harness, wait just a tad after forking a
3198 delivery process, to get repeatability of debugging output.
3200 41. Include certificate and key file names in error message when GnuTLS fails
3201 to set them up, because the GnuTLS error message doesn't include the name
3202 of the failing file when there is a problem reading it.
3204 42. Allow both -bf and -bF in the same test run.
3206 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3208 44. The "Exiscan patch" is now merged into the mainline Exim source.
3210 45. Sometimes the final signoff response after QUIT could fail to get
3211 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3212 < 0 before doing a fflush(). This bug looks as though it goes back to the
3213 introduction of TLS in release 3.20, but "sometimes" must have been rare
3214 because the tests only now provoked it.
3216 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3217 (this can affect the format of dates).
3219 47. exim_tidydb, when checking for the continued existence of a message for
3220 which it has found a message-specific retry record, was not finding
3221 messages that were in split spool directories. Consequently, it was
3222 deleting retry records that should have stayed in existence.
3224 48. Steve fixed some bugs in eximstats.
3226 49. The SPA authentication driver was not abandoning authentication and moving
3227 on to the next authenticator when an expansion was forced to fail,
3228 contradicting the general specification for all authenticators. Instead it
3229 was generating a temporary error. It now behaves as specified.
3231 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3232 (the order specifies the preference for clients). The order is now AES256,
3233 AES128, 3DES, ARCFOUR128.
3235 51. Small patch to Sieve code - explicitly set From: when generating an
3238 52. Exim crashed if a remote delivery caused a very long error message to be
3239 recorded - for instance if somebody sent an entire SpamAssassin report back
3240 as a large number of 550 error lines. This bug was coincidentally fixed by
3241 increasing the size of one of Exim's internal buffers (big_buffer) that
3242 happened as part of the Exiscan merge. However, to be on the safe side, I
3243 have made the code more robust (and fixed the comments that describe what
3246 53. Now that there can be additional text after "Completed" in log lines (if
3247 the queue_time_overall log selector is set), a one-byte patch to exigrep
3248 was needed to allow it to recognize "Completed" as not the last thing in
3251 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3252 patch that reportedly fixes this has been added. I am not expert enough to
3253 create a test for it. This is what the patch creator wrote:
3255 "I found a little strange behaviour of ldap code when working with
3256 Windows 2003 AD Domain, where users was placed in more than one
3257 Organization Units. When I tried to give exim partial DN, the exit code
3258 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3259 But simultaneously result of request was absolutely normal ldap result,
3260 so I produce this patch..."
3262 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3263 REFERENCE, so I have modified the code to exclude the patch when that macro
3266 55. Some experimental protocols are using DNS PTR records for new purposes. The
3267 keys for these records are domain names, not reversed IP addresses. The
3268 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3269 leaves it alone. Component reversal etc. now happens only for IP addresses.
3272 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3274 57. Double the size of the debug message buffer (to 2048) so that more of very
3275 long debug lines gets shown.
3277 58. The exicyclog utility now does better if the number of log files to keep
3278 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3280 59. Two changes related to the smtp_active_hostname option:
3282 (1) $smtp_active_hostname is now available as a variable.
3283 (2) The default for smtp_banner uses $smtp_active_hostname instead
3284 of $primary_hostname.
3286 60. The host_aton() function is supposed to be passed a string that is known
3287 to be a valid IP address. However, in the case of IPv6 addresses, it was
3288 not checking this. This is a hostage to fortune. Exim now panics and dies
3289 if the condition is not met. A case was found where this could be provoked
3290 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3291 components; fortuitously, this particular loophole had already been fixed
3292 by change 4.50/55 above.
3294 If there are any other similar loopholes, the new check in host_aton()
3295 itself should stop them being exploited. The report I received stated that
3296 data on the command line could provoke the exploit when Exim was running as
3297 exim, but did not say which command line option was involved. All I could
3298 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3299 running as the user.
3302 61. There was a buffer overflow vulnerability in the SPA authentication code
3303 (which came originally from the Samba project). I have added a test to the
3304 spa_base64_to_bits() function which I hope fixes it.
3307 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3308 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3310 63. The daemon start-up calls getloadavg() while still root for those OS that
3311 need the first call to be done as root, but it missed one case: when
3312 deliver_queue_load_max is set with deliver_drop_privilege. This is
3313 necessary for the benefit of the queue runner, because there is no re-exec
3314 when deliver_drop_privilege is set.
3316 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3317 This has been fixed.
3319 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3320 was in use, was not putting the data itself into the right store pool;
3321 consequently, it could be overwritten for a subsequent message in the same
3322 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3325 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3327 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3328 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3329 the function were treating the return as a boolean value, which happened to
3330 work because 0=false and not-0=true, but is not correct code.
3332 68. The host_aton() function was not handling scoped IPv6 addresses (those
3333 with, for example, "%eth0" on the end) correctly.
3335 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3336 negated items (that is, ~something) in unsigned ints. Some compilers
3337 apparently mutter when there is no cast.
3339 70. If an address verification called from an ACL failed, and did not produce a
3340 user-specific message (i.e. there was only a "system" message), nothing was
3341 put in $acl_verify_message. In this situation, it now puts the system
3344 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3345 session; change 4.31/43 added the unwanted input to the log line - except
3346 that it did not do this in the start of session case. It now does.
3348 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3349 This is wrong and can cause the other end to generate a synchronization
3350 error if it is another Exim or anything else that does the synchronization
3351 check. A QUIT command is no longer sent after a timeout.
3353 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3354 during host lookups.
3356 74. The defer_ok option of callout verification was not working if it was used
3357 when verifying addresses in header lines, that is, for this case:
3359 verify = header_sender/callout=defer_ok
3361 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3362 those file descriptors could be used for SMTP connections. If anything
3363 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3364 could be sent to the SMTP client, causing chaos. The daemon now opens
3365 stdin, stdout, and stderr to /dev/null when it puts itself into the
3368 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3369 log by default. The user can override this with suitable Perl magic.
3371 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3372 the log message when discard triggers, was not working for the DATA ACL or
3373 for the non-SMTP ACL.
3375 78. Error message wording change in sieve.c.
3377 79. If smtp_accept_max_per_host was set, the number of connections could be
3378 restricted to fewer than expected, because the daemon was trying to set up
3379 a new connection before checking whether the processes handling previous
3380 connections had finished. The check for completed processes is now done
3381 earlier. On busy systems, this bug wouldn't be noticed because something
3382 else would have woken the daemon, and it would have reaped the completed
3385 80. If a message was submitted locally by a user whose login name contained one
3386 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3387 It caused a spool format error. I have fixed the spool reading code. A
3388 related problem was that the "from" clause in the Received: line became
3389 illegal because of the space(s). It is now covered by ${quote_local_part.
3391 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3394 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3396 83. Patch from Sieve maintainer to fix unterminated string problem in
3397 "vacation" handling.
3399 84. Some minor changes to the Linux configuration files to help with other
3400 OS variants using glibc.
3402 85. One more patch for Sieve to update vacation handling to latest spec.
3405 ----------------------------------------------------
3406 See the note above about the 4.44 and 4.50 releases.
3407 ----------------------------------------------------
3413 1. Change 4.43/35 introduced a bug that caused file counts to be
3414 incorrectly computed when quota_filecount was set in an appendfile
3417 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3418 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3421 3. Give more explanation in the error message when the command for a transport
3422 filter fails to execute.
3424 4. There are several places where Exim runs a non-Exim command in a
3425 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3426 was being done only for the command run by the queryprogram router. It is
3427 now done for all such subprocesses. The other cases are: ${run, transport
3428 filters, and the commands run by the lmtp and pipe transports.
3430 5. Some older OS have a limit of 256 on the maximum number of file
3431 descriptors. Exim was using setrlimit() to set 1000 as a large value
3432 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3433 systems. I've change it so that if it can't get 1000, it tries for 256.
3435 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3436 was an oversight, and furthermore, ever since the addition of extra
3437 controls (e.g. 4.43/32), the checks on when to allow different forms of
3438 "control" were broken. There should now be diagnostics for all cases when a
3439 control that does not make sense is encountered.
3441 7. $recipients is now available in the predata ACL (oversight).
3443 8. Tidy the search cache before the fork to do a delivery from a message
3444 received from the command line. Otherwise the child will trigger a lookup
3445 failure and thereby defer the delivery if it tries to use (for example) a
3446 cached ldap connection that the parent has called unbind on.
3448 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3449 of $address_data from the recipient verification was clobbered by the
3450 sender verification.
3452 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3453 was its contents. (It was OK if the option was not defined at all.)
3455 11. A "Completed" log line is now written for messages that are removed from
3456 the spool by the -Mrm option.
3458 12. $host_address is now set to the target address during the checking of
3459 ignore_target_hosts.
3461 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3462 being passed; this would have caused $sender_host_name to have been used if
3463 matching the list had actually called for a host name (not very likely,
3464 since this list is usually IP addresses). A host name is now passed as
3467 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3468 code that specifies a non-restarting handler (typically sigaction() in
3469 modern systems) in an attempt to fix a rare and obscure crash bug.
3471 15. Narrowed the window for a race in the daemon that could cause it to ignore
3472 SIGCHLD signals. This is not a major problem, because they are used only to
3473 wake it up if nothing else does.
3475 16. A malformed maildirsize file could cause Exim to calculate negative values
3476 for the mailbox size or file count. Odd effects could occur as a result.
3477 The maildirsize information is now recalculated if the size or filecount
3480 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3481 support for a long time. Removed HAVE_SYS_VFS_H.
3483 18. Updated exipick to current release from John Jetmore.
3485 19. Allow an empty sender to be matched against a lookup in an address list.
3486 Previously the only cases considered were a regular expression, or an
3489 20. Exim went into a mad DNS lookup loop when doing a callout where the
3490 host was specified on the transport, if the DNS lookup yielded more than
3493 21. The RFC2047 encoding function was originally intended for short strings
3494 such as real names; it was not keeping to the 75-character limit for
3495 encoded words that the RFC imposes. It now respects the limit, and
3496 generates multiple encoded words if necessary. To be on the safe side, I
3497 have increased the buffer size for the ${rfc2047: expansion operator from
3500 22. Failure to deliver a bounce message always caused it to be frozen, even if
3501 there was an errors_to setting on the router. The errors_to setting is now
3504 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3505 canonical form (fully expanded) before being placed in
3506 $sender_host_address.
3508 24. Updated eximstats to version 1.33
3510 25. Include certificate and key file names in error message when GnuTLS fails
3511 to set them up, because the GnuTLS error message doesn't include the name
3512 of the failing file when there is a problem reading it.
3514 26. Expand error message when OpenSSL has problems setting up cert/key files.
3517 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3518 (this can affect the format of dates).
3520 28. exim_tidydb, when checking for the continued existence of a message for
3521 which it has found a message-specific retry record, was not finding
3522 messages that were in split spool directories. Consequently, it was
3523 deleting retry records that should have stayed in existence.
3525 29. eximstats updated to version 1.35
3526 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3527 1.35 - bugfix such that pie charts by volume are generated correctly
3529 30. The SPA authentication driver was not abandoning authentication and moving
3530 on to the next authenticator when an expansion was forced to fail,
3531 contradicting the general specification for all authenticators. Instead it
3532 was generating a temporary error. It now behaves as specified.
3534 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3535 (the order specifies the preference for clients). The order is now AES256,
3536 AES128, 3DES, ARCFOUR128.
3538 31. Small patch to Sieve code - explicitly set From: when generating an
3541 32. Exim crashed if a remote delivery caused a very long error message to be
3542 recorded - for instance if somebody sent an entire SpamAssassin report back
3543 as a large number of 550 error lines. This bug was coincidentally fixed by
3544 increasing the size of one of Exim's internal buffers (big_buffer) that
3545 happened as part of the Exiscan merge. However, to be on the safe side, I
3546 have made the code more robust (and fixed the comments that describe what
3549 33. Some experimental protocols are using DNS PTR records for new purposes. The
3550 keys for these records are domain names, not reversed IP addresses. The
3551 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3552 leaves it alone. Component reversal etc. now happens only for IP addresses.
3555 34. The host_aton() function is supposed to be passed a string that is known
3556 to be a valid IP address. However, in the case of IPv6 addresses, it was
3557 not checking this. This is a hostage to fortune. Exim now panics and dies
3558 if the condition is not met. A case was found where this could be provoked
3559 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3560 components; fortuitously, this particular loophole had already been fixed
3561 by change 4.50/55 or 4.44/33 above.
3563 If there are any other similar loopholes, the new check in host_aton()
3564 itself should stop them being exploited. The report I received stated that
3565 data on the command line could provoke the exploit when Exim was running as
3566 exim, but did not say which command line option was involved. All I could
3567 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3568 running as the user.
3571 35. There was a buffer overflow vulnerability in the SPA authentication code
3572 (which came originally from the Samba project). I have added a test to the
3573 spa_base64_to_bits() function which I hope fixes it.
3576 36. The daemon start-up calls getloadavg() while still root for those OS that
3577 need the first call to be done as root, but it missed one case: when
3578 deliver_queue_load_max is set with deliver_drop_privilege. This is
3579 necessary for the benefit of the queue runner, because there is no re-exec
3580 when deliver_drop_privilege is set.
3582 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3583 was in use, was not putting the data itself into the right store pool;
3584 consequently, it could be overwritten for a subsequent message in the same
3585 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3588 38. Sometimes the final signoff response after QUIT could fail to get
3589 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3590 < 0 before doing a fflush(). This bug looks as though it goes back to the
3591 introduction of TLS in release 3.20, but "sometimes" must have been rare
3592 because the tests only now provoked it.
3598 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3599 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3600 Now it returns an integer. A number of calls were still expecting a T/F
3601 return. Fortuitously, in all cases, the tests worked in OK situations,
3602 which is the norm. However, things would have gone wrong on any write
3603 failures on the smtp file descriptor. This function is used when sending
3604 messages over SMTP and also when doing verify callouts.
3606 2. When Exim is called to do synchronous delivery of a locally submitted
3607 message (the -odf or -odi options), it no longer closes stderr before doing
3610 3. Implemented the mua_wrapper option.
3612 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3614 5. Implemented the functions header_remove(), header_testname(),
3615 header_add_at_position(), and receive_remove_recipient(), and exported them
3618 6. If an ACL "warn" statement specified the addition of headers, Exim already
3619 inserted X-ACL-Warn: at the start if there was no header name. However, it
3620 was not making this test for the second and subsequent header lines if
3621 there were newlines in the string. This meant that an invalid header could
3622 be inserted if Exim was badly configured.
3624 7. Allow an ACL "warn" statement to add header lines at the start or after all
3625 the Received: headers, as well as at the end.
3627 8. Added the rcpt_4xx retry error code.
3629 9. Added postmaster_mailfrom=xxx to callout verification option.
3631 10. Added mailfrom=xxxx to the callout verification option, for verify=
3634 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3635 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3637 12. Inserted some casts to stop certain compilers warning when using pointer
3638 differences as field lengths or precisions in printf-type calls (mostly
3639 affecting debugging statements).
3641 13. Added optional readline() support for -be (dynamically loaded).
3643 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3644 same clock tick as a message's arrival, so that its received time was the
3645 same as the "first fail" time on the retry record, and that message
3646 remained on the queue past the ultimate address timeout, every queue runner
3647 would try a delivery (because it was past the ultimate address timeout) but
3648 after another failure, the ultimate address timeout, which should have then
3649 bounced the address, did not kick in. This was a "< instead of <=" error;
3650 in most cases the first failure would have been in the next clock tick
3651 after the received time, and all would be well.
3653 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3654 being recognized when the domain list was tested by the match_domain
3655 condition in an expansion string.
3657 16. Added the ${str2b64: operator.
3659 17. Exim was always calling setrlimit() to set a large limit for the number of
3660 processes, without checking whether the existing limit was already
3661 adequate. (It did check for the limit on file descriptors.) Furthermore,
3662 errors from getrlimit() and setrlimit() were being ignored. Now they are
3663 logged to the main and panic logs, but Exim does carry on, to try to do its
3664 job under whatever limits there are.
3666 18. Imported PCRE 5.0.
3668 19. Trivial typo in log message " temporarily refused connection" (the leading
3671 20. If the log selector return_path_on_delivery was set and an address was
3672 redirected to /dev/null, the delivery process crashed because it assumed
3673 that a return path would always be set for a "successful" delivery. In this
3674 case, the whole delivery is bypassed as an optimization, and therefore no
3677 21. Internal re-arrangement: the function for sending a challenge and reading
3678 a response while authentication was assuming a zero-terminated challenge
3679 string. It's now changed to take a pointer and a length, to allow for
3680 binary data in such strings.
3682 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3684 23. Exim was not respecting finduser_retries when seeking the login of the
3685 uid under which it was called; it was always trying 10 times. (The default
3686 setting of finduser_retries is zero.) Also, it was sleeping after the final
3687 failure, which is pointless.
3689 24. Implemented tls_on_connect_ports.
3691 25. Implemented acl_smtp_predata.
3693 26. If the domain in control=submission is set empty, Exim assumes that the
3694 authenticated id is a complete email address when it generates From: or
3695 Sender: header lines.
3697 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3698 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3699 chown and chgrp in /bin and hostname in /usr/bin.
3701 28. Exim was keeping the "process log" file open after each use, just as it
3702 does for the main log. This opens the possibility of it remaining open for
3703 long periods when the USR1 signal hits a daemon. Occasional processlog
3704 errors were reported, that could have been caused by this. Anyway, it seems
3705 much more sensible not to leave this file open at all, so that is what now
3708 29. The long-running daemon process does not normally write to the log once it
3709 has entered its main loop, and it closes the log before doing so. This is
3710 so that log files can straightforwardly be renamed and moved. However,
3711 there are a couple of unusual error situations where the daemon does write
3712 log entries, and I had neglected to close the log afterwards.
3714 30. The text of an SMTP error response that was received during a remote
3715 delivery was being truncated at 512 bytes. This is too short for some of
3716 the long messages that one sometimes sees. I've increased the limit to
3719 31. It is now possible to make retry rules that apply only when a message has a
3720 specific sender, in particular, an empty sender.
3722 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3723 it possible to be selective about when SMTP synchronization is enforced.
3725 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3727 32. Implemented hosts_connection_nolog.
3729 33. Added an ACL for QUIT.
3731 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3734 35. Added mailbox_size and mailbox_filecount to appendfile.
3736 36. Added control = no_multiline_responses to ACLs.
3738 37. There was a bug in the logic of the code that waits for the clock to tick
3739 in the case where the clock went backwards by a substantial amount such
3740 that the microsecond fraction of "now" was more than the microsecond
3741 fraction of "then" (but the whole seconds number was less).
3743 38. Added support for the libradius Radius client library this is found on
3744 FreeBSD (previously only the radiusclient library was supported).
3750 1. When certain lookups returned multiple values in the form name=value, the
3751 quoting of the values was not always being done properly. Specifically:
3752 (a) If the value started with a double quote, but contained no whitespace,
3754 (b) If the value contained whitespace other than a space character (i.e.
3755 tabs or newlines or carriage returns) it was not quoted.
3756 This fix has been applied to the mysql and pgsql lookups by writing a
3757 separate quoting function and calling it from the lookup code. The fix
3758 should probably also be applied to nisplus, ibase and oracle lookups, but
3759 since I cannot test any of those, I have not disturbed their existing code.
3761 2. A hit in the callout cache for a specific address caused a log line with no
3762 reason for rejecting RCPT. Now it says "Previous (cached) callout
3763 verification failure".
3765 3. There was an off-by-one bug in the queryprogram router. An over-long
3766 return line was truncated at 256 instead of 255 characters, thereby
3767 overflowing its buffer with the terminating zero. As well as fixing this, I
3768 have increased the buffer size to 1024 (and made a note to document this).
3770 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3771 when Exim is waiting for an SMTP response from a remote server, Exim
3772 restarts its select() call on the socket, thereby resetting its timeout.
3773 This is not a problem when such interrupts are rare. Somebody set up a cron
3774 job to run exiwhat every 2 minutes, which is less than the normal select()
3775 timeout (5 or 10 minutes). This meant that the select() timeout never
3776 kicked in because it was always reset. I have fixed this by comparing the
3777 time when an interrupt arrives with the time at the start of the first call
3778 to select(). If more time than the timeout has elapsed, the interrupt is
3779 treated as a timeout.
3781 5. Some internal re-factoring in preparation for the addition of Sieve
3782 extensions (by MH). In particular, the "personal" test is moved to a
3783 separate function, and given an option for scanning Cc: and Bcc: (which is
3784 not set for Exim filters).
3786 6. When Exim created an email address using the login of the caller as the
3787 local part (e.g. when creating a From: or Sender: header line), it was not
3788 quoting the local part when it contained special characters such as @.
3790 7. Installed new OpenBSD configuration files.
3792 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3793 try to make them clearer.
3795 9. Callout options, other than the timeout value, were being ignored when
3796 verifying sender addresses in header lines. For example, when using
3798 verify = header_sender/callout=no_cache
3800 the cache was (incorrectly) being used.
3802 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3803 only the Cygwin environment.
3805 11. When return_path_on_delivery was set as a log selector, if different remote
3806 addresses in the same message used different return paths and parallel
3807 remote delivery occurred, the wrong values would sometimes be logged.
3808 (Whenever a remote delivery process finished, the return path value from
3809 the most recently started remote delivery process was logged.)
3811 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3812 lines when AUTH and/or TLS are in use. This is the "received protocol"
3813 field. Exim used to use "asmtp" for authenticated SMTP, without any
3814 indication (in the protocol name) for TLS use. Now it follows the RFC and
3815 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3816 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3817 names appear in log lines as well as in Received: header lines.
3819 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3820 extensions, and comparison tests, and to fix some bugs.
3822 14. Changes to the "personal" filter test:
3824 (1) The test was buggy in that it was just doing the equivalent of
3825 "contains" tests on header lines. For example, if a user's address was
3826 anne@some.where, the "personal" test would incorrectly be true for
3828 To: susanne@some.where
3830 This test is now done by extracting each address from the header in turn,
3831 and checking the entire address. Other tests that are part of "personal"
3832 are now done using regular expressions (for example, to check local parts
3833 of addresses in From: header lines).
3835 (2) The list of non-personal local parts in From: addresses has been
3836 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3837 taken from the Sieve specification recommendations.
3839 (3) If the message contains any header line starting with "List-" it is
3840 treated as non-personal.
3842 (4) The test for "circular" in the Subject: header line has been removed
3843 because it now seems ill-conceived.
3845 15. Minor typos in src/EDITME comments corrected.
3847 16. Installed latest exipick from John Jetmore.
3849 17. If headers_add on a router specified a text string that was too long for
3850 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3851 of string_sprintf() is now avoided.
3853 18. $message_body_size was not set (it was always zero) when running the DATA
3854 ACL and the local_scan() function.
3856 19. For the "mail" command in an Exim filter, no default was being set for
3857 the once_repeat time, causing a random time value to be used if "once" was
3858 specified. (If the value happened to be <= 0, no repeat happened.) The
3859 default is now 0s, meaning "never repeat". The "vacation" command was OK
3860 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3861 (I found it when inspecting the code).
3863 20. There is now an overall timeout for performing a callout verification. It
3864 defaults to 4 times the callout timeout, which applies to individual SMTP
3865 commands during the callout. The overall timeout applies when there is more
3866 than one host that can be tried. The timeout is checked before trying the
3867 next host. This prevents very long delays if there are a large number of
3868 hosts and all are timing out (e.g. when the network connections are timing
3869 out). The value of the overall timeout can be changed by specifying an
3870 additional sub-option for "callout", called "maxwait". For example:
3872 verify = sender/callout=5s,maxwait=20s
3874 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3875 to the end before writing, but this should make it even safer).
3877 22. Exim was forgetting that it had advertised PIPELINING for the second and
3878 subsequent messages on an SMTP connection. It was also not resetting its
3879 memory on STARTTLS and an internal HELO.
3881 23. When Exim logs an SMTP synchronization error within a session, it now
3882 records whether PIPELINING has been advertised or not.
3884 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3885 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3886 rather than long int.
3888 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3890 26. Added the never_mail option to autoreply.
3896 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3897 crash if the getsockname() call failed; this can happen if a connection is
3898 closed very soon after it is established. The problem was simply in the
3899 order in which certain operations were done, causing Exim to try to write
3900 to the SMTP stream before it had set up the file descriptor. The bug has
3901 been fixed by making things happen in the correct order.
3907 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3908 before the connection was closed, thus losing the rejection response.
3910 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3911 some early Solaris releases, but causes trouble in current releases where
3912 socklen_t is defined.
3914 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3917 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3920 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3921 that releasing the top of it at the end releases what was used for sub-
3922 expansions (unless the block got too big). However, discard this block if
3923 the first thing is a variable or header, so that we can use its block when
3924 it is dynamic (useful for very large $message_headers, for example).
3926 6. Lookups now cache *every* query, not just the most recent. A new, separate
3927 store pool is used for this. It can be recovered when all lookup caches are
3928 flushed. Lookups now release memory at the end of their result strings.
3929 This has involved some general refactoring of the lookup sources.
3931 7. Some code has been added to the store_xxx() functions to reduce the amount
3932 of flapping under certain conditions.
3934 8. log_incoming_interface used to affect only the <= reception log lines. Now
3935 it causes the local interface and port to be added to several more SMTP log
3936 lines, for example "SMTP connection from", and rejection lines.
3938 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3940 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3942 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3943 could be overwritten at the end of the current message (or the start of a
3944 new message if it was set in a HELO ACL). The value is now preserved for
3945 the duration of the SMTP connection.
3947 12. If a transport had a headers_rewrite setting, and a matching header line
3948 contained an unqualified address, that address was qualified, even if it
3949 did not match any rewriting rules. The underlying bug was that the values
3950 of the flags that permit the existence of unqualified sender and recipient
3951 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3952 non-local messages, and by -bnq for local messages) were not being
3953 preserved with the message after it was received.
3955 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3956 "next input=" as part of the text comprising the host identity instead of
3957 the correct text. The code was using the same buffer for two different
3958 strings. However, depending on which order the printing function evaluated
3959 its arguments, the bug did not always show up. Under Linux, for example, my
3960 test suite worked just fine.
3962 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3963 doesn't work with some older versions of Perl. It has been changed to "my",
3964 which in any case is probably the better facility to use.
3966 15. A really picky compiler found some instances of statements for creating
3967 error messages that either had too many or two few arguments for the format
3970 16. The size of the buffer for calls to the DNS resolver has been increased
3971 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3972 for addresses that have a lot of PTR records. This alleviates a problem; it
3973 does not fully solve it.
3975 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3976 buffer now truncates the list and logs the incident, which is the same
3977 action as happens when Exim is looking up a host name and its aliases.
3978 Previously in this situation something unpredictable would happen;
3979 sometimes it was "internal error: store_reset failed".
3981 18. If a server dropped the connection unexpectedly when an Exim client was
3982 using GnuTLS and trying to read a response, the client delivery process
3983 crashed while trying to generate an error log message.
3985 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3986 string, for example:
3988 warn message = H1: something\nH2: something
3990 the text was added as a single header line from Exim's point of view
3991 though it ended up OK in the delivered message. However, searching for the
3992 second and subsequent header lines using $h_h2: did not work. This has been
3993 fixed. Similarly, if a system filter added multiple headers in this way,
3994 the routers could not see them.
3996 20. Expanded the error message when iplsearch is called with an invalid key to
3997 suggest using net-iplsearch in a host list.
3999 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4000 ACLs are no longer actually imposed (and a message to that effect is
4003 22. If a "gecos" field in a passwd entry contained escaped characters, in
4004 particular, if it contained a \" sequence, Exim got it wrong when building
4005 a From: or a Sender: header from that name. A second bug also caused
4006 incorrect handling when an unquoted " was present following a character
4007 that needed quoting.
4009 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4010 was not being matched caselessly.
4012 24. Arranged for all hyphens in the exim.8 source to be escaped with
4015 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4016 properly. Recipient callout cache records were still being keyed to include
4017 the sender, even when use_sender was set false. This led to far more
4018 callouts that were necessary. The sender is no longer included in the key
4019 when use_sender is false.
4021 26. Added "control = submission" modifier to ACLs.
4023 27. Added the ${base62d: operator to decode base 62 numbers.
4025 28. dnsdb lookups can now access SRV records.
4027 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4028 the configuration file.
4030 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4031 (-v) mode. This makes the output for a verbose queue run more intelligible.
4033 31. Added a use_postmaster feature to recipient callouts.
4035 32. Added the $body_zerocount variable, containing the number of binary zero
4036 bytes in the message body.
4038 33. The time of last modification of the "new" subdirectory is now used as the
4039 "mailbox time last read" when there is a quota error for a maildir
4042 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4044 35. Added +ignore_unknown as a special item in host lists.
4046 36. Code for decoding IPv6 addresses in host lists is now included, even if
4047 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4048 address was recognized as an IP address, but was then not correctly decoded
4049 into binary, causing unexpected and incorrect effects when compared with
4056 1. Very minor rewording of debugging text in manualroute to say "list of
4057 hosts" instead of "hostlist".
4059 2. If verify=header_syntax was set, and a header line with an unqualified
4060 address (no domain) and a large number of spaces between the end of the
4061 name and the colon was received, the reception process suffered a buffer
4062 overflow, and (when I tested it) crashed. This was caused by some obsolete
4063 code that should have been removed. The fix is to remove it!
4065 3. When running in the test harness, delay a bit after writing a bounce
4066 message to get a bit more predictability in the log output.
4068 4. Added a call to search_tidyup() just before forking a reception process. In
4069 theory, someone could use a lookup in the expansion of smtp_accept_max_
4070 per_host which, without the tidyup, could leave open a database connection.
4072 5. Added the variables $recipient_data and $sender_data which get set from a
4073 lookup success in an ACL "recipients" or "senders" condition, or a router
4074 "senders" option, similar to $domain_data and $local_part_data.
4076 6. Moved the writing of debug_print from before to after the "senders" test
4079 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4080 problems for message scanning, either using a data ACL, or using
4081 local_scan() because the Received: header was not generated till after they
4082 were called (in order to set the time as the time of reception completion).
4083 I have revised the way this works. The header is now generated after the
4084 body is received, but before the ACL or local_scan() are called. After they
4085 are run, the timestamp in the header is updated.
4091 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4092 before starting a queue runner without re-exec. This happened only when
4093 deliver_drop_privilege was set or when the Exim user was set to root. The
4094 effect of the bug was that timeouts during subsequent deliveries caused
4095 crashes instead of being properly handled. The handler is now left at its
4096 default (and expected) setting.
4098 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4099 message, again when deliver_drop_privilege is set or Exim is run as root.
4100 The bug described in (1) was not present in this case, but the tidying up
4101 of the other signals was missing. I have made the two cases consistent.
4103 3. The ignore_target_hosts setting on a manualroute router was being ignored
4104 for hosts that were looked up using the /MX notation.
4106 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4109 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4110 operated on the sender address. After changing the $sender_address to <>
4111 for the sender address verify, Exim was re-instated it as the original
4112 (before rewriting) address, but remembering that it had rewritten it, so it
4113 wasn't rewriting it again. This bug also had the effect of breaking the
4114 sender address verification caching when the sender address was rewritten.
4116 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4117 This has been changed so that if the ip literal address matches
4118 ignore_target_hosts, the router declines.
4120 7. Added expansion conditions match_domain, match_address, and match_local_
4121 part (NOT match_host).
4123 8. The placeholder for the Received: header didn't have a length field set.
4125 9. Added code to Exim itself and to exim_lock to test for a specific race
4126 condition that could lead to file corruption when using MBX delivery. The
4127 issue is with the lockfile that is created in /tmp. If this file is removed
4128 after a process has opened it but before that process has acquired a lock,
4129 there is the potential for a second process to recreate the file and also
4130 acquire a lock. This could lead to two Exim processes writing to the file
4131 at the same time. The added code performs the same test as UW imapd; it
4132 checks after acquiring the lock that its file descriptor still refers to
4133 the same named file.
4135 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4136 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4139 11. Added the smtp_active_hostname option. If used, this will typically be made
4140 to depend on the incoming interface address. Because $interface_address is
4141 not set up until the daemon has forked a reception process, error responses
4142 that can happen earlier (such as "too many connections") no longer contain
4145 12. If an expansion in a condition on a "warn" statement fails because a lookup
4146 defers, the "warn" statement is abandoned, and the next ACL statement is
4147 processed. Previously this caused the whole ACL to be aborted.
4149 13. Added the iplsearch lookup type.
4151 14. Added ident_timeout as a log selector.
4153 15. Added tls_certificate_verified as a log selector.
4155 16. Added a global option tls_require_ciphers (compare the smtp transport
4156 option of the same name). This controls incoming TLS connections.
4158 17. I finally figured out how to make tls_require_ciphers do a similar thing
4159 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4160 before starting the TLS session.
4162 18. Tabs are now shown as \t in -bP output.
4164 19. If the log selector return_path_on_delivery was set, Exim crashed when
4165 bouncing a message because it had too many Received: header lines.
4167 20. If two routers both had headers_remove settings, and the first one included
4168 a superfluous trailing colon, the final name in the first list and the
4169 first name in the second list were incorrectly joined into one item (with a
4170 colon in the middle).
4176 1. Added -C and -D options to the exinext utility, mainly to make it easier
4177 to include in the automated testing, but these could be helpful when
4178 multiple configurations are in use.
4180 2. The exinext utility was not formatting the output nicely when there was
4181 an alternate port involved in the retry record key, nor when there was a
4182 message id as well (for retries that were specific to a specific message
4183 and a specific host). It was also confused by IPv6 addresses, because of
4184 the additional colons they contain. I have fixed the IPv4 problem, and
4185 patched it up to do a reasonable job for IPv6.
4187 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4188 delivery, the log line now contains "pipelined" if PIPELINING was used.
4190 4. An SMTP transport process used to panic and die if the bind() call to set
4191 an explicit outgoing interface failed. This has been changed; it is now
4192 treated in the same way as a connect() failure.
4194 5. A reference to $sender_host_name in the part of a conditional expansion
4195 that was being skipped was still causing a DNS lookup. This no longer
4198 6. The def: expansion condition was not recognizing references to header lines
4199 that used bh_ and bheader_.
4201 7. Added the _cache feature to named lists.
4203 8. The code for checking quota_filecount in the appendfile transport was
4204 allowing one more file than it should have been.
4206 9. For compatibility with Sendmail, the command line option
4214 and sets the incoming protocol and host name (for trusted callers). The
4215 host name and its colon can be omitted when only the protocol is to be set.
4216 Note the Exim already has two private options, -pd and -ps, that refer to
4217 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4218 "s", but I don't think that's a major issue.
4220 10. A number of refactoring changes to the code, none of which should affect
4223 (a) The number of logging options was getting close to filling up the
4224 32-bit word that was used as a bit map. I have split them into two classes:
4225 those that are passed in the argument to log_write(), and those that are
4226 only ever tested independently outside of that function. These are now in
4227 separate 32-bit words, so there is plenty of room for expansion again.
4228 There is no change in the user interface or the logging behaviour.
4230 (b) When building, for example, log lines, the code previously used a
4231 macro that called string_cat() twice, in order to add two strings. This is
4232 not really sufficiently general. Furthermore, there was one instance where
4233 it was actually wrong because one of the argument was used twice, and in
4234 one call a function was used. (As it happened, calling the function twice
4235 did not affect the overall behaviour.) The macro has been replaced by a
4236 function that can join an arbitrary number of extra strings onto a growing
4239 (c) The code for expansion conditions now uses a table and a binary chop
4240 instead of a serial search (which was left over from when there were very
4241 few conditions). Also, it now recognizes conditions like "pam" even when
4242 the relevant support is not compiled in: a suitably worded error message is
4243 given if an attempt is made to use such a condition.
4245 11. Added ${time_interval:xxxxx}.
4247 12. A bug was causing one of the ddress fields not to be passed back correctly
4248 from remote delivery subprocesses. The field in question was not being
4249 subsequently used, so this caused to problems in practice.
4251 13. Added new log selectors queue_time and deliver_time.
4253 14. Might have fixed a bug in maildirsizefile handling that threw up
4254 "unexpected character" debug warnings, and recalculated the data
4255 unnecessarily. In any case, I expanded the warning message to give more
4258 15. Added the message "Restricted characters in address" to the statements in
4259 the default ACL that block characters like @ and % in local parts.
4261 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4262 Three changes have been made:
4264 (a) There was a serious bug; a negative response to MAIL caused the whole
4265 recipient domain to be cached as invalid, thereby blocking all messages
4266 to all local parts at the same domain, from all senders. This bug has
4267 been fixed. The domain is no longer cached after a negative response to
4268 MAIL if the sender used is not empty.
4270 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4273 (c) A new callout option, "use_sender" has been added for people who want
4274 the modified behaviour.
4280 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4283 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4284 indeed breaks things for older releases.
4286 3. Added additional logging to the case where there is a problem reading data
4287 from a filter that is running in a subprocess using a pipe, in order to
4288 try to track down a specific problem.
4290 4. Testing facility fudge: when running in the test harness and attempting
4291 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4292 getting "No route to host". Convert this to a timeout.
4294 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4297 6. Some OS don't have socklen_t but use size_t instead. This affects the
4298 fifth argument of getsockopt() amongst other things. This is now
4299 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4300 can be set for individual OS. I have set it for SunOS5, OSF1, and
4301 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4302 some earlier ones do not.
4304 7. Change 4.30/15 was not doing the test caselessly.
4306 8. The standard form for an IPv6 address literal was being rejected by address
4307 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4308 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4309 this, as well as the form without the "IPv6" on the front (but only when
4310 address literals are enabled, of course).
4312 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4314 10. Exim crashed if a message with an empty sender address specified by -f
4315 encountered a router with an errors_to setting. This could be provoked only
4316 by a command such as
4320 where an empty string was supplied; "<>" did not hit this bug.
4322 11. Installed PCRE release 4.5.
4324 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4325 remained set. It is now erased.
4327 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4328 times from message ids (which are base 36 rather than the normal 62).
4330 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4331 were being counted as actual protocol errors, and logged if the log
4332 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4333 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4334 and DATA following a set of rejected RCPTs do not count as protocol errors.
4335 In other words, Exim assumes they were pipelined, though this may not
4336 actually be the case. Of course, in all cases the client gets an
4337 appropriate error code.
4339 15. If a lookup fails in an ACL condition, a message about the failure may
4340 be available; it is used if testing the ACL cannot continue, because most
4341 such messages specify what the cause of the deferral is. However, some
4342 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4343 that caused an old message to be retained and used if a later statement
4344 caused a defer, replacing the real cause of the deferral.
4346 16. If an IP address had so many PTR records that the DNS lookup buffer
4347 was not large enough to hold them, Exim could crash while trying to process
4348 the truncated data. It now detects and logs this case.
4350 17. Further to 4.21/58, another change has been made: if (and only if) the
4351 first line of a message (the first header line) ends with CRLF, a bare LF
4352 in a subsequent header line has a space inserted after it, so as not to
4353 terminate the header.
4355 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4356 unnecessarily, used atoi() instead of strtol(), and didn't check the
4357 termination when getting file sizes from file names by regex.
4359 19. Completely re-implemented the support for maildirsize files, in the light
4360 of a number of problems with the previous contributed implementation
4361 (4.30/29). In particular:
4363 . If the quota is zero, the maildirsize file is maintained, but no quota is
4366 . If the maildir directory does not exist, it is created before any attempt
4367 to write a maildirsize file.
4369 . The quota value in the file is just a cache; if the quota is changed in
4370 the transport, the new value overrides.
4372 . A regular expression is available for excluding directories from the
4375 20. The autoreply transport checks the characters in options that define the
4376 message's headers; it allows continued headers, but it was checking with
4377 isspace() after an embedded newline instead of explicitly looking for a
4380 21. If all the "regular" hosts to which an address was routed had passed their
4381 expiry times, and had not reached their retry times, the address was
4382 bounced, even if fallback hosts were defined. Now Exim should go on to try
4385 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4386 equivalent code in the SMTP transport. Some hosts send humungous responses
4387 to HELO/EHLO, more than 1024 it seems.
4389 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4390 gives compiler warnings in some environments. I've now done it "properly",
4393 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4394 (because of problems with the built-in one) was declared to return uschar *
4395 instead of char *, causing compiler failure.
4397 25. Fixed a file descriptor leak when processing alias/forward files.
4399 26. Fixed a minor format string issue in dbfn.c.
4401 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4403 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4404 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4407 29. When a sender address is verified, it is cached, to save repeating the test
4408 when there is more than one recipient in a message. However, when the
4409 verification involves a callout, it is possible for different callout
4410 options to be set for different recipients. It is too complicated to keep
4411 track of this in the cache, so now Exim always runs a verification when a
4412 callout is required, relying on the callout cache for the optimization.
4413 The overhead is duplication of the address routing, but this should not be
4416 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4417 to be verified with callout=postmaster, and the main callout worked but the
4418 postmaster check failed, the verification correctly failed. However, if a
4419 subsequent RCPT command asked for sender verification *without* the
4420 postmaster check, incorrect caching caused this verification also to fail,
4423 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4424 it was not caching the DNS options (qualify_single, search_parents) that
4425 were used when the lookup failed. A subsequent lookup with different
4426 options therefore always gave the same answer, though there were cases
4427 where it should not have. (Example: a "domains = !$mx_any" option on a
4428 dnslookup router: the "domains" option is always processed without any
4429 widening, but the router might have qualify_single set.) Now Exim uses the
4430 cached value only when the same options are set.
4432 32. Added John Jetmore's "exipick" utility to the distribution.
4434 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4435 than a timeout (e.g. a certificate is required, and is not provided), an
4436 Exim server now closes the connection immediately. Previously it waited for
4437 the client to close - but if the client is SSL, it seems that they each
4438 wait for each other, leading to a delay before one of them times out.
4440 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4441 maintained 0.8.x compatibility because I don't think many are using it, and
4442 it is clearly obsolete.
4444 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4447 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4448 client certificate was expired. A simple patch fixes this, though I don't
4449 understand the full logic of why the verify callback is called multiple
4452 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4453 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4454 which causes problems with some clients (such as the Certicom SSL Plus
4455 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4456 disables the coutermeasure allowing Eudora to connect."
4458 38. Exim was not checking that a write() to a log file succeeded. This could
4459 lead to Bad Things if a log got too big, in particular if it hit a file
4460 size limit. Exim now panics and dies if it cannot write to a log file, just
4461 as it does if it cannot open a log file.
4463 39. Modified OS/Makefile-Linux so that it now contains
4465 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4467 The two -D definitions ensure that Exim is compiled with large file
4468 support, which makes it possible to handle log files that are bigger than
4471 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4472 instance) a domain was checked against a named list that involved a lookup,
4473 causing $domain_data to be set, then another domain was checked against the
4474 same list, then the first domain was re-checked, the value of $domain_data
4475 after the final check could be wrong. In particular, if the second check
4476 failed, it could be set empty. This bug probably also applied to
4479 41. The strip_trailing_dot option was not being applied to the address given
4480 with the -f command-line option.
4482 42. The code for reading a message's header from the spool was incrementing
4483 $received_count, but never initializing it. This meant that the value was
4484 incorrect (doubled) while delivering a message in the same process in which
4485 it was received. In the most common configuration of Exim, this never
4486 happens - a fresh exec is done - but it can happen when
4487 deliver_drop_privilege is set.
4489 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4490 it now includes up to 150 characters of the unexpected data in the log
4493 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4494 and building data strings. The size of both of these buffers was 10 000
4495 bytes - far larger than anybody would *ever* want, thought I. Needless to
4496 say, somebody hit the limit. I have increased the maximum line length to
4497 20 000 and the maximum data length of concatenated lines to 100 000. I have
4498 also fixed two bugs, because there was no checking on these buffers. Tsk,
4499 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4500 buffer is too small.
4502 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4503 lsearch lookups. Now it does.
4505 46. When parsing a route_list item in a manualroute router, a fixed-length
4506 buffer was used for the list of hosts. I made this 1024 bytes long,
4507 thinking that nobody would ever have a list of hosts that long. Wrong.
4508 Somebody had a whole pile of complicated expansion conditions, and the
4509 string was silently truncated, leading to an expansion error. It turns out
4510 that it is easier to change to an unlimited length (owing to other changes
4511 that have happened since this code was originally written) than to build
4512 structure for giving a limitation error. The length of the item that
4513 expands into the list of hosts is now unlimited.
4515 47. The lsearch lookup could not handle data where the length of text line was
4516 more than 4095 characters. Such lines were truncated, leading to shortened
4517 data being returned. It should now handle lines of any length.
4519 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4520 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4523 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4524 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4525 no longer happens in comment lines. A list of replacements is now placed
4526 at the head of all of the source files, except those whose only change is
4527 to replace PERL_COMMAND in the very first #! line.
4529 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4530 messages on the queue, with a bottom-up merge sort, using code contributed
4531 by Michael Haardt. This should make operations like -bp somewhat faster on
4532 large queues. It won't affect queue runners, except when queue_run_in_order
4535 51. Installed eximstats 1.31 in the distribution.
4537 52. Added support for SRV lookups to the dnslookup router.
4539 53. If an ACL referred to $message_body or $message_body_end, the value was not
4540 reset for any messages that followed in the same SMTP session.
4542 54. The store-handling optimization for building very long strings was not
4543 differentiating between the different store pools. I don't think this
4544 actually made any difference in practice, but I've tidied it.
4546 55. While running the routers to verify a sender address, $sender_address
4547 was still set to the sender address. This is wrong, because when routing to
4548 send a bounce to the sender, it would be empty. Therefore, I have changed
4549 it so that, while verifying a sender address, $sender_address is set to <>.
4550 (There is no change to what happens when verifying a recipient address.)
4552 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4553 target A or AAAA records (if not already returned) without resetting the
4554 qualify_single or search_parents options of the DNS resolver. These are
4555 inappropriate in this case because the targets of MX and SRV records must
4556 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4557 target that, when qualified, matched something in the local domain. These
4558 two options are now turned off when doing these lookups.
4560 57. It seems that at least some releases of Reiserfs (which does not have the
4561 concept of a fixed number of inodes) returns zero and not -1 for the
4562 number of available inodes. This interacted badly with check_spool_inodes,
4563 which assumed that -1 was the "no such thing" setting. What I have done is
4564 to check that the total number of inodes is greater than zero before doing
4565 the test of how many are available.
4567 58. When a "warn" ACL statement has a log_message modifier, the message is
4568 remembered, and not repeated. This is to avoid a lot of repetition when a
4569 message has many recipients that cause the same warning to be written.
4570 Howewer, Exim was preserving the list of already written lines for an
4571 entire SMTP session, which doesn't seem right. The memory is now reset if a
4572 new message is started.
4574 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4575 debugging output unless log_rewrite was also set.
4577 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4578 of (int)(handle) when we know that handle contains (void *)(-1).
4580 61. The Exim daemon panic-logs an error return when it closes the incoming
4581 connection. However "connection reset by peer" seems to be common, and
4582 isn't really an error worthy of noting specially, so that particular error
4585 62. When Exim is trying to find all the local interfaces, it used to panic and
4586 die if the ioctl to get the interface flags failed. However, it seems that
4587 on at least one OS (Solaris 9) it is possible to have an interface that is
4588 included in the list of interfaces, but for which you get a failure error
4589 for this call. This happens when the interface is not "plumbed" into a
4590 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4591 failure of the "get flags" call assumes that the interface is down.
4593 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4594 makes life easier for people who are doing arithmetic on fields extracted
4595 from dates, where you often get leading zeros that should not be
4596 interpreted as octal.
4598 64. Added qualify_domain to the redirect router, to override the global
4601 65. If a pathologically long header line contained very many addresses (the
4602 report of this problem mentioned 10 000) and each of them was rewritten,
4603 Exim could use up a very large amount of memory. (It kept on making new
4604 copies of the header line as it rewrote, and never released the old ones.)
4605 At the expense of a bit more processing, the header rewriting function has
4606 been changed so that it no longer eats memory in this way.
4608 66. The generation of the Received: header has been moved from the time that a
4609 message starts to be received, to the time that it finishes. The timestamp
4610 in the Received: header should now be very close to that of the <= log
4611 line. There are two side-effects of this change:
4613 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4614 logged header lines no longer include the local Received: line, because
4615 it has not yet been created. The same applies to a copy of the message
4616 that is returned to a non-SMTP sender when a message is rejected.
4618 (b) When a filter file is tested using -bf, no additional Received: header
4619 is added to the test message. After some thought, I decided that this
4622 This change does not affect the value of $received_for. It is still set
4623 after address rewriting, but before local_scan() is called.
4625 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4627 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4628 gave an unhelpful panic error message, and a defer error. I have managed to
4629 change this behaviour so that it now rejects any supplied certificate,
4630 which seems right, as the list of acceptable certificates is empty.
4632 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4633 gave an unhelpful defer error. I have not managed to make this reject any
4634 supplied certificates, but the error message it gives is "no certificate
4635 supplied", which is not helpful.
4637 70. exigrep's output now also includes lines that are not associated with any
4638 message, but which match the given pattern. Implemented by a patch from
4639 Martin Sluka, which also tidied up the Perl a bit.
4641 71. Recipient callout verification, like sender verification, was using <> in
4642 the MAIL FROM command. This isn't really the right thing, since the actual
4643 sender may affect whether the remote host accepts the recipient or not. I
4644 have changed it to use the actual sender in the callout; this means that
4645 the cache record is now keyed on a recipient/sender pair, not just the
4646 recipient address. There doesn't seem to be a real danger of callout loops,
4647 since a callout by the remote host to check the sender would use <>.
4648 [SEE ABOVE: changed after hitting problems.]
4650 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4651 temporary errors. However, in the case of such a code being given after
4652 the end of a data transmission (i.e. after ".") Exim was failing to write
4653 a retry record for the message. (Yes, there was some broken host that was
4654 actually sending 8xx at this point.)
4656 73. An unknown lookup type in a host list could cause Exim to panic-die when
4657 the list was checked. (An example that provoked this was putting <; in the
4658 middle of a list instead of at the start.) If this happened during a DATA
4659 ACL check, a -D file could be left lying around. This kind of configuration
4660 error no longer causes Exim to die; instead it causes a defer errror. The
4661 incident is still logged to the main and panic logs.
4663 74. Buglet left over from Exim 3 conversion. The message "too many messages
4664 in one connection" was written to the rejectlog but not the mainlog, except
4665 when address rewriting (yes!) was being logged.
4667 75. Added write_rejectlog option.
4669 76. When a system filter was run not as root (that is, when system_filter_user
4670 was set), the values of the $n variables were not being returned to the
4671 main process; thus, they were not subsequently available in the $sn
4674 77. Added +return_path_on_delivery log selector.
4676 78. A connection timeout was being treated differently from recipients deferred
4677 when testing hosts_max_try with a message that was older than the host's
4678 retry timeout. (The host should not be counted, thus allowing all hosts to
4679 be tried at least once before bouncing.) This may have been the cause of an
4680 occasionally reported bug whereby a message would remain on the queue
4681 longer than the retry timeout, but would be bounced if a delivery was
4682 forced. I say "may" because I never totally pinned down the problem;
4683 setting up timeout/retry tests is difficult. See also the next item.
4685 79. The ultimate address timeout was not being applied to errors that involved
4686 a combination of host plus message (for example, a timeout on a MAIL
4687 command). When an address resolved to a number of possible hosts, and they
4688 were not all tried for each delivery (e.g. because of hosts_max_try), a
4689 message could remain on the queue longer than the retry timeout.
4691 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4694 81. Fixed an obscure SMTP outgoing bug which required at least the following
4695 conditions: (a) there was another message waiting for the same server;
4696 (b) the server returned 5xx to all RCPT commands in the first message so
4697 that the message was not completed; (c) the server dropped the connection
4698 or gave a negative response to the RSET that Exim sends to abort the
4699 transaction. The observed case was a dropped connection after DATA that had
4700 been sent in pipelining mode. That is, the server had advertised PIPELINING
4701 but was not implementing it correctly. The effect of the bug was incorrect
4702 behaviour, such as trying another host, and this could lead to a crash.
4708 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4709 and daemon.c were passed as pointers to ints; they should have been
4710 pointers to socklen_t variables (which are typically unsigned ints).
4712 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4715 3. Fixed a really odd bug that affected only the testing scheme; patching a
4716 certain fixed string in the binary changed the value of another string that
4717 happened to be identical to the end of the original first string.
4719 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4720 name", it returns that address as the IP address. On some operating
4721 systems (e.g. Solaris), it also passes back the IP address string as the
4722 "host name". However, on others (e.g. Linux), it passes back an empty
4723 string. Exim wasn't checking for this, and was changing the host name to an
4724 empty string, assuming it had been canonicized.
4726 5. Although rare, it is permitted to have more than one PTR record for a given
4727 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4728 all the names associated with an address, because they do in Solaris.
4729 However, it seems that they do not in Linux for data that comes from the
4730 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4731 I found this out when I moved to a new Linux workstation and tried to run
4732 the Exim test suite.
4734 To get round this problem I have changed the code so that it now does its
4735 own call to the DNS to look up PTR records when searching for a host name.
4736 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4737 addresses that are only in /etc/hosts are still found.
4739 This behaviour is, however, controlled by an option called host_lookup_
4740 order, which defaults to "bydns:byaddr". If people want to use the other
4741 order, or indeed, just use one or the other means of lookup, they can
4742 specify it in this variable.
4744 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4745 some operating systems, this comes back from gethostbyaddr() as an empty
4746 string, and this is what Exim used to test for. However, it seems that in
4747 other systems, "." is yielded. Exim now tests for this case too.
4749 7. The values of check_spool_space and check_log_space are now held internally
4750 as a number of kilobytes instead of an absolute number of bytes. If a
4751 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4752 kilobyte. This means that much larger values can be stored.
4754 8. Exim monitor: an attempt to get the action menu when not actually pointing
4755 at a message produces an empty menu entitled "No message selected". This
4756 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4757 no entries in it ("Shell widget menu has zero width and/or height"). So I
4758 have added a single, blank menu entry in this case.
4760 9. Added ${quote_local_part.
4762 10. MIME decoding is now applied to the contents of Subject: header lines when
4765 11. Now that a reference to $sender_host_address automatically causes a reverse
4766 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4767 host lookup before query-style lookups in lists that might use this
4768 variable. This has therefore been abolished, and the "net-" prefix is no
4769 longer necessary for query-style lookups.
4771 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4772 have been a typo for LFLAGS, so it has been changed.
4774 13. The install script calls Exim with "-C /dev/null" in order to find the
4775 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4776 to be output. Howeve, since Exim outputs its version number before the
4777 error, it didn't break the script. It just looked ugly. I fixed this by
4778 always allowing "-C /dev/null" if the caller is root.
4780 14. Ignore overlarge ACL variable number when reading spool file - insurance
4781 against a later release with more variables having written the file.
4783 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4784 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4785 this, as well as the form without the "IPv6" on the front.
4787 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4788 OS/Makefile-Darwin file.
4790 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4791 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4793 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4794 variable. However, some OS use ints for inodes. I've added cast to long int
4795 to get rid of the compiler warning.
4797 19. I had forgotten to lock out "/../" in configuration file names when
4798 ALT_CONFIG_PREFIX was set.
4800 20. Routers used for verification do not need to specify transports. However,
4801 if such a router generated a host list, and callout was configured, Exim
4802 crashed, because it could not find a port number from the (non-existent)
4803 transport. It now assumes port 25 in this circumstance.
4805 21. Added the -t option to exigrep.
4807 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4808 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4809 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4810 them is set, LOOKUP_LSEARCH is forced.
4812 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4814 24. Added sender and host information to the "rejected by local_scan()" log
4815 line; previously there was no indication of these.
4817 25. Added .include_if_exists.
4819 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4820 Linux. It turns out that not all file systems support this. Apparently some
4821 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4822 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4823 not supported on the file descriptor, is now ignored when Exim is trying to
4824 sync a directory. This applies only to Linux.
4826 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4828 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4829 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4832 29. Added support for maildirsize files from supplied patch (modified a bit).
4834 30. The use of :fail: followed by an empty string could lead Exim to respond to
4835 sender verification failures with (e.g.):
4837 550 Verification failed for <xxx>
4838 550 Sender verify failed
4840 where the first response line was missing the '-' that indicates it is not
4841 the final line of the response.
4843 31. The loop for finding the name of the user that called Exim had a hardwired
4844 limit of 10; it now uses the value of finduser_retries, which is used for
4845 all other user lookups.
4847 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4850 33. Exim was neglecting to zero errno before one call of strtol() when
4851 expanding a string and expecting an integer value. On some systems this
4852 resulted in spurious "integer overflow" errors. Also, it was casting the
4853 result into an int without checking.
4855 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4856 did not work. The code looks as if it has *never* worked, though it appears
4857 to have been documented since at least releast 1.62. I have made it work.
4859 35. The "timeout_DNS" error in retry rules, also documented since at least
4860 1.62, also never worked. As it isn't clear exactly what this means, and
4861 clearly it isn't a major issue, I have abolished the feature by treating it
4862 as "timeout", and writing a warning to the main and panic logs.
4864 36. The display of retry rules for -brt wasn't always showing the error code
4867 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4868 timeout_connect_A, timeout_connect_MX.
4870 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4871 to the empty sender.
4873 39. The daemon was not analysing the content of -oX till after it had closed
4874 stderr and disconnected from the controlling terminal. This meant that any
4875 syntax errors were only noted on the panic log, and the return code from
4876 the command was 0. By re-arranging the code a little, I've made the
4877 decoding happen first, so such errors now appear on stderr, and the return
4878 code is 1. However, the actual setting up of the sockets still happens in
4879 the disconnected process, so errors there are still only recorded on the
4882 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4883 connections (as happens on some IP stacks) was logged at start up time as
4884 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4885 it from "IPv6 and IPv4", which means that two separate sockets are being
4888 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4889 says whether AF_INET or AF_INET6 was passed as an argument.
4891 42. Exiwhat output was messed up when time zones were included in log
4894 43. Exiwhat now gives more information about the daemon's listening ports,
4895 and whether -tls-on-connect was used.
4897 44. The "port" option of the smtp transport is now expanded.
4899 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4900 silently ignored. Now an error message is written to the main and panic
4903 46. There's a new ACL modifier called "logwrite" which writes to a log file
4904 as soon as it is encountered.
4906 47. Added $local_user_uid and $local_user_gid at routing time.
4908 48. Exim crashed when trying to verify a sender address that was being
4911 49. Exim was recognizing only a space character after ".include". It now also
4912 recognizes a tab character.
4914 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4915 extracting the relevant information from the specification. The man page no
4916 longer contains scrambled data for the -d option, and I've added a section
4917 at the front about calling Exim under different names.
4919 51. Added "extra_headers" argument to the "mail" command in filter files.
4921 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4924 53. Installed eximstats 1.29.
4926 54. Added transport_filter_timeout as a generic transport option.
4928 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4929 Cc: header lines. This was required by RFC 822, but it not required by RFC
4932 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4933 incoming messages that did not have them. Now it does so only if the
4934 message originates locally, that is, if there is no associated remote host
4935 address. When Resent- header lines are present, this applies to the Resent-
4936 lines rather than the non-Resent- lines.
4938 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4939 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4941 58. Messages for configuration errors now include the name of the main
4942 configuration file - useful now that there may be more than one file in a
4943 list (.included file names were always shown).
4945 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4946 for those rare installations that do not start the daemon as root or run it
4947 setuid root. I've cut out the call to initgroups() if the daemon is not
4950 60. The Exim user and group can now be bound into the binary as text strings
4951 that are looked up at the start of Exim's processing.
4953 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4955 62. Added $mailstore_basename variable.
4957 63. Installed patch to sieve.c from Michael Haardt.
4959 64. When Exim failed to open the panic log after failing to open the main log,
4960 the original message it was trying to log was written to stderr and debug
4961 output, but if they were not available (the usual case in production), it
4962 was lost. Now it is written to syslog before the two lines that record the
4963 failures to open the logs.
4965 65. Users' Exim filters run in subprocesses under the user's uid. It is
4966 possible for a "deliver" command or an alias in a "personal" command to
4967 provoke an address rewrite. If logging of address rewriting is configured,
4968 this fails because the process is not running as root or exim. There may be
4969 a better way of dealing with this, but for the moment (because 4.30 needs
4970 to be released), I have disabled address rewrite logging when running a
4971 filter in a non-root, non-exim process.
4977 1. The buildconfig auxiliary program wasn't quoting the value set for
4978 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4979 not defined. This bug was masked in 4.22 by the effect that was fixed in
4982 2. Some messages that were rejected after a message id was allocated were
4983 shown as "incomplete" by exigrep. It no longer does this for messages that
4984 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4986 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4987 have allow_domain_literals set, the ID did not get logged in the <= line.
4988 Domain literals are now always recognized in Message-ID: header lines.
4990 4. The first argument for a ${extract expansion item is the key name or field
4991 number. Leading and trailing spaces in this item were not being ignored,
4992 causing some misleading effects.
4994 5. When deliver_drop_privilege was set, single queue runner processes started
4995 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4996 same command in the process it spins off) were not dropping privilege.
4998 6. When the daemon running as "exim" started a queue runner, it always
4999 re-executed Exim in the spun-off process. This is a waste of effort when
5000 deliver_drop_privilege is set. The new process now just calls the
5001 queue-runner function directly.
5007 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5010 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5011 ignored. Though the use of -oP was forcing the writing of a pid file, it
5012 was always written to the default place.
5014 3. If the message "no IP address found for host xxxx" is generated during
5015 incoming verification, it is now followed by identification of the incoming
5016 connection (so you can more easily find what provoked it).
5018 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5020 5. Added some features to "harden" Exim a bit more against certain attacks:
5022 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5023 be put in Local/Makefile. This is like the never_users runtime option,
5024 but it cannot be overridden. The default setting is "root".
5026 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5027 prefix string with which any file named in a -C command line option
5030 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5031 is retained for -C and -D only if the caller of Exim is root. Without
5032 it, the exim user may also use -C and -D and retain privilege.
5034 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5035 command line option is disabled.
5037 6. Macro names set by the -D option must start with an upper case letter, just
5038 like macro names defined in the configuration file.
5040 7. Added "dereference=" facility to LDAP.
5042 8. Two instances of the typo "uknown" in the source files are fixed.
5044 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5045 the Configure-Makefile script screwed up while processing it.
5047 10. Incorporated PCRE 4.4.
5049 11. The SMTP synchronization check was not operating right at the start of an
5050 SMTP session. For example, it could not catch a HELO sent before the client
5051 waited for the greeting. There is now a check for outstanding input at the
5052 point when the greeting is written. Because of the duplex, asynchronous
5053 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5054 way, but not yet received, when the check is performed.
5056 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5057 on TCP/IP sockets, because this apparently causes some broken clients to
5060 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5061 unchanged) from the Cygwin maintainer.
5063 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5064 was supported instead of testing for mbx. Effectively a typo.
5066 15. The spa authenticator server code was not checking that the input it
5067 received was valid base64.
5069 16. The debug output line for the "set" modifier in ACLs was not showing the
5070 name of the variable that was being set.
5072 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5074 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5075 lookup on its own. Something else was needed to trigger the lookup. For
5076 example, a match in host_lookup or the need for a host name in a host list.
5077 Now, if $sender_host_name is referenced and the host name has not yet been
5078 looked up, a lookup is performed. If the lookup fails, the variable remains
5079 empty, and $host_lookup_failed is set to "1".
5081 19. Added "eqi" as a case-independent comparison operator.
5083 20. The saslauthd authentication condition could segfault if neither service
5084 nor realm was specified.
5086 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5087 error message that was logged was misleading, and incoming SMTP
5088 connections were dropped. The message is now more accurate, and temporary
5089 errors are given to SMTP connections.
5091 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5092 (except RSET) with a 421 error, until QUIT is received. However, it was
5093 failing to send a response to QUIT.
5095 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5096 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5097 ACL is now run after the helo_try_verify_hosts code.
5099 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5100 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5101 case-independent, so other case variants are also recognized). Apparently
5102 some systems use these upper case variants.
5104 25. If more than two messages were waiting for the same host, and a transport
5105 filter was specified for the transport, Exim sent two messages over the
5106 same TCP/IP connection, and then failed with "socket operation on non-
5107 socket" when it tried to send the third.
5109 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5111 27. The extern definition of crypt16() in expand.c was not being excluded when
5112 the OS had its own crypt16() function.
5114 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5115 as a preferred synonym for return_size_limit, both as an option and as an
5118 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5120 30. Changed the default configuration ACL to relax the local part checking rule
5121 for addresses that are not in any local domains. For these addresses,
5122 slashes and pipe symbols are allowed within local parts, but the sequence
5123 /../ is explicitly forbidden.
5125 31. SPA server authentication was not clearing the challenge buffer before
5128 32. log_message in a "warn" ACL statement was writing to the reject log as
5129 well as to the main log, which contradicts the documentation and doesn't
5130 seem right (because no rejection is happening). So I have stopped it.
5132 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5133 However, I am unable to do any testing of this.
5135 34. Fixed an infelicity in the appendfile transport. When checking directories
5136 for a mailbox, to see if any needed to be created, it was accidentally
5137 using path names with one or more superfluous leading slashes; tracing
5138 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5140 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5141 contents are added to the log line that is written for every discarded
5142 recipient. (Previously a log_message setting was ignored.)
5144 36. The ${quote: operator now quotes the string if it is empty.
5146 37. The install script runs exim in order to find its version number. If for
5147 some reason other than non-existence or emptiness, which it checks, it
5148 could not run './exim', it was installing it with an empty version number,
5149 i.e. as "exim-". This error state is now caught, and the installation is
5152 38. An argument was missing from the function that creates an error message
5153 when Exim fails to connect to the socket for saslauthd authentication.
5154 This could cause Exim to crash, or give a corrupted message.
5156 39. Added isip, isip4, and isip6 to ${if conditions.
5158 40. The ACL variables $acl_xx are now saved with the message, and can be
5159 accessed later in routers, transports, and filters.
5161 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5162 strings in the file are not string-expanded.
5164 42. If a MAIL command specified a SIZE value that was too large to fit into an
5165 int variable, the check against message_size_limit failed. Such values are
5166 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5167 day this will have to be increased, but I don't think I want to be around
5168 when emails are that large.
5175 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5176 iconv() is not standard in FreeBSD.
5178 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5179 IPv6 enabled. The observed symptom was a segmentation fault on return from
5180 the function os_common_find_running_interfaces() in src/os.c.
5182 3. In the check_special_case() function in daemon.c I had used "errno" as an
5183 argument name, which causes warnings on some systems. This was basically a
5184 typo, since it was named "eno" in the comments!
5186 4. The code that waits for the clock to tick (at a resolution of some fraction
5187 of a second) so as to ensure message-id uniqueness was always waiting for
5188 at least one whole tick, when it could have waited for less. [This is
5189 almost certainly not relevant at current processor speeds, where it is
5190 unlikely to ever wait at all. But we try to future-proof.]
5192 5. The function that sleeps for a time interval that includes fractions of a
5193 second contained a race. It did not block SIGALRM between setting the
5194 timer, and suspending (a couple of lines later). If the interval was short
5195 and the sigsuspend() was delayed until after it had expired, the suspension
5196 never ended. On busy systems this could lead to processes getting stuck for
5199 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5200 process, before it forks any delivery processes. The open lookup caching
5201 mechanism meant that the open file or database connection was passed into
5202 the delivery process. The problem was that delivery processes always tidy
5203 up cached lookup data. This could cause a problem for the next delivery
5204 process started by the queue runner, because the external queue runner
5205 process does not know about the closure. So the next delivery process
5206 still has data in the lookup cache. In the case of a file lookup, there was
5207 no problem because closing a file descriptor in a subprocess doesn't affect
5208 the parent. However, if the lookup was caching a connection to a database,
5209 the connection was closed, and the second delivery process was likely to
5210 see errors such as "PGSQL: query failed: server closed the connection
5211 unexpectedly". The problem has been fixed by closing all cached lookups
5212 in a queue runner before running a delivery process.
5214 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5215 seem to have the "const" qualifier which it has on other OS. I've
5218 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5219 *of the same type* (client or server) with the same public name that an
5220 error should be diagnosed.
5222 9. When Exim looked up a host name for an IP address, but failed to find the
5223 original IP address when looking up the host name (a safety check), it
5224 output the message "<ip address> does not match any IP for NULL", which was
5225 confusing, to say the least. The bug was that the host name should have
5226 appeared instead of "NULL".
5228 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5229 user that is built into the binary, and the -C or -D options is used, root
5230 privilege is dropped before the configuration file is read. In addition,
5231 logging is switched to stderr instead of the normal log files. If the
5232 configuration then re-defines the Exim user, the unprivileged environment
5233 is probably not what is expected, so Exim logs a panic warning message (but
5236 However, if deliver_drop_privilege is set, the unprivileged state may well
5237 be exactly what is intended, so the warning has been cut out in that case,
5238 and Exim is allowed to try to write to its normal log files.
5244 1. smtp_return_error_details was not giving details for temporary sender
5245 or receiver verification errors.
5247 2. Diagnose a configuration error if two authenticators have the same public
5250 3. Exim used not to create the message log file for a message until the first
5251 delivery attempt. This could be confusing when incoming messages were held
5252 for policy or load reasons. The message log file is now created at the time
5253 the message is received, and an initial "Received" line is written to it.
5255 4. The automatically generated man page for command line options had a minor
5256 bug that caused no ill effects; however, a more serious problem was that
5257 the procedure for building the man page automatically didn't always
5258 operate. Consequently, release 4.20 contains an out-of-date version. This
5259 shouldn't happen again.
5261 5. When building Exim with embedded Perl support, the script that builds the
5262 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5263 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5265 6. The freeze_tell option was not being used for messages that were frozen on
5266 arrival, either by an ACL or by local_scan().
5268 7. Added the smtp_incomplete_transaction log selector.
5270 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5271 was accepting AUTH without a new EHLO.
5273 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5274 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5277 10. Logging of TCP/IP connections (when configured) now happens in the main
5278 daemon process instead of the child process, so that the TCP/IP connection
5279 count is more accurate (but it can never be perfect).
5281 11. The use of "drop" in a nested ACL was not being handled correctly in the
5282 outer ACL. Now, if condition failure induced by the nested "drop" causes
5283 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5284 or "require"), the connection is dropped.
5286 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5287 that yield "discard" can now be used with an "accept" or a "discard" verb,
5288 but an error is generated for any others (because I can't see a useful way
5289 to define what should happen).
5291 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5292 are now processed in the same way as lines in the Exim configuration file.
5293 In particular, continuation lines are supported.
5295 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5297 15. Added -ti meaning -t -i.
5299 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5300 domains, and warn by logging if others are found.
5302 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5303 returned from GIFCONF when Exim is trying to find the list of interfaces on
5304 a host. The code in os.c has been modified to copy each ifreq to an aligned
5305 structure in all cases.
5307 Also, in some cases, the returned ifreq's were being copied to a 'struct
5308 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5309 means the last couple of bytes of an IPv6 address could be chopped if the
5310 ifreq contained only a normal sockaddr (14 bytes storage).
5312 18. Named domain lists were not supported in the hosts_treat_as_local option.
5313 An entry such as +xxxx was not recognized, and was treated as a literal
5316 19. Ensure that header lines added by a DATA ACL are included in the reject log
5317 if the ACL subsequently rejects the message.
5319 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5320 MD5 (which is deprecated).
5322 21. When testing a filter file using -bf, Exim was writing a message when it
5323 took the sender from a "From " line in the message, but it was not doing so
5324 when it took $return_path from a Return-Path: header line. It now does.
5326 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5327 with a valid header line field name (a series of printing characters
5328 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5330 23. Changed "disc" in the source to "disk" to conform to the documentation and
5331 the book and for uniformity.
5333 24. Ignore Sendmail's -Ooption=value command line item.
5335 25. When execve() failed while trying to run a command in a pipe transport,
5336 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5337 could be confused with a return value of 69 from the command itself. This
5338 has been changed to 127, the value the shell returns if it is asked to run
5339 a non-existent command. The wording for the related log line suggests a
5340 non-existent command as the problem.
5342 26. If received_header_text expands to an empty string, do not add a Received:
5343 header line to the message. (Well, it adds a token one on the spool, but
5344 marks it "old" so that it doesn't get used or transmitted.)
5346 27. Installed eximstats 1.28 (addition of -nt option).
5348 28. There was no check for failure on the call to getsockname() in the daemon
5349 code. This can fail if there is a shortage of resources on the system, with
5350 ENOMEM, for example. A temporary error is now given on failure.
5352 29. Contrary to the C standard, it seems that in some environments, the
5353 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5354 program. Exim now does this explicitly; it affects the formatting of
5355 timestamps using strftime().
5357 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5358 complaints. I've now initialized all the variables, to avoid this.
5360 32. Header lines added by a system filter were not being "seen" during
5361 transport-time rewrites.
5363 33. The info_callback() function passed to OpenSSL is set up with type void
5364 (*)(SSL *, int, int), as described somewhere. However, when calling the
5365 function (actually a macro) that sets it up, the type void(*)() is
5366 expected. I've put in a cast to prevent warnings from picky compilers.
5368 34. If a DNS black list lookup found a CNAME record, but there were no A
5369 records associated with the domain it pointed at, Exim crashed.
5371 35. If a DNS black list lookup returned more than one A record, Exim ignored
5372 all but the first. It now scans all returned addresses if a particular IP
5373 value is being sought. In this situation, the contents of the
5374 $dnslist_value variable are a list of all the addresses, separated by a
5377 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5378 to accept a host name and all its aliases if the forward lookup for any of
5379 them yielded the IP address of the incoming connection. Now it accepts only
5380 those names whose forward lookup yields the correct IP address. Any other
5381 names are discarded. This closes a loophole whereby a rogue DNS
5382 administrator could create reverse DNS records to break through a
5383 wildcarded host restriction in an ACL.
5385 37. If a user filter or a system filter that ran in a subprocess used any of
5386 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5387 the wrong values were passed to the pipe command ($thisaddress had the
5388 value of $0, $0 had the value of $1, etc). This bug was introduced by
5389 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5391 38. Improved the line breaking for long SMTP error messages from ACLs.
5392 Previously, if there was no break point between 40 and 75 characters, Exim
5393 left the rest of the message alone. Two changes have been made: (a) I've
5394 reduced the minimum length to 35 characters; (b) if it can't find a break
5395 point between 35 and 75 characters, it looks ahead and uses the first one
5396 that it finds. This may give the occasional overlong line, but at least the
5397 remaining text gets split now.
5399 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5400 file descriptors might be low, and that setting 1000 would always raise it.
5401 It turns out that in some environments, the limit is already over 1000 and
5402 that lowering it causes trouble. So now Exim takes care not to decrease it.
5404 40. When delivering a message, the value of $return_path is set to $sender_
5405 address at the start of routing (routers may change the value). By an
5406 oversight, this default was not being set up when an address was tested by
5407 -bt or -bv, which affected the outcome if any router or filter referred to
5410 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5411 log only when "message" or "log_message" are set. However, if one of the
5412 conditions was an address verification, or a call to a nested ACL, the
5413 messages generated by the underlying test were being passed through. This
5414 no longer happens. The underlying message is available in $acl_verify_
5415 message for both "message" and "log_message" expansions, so it can be
5416 passed through if needed.
5418 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5419 new expansion $bh_ to give the encoded byte string without charset
5420 translation. Translation happens only if iconv() is available; HAVE_ICONV
5421 indicates this at build time. HEADERS_CHARSET gives the charset to
5422 translate to; headers_charset can change it in the configuration, and
5423 "headers charset" can change it in an individual filter file.
5425 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5426 that creates RFC 2047 encoded "words" labels them as that charset instead
5427 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5428 expansion operator; (ii) when Exim creates a From: line for a local
5429 message; (iii) when a header line is rewritten to include a "phrase" part.
5431 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5432 buggy, causing it to skip the first lines of messages whose message ID
5433 ended in 'D'. This would not have bitten before Exim release 4.14, because
5434 message IDs were unlikely to end in 'D' before then. The effect was to have
5435 incorrect size information for certain domains.
5437 45. #include "config.h" was missing at the start of the crypt16.c module. This
5438 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5441 46. If there was a timeout during a "random" callout check, Exim treated it as
5442 a failure of the random address, and carried on sending RSET and the real
5443 address. If the delay was just some slowness somewhere, the response to the
5444 original RCPT would be taken as a response to RSET and so on, causing
5445 mayhem of various kinds.
5447 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5448 when I implemented it. It didn't allow for the fact that some option values
5449 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5450 the right test for positive values.
5452 48. Domain names in DNS records are case-independent. Exim always looks them up
5453 in lower case. Some resolvers return domain names in exactly the case they
5454 appear in the zone file, that is, they may contain uppercase letters. Not
5455 all resolvers do this - some return always lower case. Exim was treating a
5456 change of case by a resolver as a change of domain, similar to a widening
5457 of a domain abbreviation. This triggered its re-routing code and so it was
5458 trying to route what was effectively the same domain again. This normally
5459 caused routing to fail (because the router wouldn't handle the domain
5460 twice). Now Exim checks for this case specially, and just changes the
5461 casing of the domain that it ultimately uses when it transmits the message
5464 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5467 50. If a filter generated a file delivery with a non-absolute name (possible if
5468 no home directory exists for the router), the forbid_file option was not
5471 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5472 the existing equality matching.
5474 52. Exim was using ints instead of ino_t variables in some places where it was
5475 dealing with inode numbers.
5477 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5478 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5479 called TMPDIR, and if it finds it is different, it changes its value.
5481 54. The smtp_printf() function is now made available to local_scan() so
5482 additional output lines can be written before returning. There is also an
5483 smtp_fflush() function to enable the detection of a dropped connection.
5484 The variables smtp_input and smtp_batched_input are exported to
5487 55. Changed the default runtime configuration: the message "Unknown user"
5488 has been removed from the ACL, and instead placed on the localuser router,
5489 using the cannot_route_message feature. This means that any verification
5490 failures that generate their own messages won't get overridden. Similarly,
5491 the "Unrouteable address" message that was in the ACL for unverifiable
5492 relay addresses has also been removed.
5494 56. Added hosts_avoid_esmtp to the smtp transport.
5496 57. The exicyclog script was not checking for the esoteric option
5497 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5498 will work only if exicyclog is run under the appropriate euid.
5500 58. Following a discussion on the list, the rules by which Exim recognises line
5501 endings on incoming messages have been changed. The -dropcr and drop_cr
5502 options are now no-ops, retained only for backwards compatibility. The
5503 following line terminators are recognized: LF CRLF CR. However, special
5504 processing applies to CR:
5506 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5507 nor a local message in the state where . is a terminator.
5509 (ii) If a bare CR is encountered in a header line, an extra space is added
5510 after the line terminator so as not to end the header. The reasoning
5511 behind this is that bare CRs in header lines are most likely either
5512 to be mistakes, or people trying to play silly games.
5514 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5515 was being incorrectly given as 18 bytes larger than it should have been.
5516 This is a VOB (very old bug).
5518 60. This may never have affected anything current, but just in case it has:
5519 When the local host is found other than at the start of a list of hosts,
5520 the local host, those with the same MX, and any that follow, are discarded.
5521 When the list in question was part of a longer list of hosts, the following
5522 hosts (not currently being processed) were also being discarded. This no
5523 longer happens. I'm not sure if this situation could ever has previously
5526 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5529 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5530 header. This is something that is recommended in a new Internet Draft, and
5531 is something that is documented as being done by Sendmail. There are two
5532 possible values. For messages generated by the autoreply transport, Exim
5535 Auto-Submitted: auto-replied
5537 whereas for all other generated messages (e.g. bounces) it adds
5539 Auto-Submitted: auto-generated
5541 63. The "personal" condition in filters now includes a test for the
5542 Auto-Submitted: header. If it contains the string "auto-" the message it
5543 not considered personal.
5545 64. Added rcpt_include_affixes as a generic transport option.
5547 65. Added queue_only_override (default true).
5549 66. Added the syslog_duplication option.
5551 67. If what should have been the first header line of a message consisted of
5552 a space followed by a colon, Exim was mis-interpreting it as a header line.
5553 It isn't of course - it is syntactically invalid and should therefore be
5554 treated as the start of the message body. The misbehaviour could have
5555 caused a number of strange effects, including loss of data in subsequent
5556 header lines, and spool format errors.
5558 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5559 client host had authenticated. This control can now be exercised by an ACL
5560 for more flexibility.
5562 69. By default, callouts do not happen when testing with -bh. There is now a
5563 variant, -bhc, which does actually run the callout code, including
5564 consulting and updating the callout cache.
5566 70. Added support for saslauthd authentication, courtesy of Alexander
5569 71. If statvfs() failed on the spool or log directories while checking their
5570 size for availability, Exim confusingly gave the error "space shortage".
5571 Furthermore, in debugging mode it crashed with a floating point exception.
5572 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5573 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5574 serious problem, Exim now writes to the main and panic logs when this
5575 happens, with details of the failure. It then refuses to accept the
5576 incoming message, giving the message "spool directory problem" or "log
5577 directory problem" with a 421 code for SMTP messages.
5579 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5580 0, 1, and 2 exist, because some OS complain for execs without them (see
5581 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5582 descriptors. However, the code omitted to check that the open succeeded,
5583 causing mysterious errors if for some reason the permissions on /dev/null
5584 got screwed. Now Exim writes a message to the main and panic logs, and
5585 bombs out if it can't open /dev/null.
5587 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5588 interact so that it is all more flexible. It is supposed to remain
5589 backwards compatible. Also added extra_local_interfaces.
5591 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5592 to bomb out with an assertion failure - to the client this appears as a
5593 connection drop. This problem occurs in the part of the code that was taken
5594 from the Samba project. Fortunately, the assertion is in a very simple
5595 function, so I have fixed this by reproducing the function inline in the
5596 one place where it is called, and arranging for authentication to fail
5597 instead of killing the process with assert().
5599 75. The SPA client code was not working when the server requested OEM rather
5600 than Unicode encoding.
5602 76. Added code to make require_files with a specific uid setting more usable in
5603 the case where statting the file as root fails - usually a non-root-mounted
5604 NFS file system. When this happens and the failure is EACCES, Exim now
5605 forks a subprocess and does the per-uid checking as the relevant uid.
5607 77. Added process_log_path.
5609 78. If log_file_path was not explicitly set, a setting of check_log_space or
5610 check_log_inodes was ignored.
5612 79. If a space check for the spool or log partitions fails, the incident is now
5613 logged. Of course, in the latter case the data may get lost...
5615 80. Added the %p formatting code to string_format() so that it can be used to
5616 print addresses in debug_print(). Adjusted all the address printing in the
5617 debugging in store.c to use %p rather than %d.
5619 81. There was a concern that a line of code in smtp_in.c could overflow a
5620 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5621 initially expressed, the concern was not well-founded, because trailing
5622 spaces are removed early. However, if the trailing spaces were followed by
5623 a NULL, they did not get removed, so the overflow was possible. Two fixes
5626 (a) I re-wrote the offending code in a cleaner fashion.
5627 (b) If an incoming SMTP command contains a NULL character, it is rejected
5630 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5631 runs initgroups(), so that if the Exim user is in any additional groups,
5632 they will be used during message reception.
5638 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob