1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
9 PP/02 Make -n do something, by making it not do something.
10 When combined with -bP, the name of an option is not output.
12 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
15 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
16 $sender_host_name and config options to manage this, and basic check
19 PP/05 DSCP support for outbound connections and control modifier for inbound.
21 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
22 (Only plugin which currently uses this is kerberos4, which nobody should
23 be using, but we should make it available and other future plugins might
24 conceivably use it, even though it would break NAT; stuff *should* be
25 using channel bindings instead).
27 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
28 name; added for Sendmail compatibility; requires admin caller.
29 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
30 just ignore it); requires trusted caller.
31 Also parse but ignore: -Ac -Am -X<logfile>
34 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
36 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
38 JH/02 Support "G" suffix to numbers in ${if comparisons.
43 PP/01 Handle short writes when writing local log-files.
44 In practice, only affects FreeBSD (8 onwards).
45 Bugzilla 1053, with thanks to Dmitry Isaikin.
47 NM/01 Bugzilla 949 - Documentation tweak
49 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
52 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
54 PP/02 Implemented gsasl authenticator.
56 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
58 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
59 `pkg-config foo` for cflags/libs.
61 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
62 with rest of GSASL and with heimdal_gssapi.
64 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
65 `pkg-config foo` for cflags/libs for the TLS implementation.
67 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
68 properties get this fed in as external SSF. A number of robustness
69 and debugging improvements to the cyrus_sasl authenticator.
71 PP/08 cyrus_sasl server now expands the server_realm option.
73 PP/09 Bugzilla 1214 - Log authentication information in reject log.
74 Patch by Jeremy Harris.
76 PP/10 Added dbmjz lookup type.
78 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
80 PP/12 MAIL args handles TAB as well as SP, for better interop with
81 non-compliant senders.
82 Analysis and variant patch by Todd Lyons.
84 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
85 Bug report from Lars Müller <lars@samba.org> (via SUSE),
86 Patch from Dirk Mueller <dmueller@suse.com>
88 PP/13 tls_peerdn now print-escaped for spool files.
89 Observed some $tls_peerdn in wild which contained \n, which resulted
90 in spool file corruption.
92 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
93 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
94 or write after TLS renegotiation, which otherwise led to messages
97 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
98 as a tracking header (ie: a signed header comes before the signature).
99 Patch from Wolfgang Breyha.
101 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
102 comma-sep list; embedded commas doubled.
104 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
106 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
108 Report and patch from Dmitry Banschikov.
110 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
111 Removed SSL_clear() after SSL_new() which led to protocol negotiation
112 failures. We appear to now support TLS1.1+ with Exim.
114 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
115 lets Exim select keys and certificates based upon TLS SNI from client.
116 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
117 before an outbound SMTP session. New log_selector, +tls_sni.
119 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
120 NULL dereference. Report and patch from Alun Jones.
122 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
123 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
124 Not seeing resolver debug output on NetBSD, but suspect this is a
125 resolver implementation change.
127 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
128 Left warnings. Added "eximon gdb" invocation mode.
130 PP/21 Defaulting "accept_8bitmime" to true, not false.
132 PP/22 Added -bw for inetd wait mode support.
134 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
135 locate the relevant includes and libraries. Made this the default.
137 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
138 Bugzilla 1246, report and most of solution from Tomasz Kusy.
140 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
141 This may cause build issues on older platforms.
143 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
144 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
145 gnutls_require_mac & gnutls_require_protocols (no longer supported).
146 Added SNI support via GnuTLS too.
147 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
149 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
151 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
152 Applied second patch from Janne, implementing suggestion to default
153 multiple-strings-in-record handling to match SPF spec.
155 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
157 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
158 read-only, out of scope).
159 Patch from Wolfgang Breyha, report from Stuart Northfield.
161 PP/29 Fix three issues highlighted by clang analyser static analysis.
162 Only crash-plausible issue would require the Cambridge-specific
163 iplookup router and a misconfiguration.
164 Report from Marcin Mirosław.
166 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
168 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
169 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
170 As part of this, removing so much warning spew let me fix some minor
171 real issues in debug logging.
173 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
174 assignment on my part. Fixed.
176 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
177 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
178 Janne Snabb (who went above and beyond: thank you).
180 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
181 string otherwise requires a connection and a bunch more work and it's
182 relatively easy to get wrong. Should also expose TLS library linkage
185 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
186 64-bit ${eval} (JH/03).
188 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
189 GNU libc to support some of the 64-bit stuff, should not lead to
190 conflicts. Defined before os.h is pulled in, so if a given platform
191 needs to override this, it can.
193 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
194 protection layer was required, which is not implemented.
195 Bugzilla 1254, patch from Wolfgang Breyha.
197 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
198 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
199 tls_dhparam take prime identifiers. Also unbreak combination of
200 OpenSSL+DH_params+TLSSNI.
202 PP/39 Disable SSLv2 by default in OpenSSL support.
208 PP/01 Solaris build fix for Oracle's LDAP libraries.
209 Bugzilla 1109, patch from Stephen Usher.
211 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
213 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
216 TF/02 Fix a couple more cases where we did not log the error message
217 when unlink() failed. See also change 4.74-TF/03.
219 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
220 lock up or crash if it happened to be inside a call to libc when it
221 got a SIGUSR1 from exiwhat.
223 The SIGUSR1 handler appends the current process status to the process
224 log which is later printed by exiwhat. It used to use the general
225 purpose logging code to do this, but several functions it calls are
226 not safe for signals.
228 The new output code in the SIGUSR1 handler is specific to the process
229 log, and simple enough that it's easy to inspect for signal safety.
230 Removing some special cases also simplifies the general logging code.
231 Removing the spurious timestamps from the process log simplifies
234 TF/04 Improved ratelimit ACL condition.
236 The /noupdate option has been deprecated in favour of /readonly which
237 has clearer semantics. The /leaky, /strict, and /readonly update modes
238 are mutually exclusive. The update mode is no longer included in the
239 database key; it just determines when the database is updated. (This
240 means that when you upgrde Exim will forget old rate measurements.)
242 Exim now checks that the per_* options are used with an update mode that
243 makes sense for the current ACL. For example, when Exim is processing a
244 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
245 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
246 must specify per_mail/readonly. If you omit the update mode it defaults to
247 /leaky where that makes sense (as before) or /readonly where required.
249 The /noupdate option is now undocumented but still supported for
250 backwards compatibility. It is equivalent to /readonly except that in
251 ACLs where /readonly is required you may specify /leaky/noupdate or
252 /strict/noupdate which are treated the same as /readonly.
254 A useful new feature is the /count= option. This is a generalization
255 of the per_byte option, so that you can measure the throughput of other
256 aggregate values. For example, the per_byte option is now equivalent
257 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
259 The per_rcpt option has been generalized using the /count= mechanism
260 (though it's more complicated than the per_byte equivalence). When it is
261 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
262 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
263 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
264 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
265 Note that using per_rcpt with a non-readonly update mode in more than
266 one ACL will cause the recipients to be double-counted. (The per_mail
267 and per_byte options don't have this problem.)
269 The handling of very low rates has changed slightly. If the computed rate
270 is less than the event's count (usually one) then this event is the first
271 after a long gap. In this case the rate is set to the same as this event's
272 count, so that the first message of a spam run is counted properly.
274 The major new feature is a mechanism for counting the rate of unique
275 events. The new per_addr option counts the number of different
276 recipients that someone has sent messages to in the last time period. It
277 behaves like per_rcpt if all the recipient addresses are different, but
278 duplicate recipient addresses do not increase the measured rate. Like
279 the /count= option this is a general mechanism, so the per_addr option
280 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
281 example, measure the rate that a client uses different sender addresses
282 with the options per_mail/unique=$sender_address. There are further
283 details in the main documentation.
285 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
287 TF/06 Removed a few PCRE remnants.
289 TF/07 Automatically extract Exim's version number from tags in the git
290 repository when doing development or release builds.
292 PP/02 Raise smtp_cmd_buffer_size to 16kB.
293 Bugzilla 879. Patch from Paul Fisher.
295 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
296 Heavily based on revision 40f9a89a from Simon Arlott's tree.
299 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
301 PP/05 Variable $av_failed, true if the AV scanner deferred.
302 Bugzilla 1078. Patch from John Horne.
304 PP/06 Stop make process more reliably on build failure.
305 Bugzilla 1087. Patch from Heiko Schlittermann.
307 PP/07 Make maildir_use_size_file an _expandable_ boolean.
308 Bugzilla 1089. Patch from Heiko Schlittermann.
310 PP/08 Handle ${run} returning more data than OS pipe buffer size.
311 Bugzilla 1131. Patch from Holger Weiß.
313 PP/09 Handle IPv6 addresses with SPF.
314 Bugzilla 860. Patch from Wolfgang Breyha.
316 PP/10 GnuTLS: support TLS 1.2 & 1.1.
318 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
321 PP/11 match_* no longer expand right-hand-side by default.
322 New compile-time build option, EXPAND_LISTMATCH_RHS.
323 New expansion conditions, "inlist", "inlisti".
325 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
327 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
329 PP/14 fix log_write() format string regression from TF/03.
330 Bugzilla 1152. Patch from Dmitry Isaikin.
336 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
338 PP/02 Harmonised TLS library version reporting; only show if debugging.
339 Layout now matches that introduced for other libraries in 4.74 PP/03.
341 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
343 PP/04 New "dns_use_edns0" global option.
345 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
348 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
349 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
351 TK/01 Updated PolarSSL code to 0.14.2.
352 Bugzilla 1097. Patch from Andreas Metzler.
354 PP/07 Catch divide-by-zero in ${eval:...}.
357 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
360 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
361 format-string attack -- SECURITY: remote arbitrary code execution.
363 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
364 time unintentionally subject to list matching rules, letting the header
365 cause arbitrary Exim lookups (of items which can occur in lists, *not*
366 arbitrary string expansion). This allowed for information disclosure.
368 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
369 INT_MIN/-1 -- value coerced to INT_MAX.
375 NM/01 Workround for PCRE version dependency in version reporting
378 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
379 This fixes portability to compilers other than gcc, notably
380 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
382 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
383 makefiles for portability to HP-UX and POSIX correctness.
385 PP/01 Permit LOOKUP_foo enabling on the make command-line.
386 Also via indented variable definition in the Makefile.
387 (Debugging by Oliver Heesakkers).
389 PP/02 Restore caching of spamd results with expanded spamd_address.
390 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
392 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
393 Improves build reliability. Fix from: Frank Elsner
395 NM/02 Fix wide character breakage in the rfc2047 coding
396 Fixes bug 1064. Patch from Andrey N. Oktyabrski
398 NM/03 Allow underscore in dnslist lookups
399 Fixes bug 1026. Patch from Graeme Fowler
401 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
402 Code patches from Adam Ciarcinski of NetBSD.
404 NM/04 Fixed exiqgrep to cope with mailq missing size issue
407 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
408 is logged, to avoid truncation. Patch from John Horne.
410 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
411 Patch from Jakob Hirsch.
413 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
414 SQL string expansion failure details.
415 Patch from Andrey Oktyabrski.
417 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
418 Patch from Simon Arlott.
420 PP/09 New lookups functionality failed to compile on old gcc which rejects
421 extern declarations in function scope.
422 Patch from Oliver Fleischmann
424 PP/10 Use sig_atomic_t for flags set from signal handlers.
425 Check getgroups() return and improve debugging.
426 Fixed developed for diagnosis in bug 927 (which turned out to be
429 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
430 Patch from Mark Zealey.
432 PP/12 Bugzilla 1056: Improved spamd server selection.
433 Patch from Mark Zealey.
435 PP/13 Bugzilla 1086: Deal with maildir quota file races.
436 Based on patch from Heiko Schlittermann.
438 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
439 Patch from Uwe Doering, sign-off by Michael Haardt.
441 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
442 variable declaration deep within a block. Bug and patch from
445 PP/15 lookups-Makefile IRIX compatibilty coercion.
447 PP/16 Make DISABLE_DKIM build knob functional.
449 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
450 Patch by Simon Arlott
452 TF/03 Fix valgrind.h portability to C89 compilers that do not support
453 variable argument macros. Our copy now differs from upstream.
459 TF/01 Failure to get a lock on a hints database can have serious
460 consequences so log it to the panic log.
462 TF/02 Log LMTP confirmation messages in the same way as SMTP,
463 controlled using the smtp_confirmation log selector.
465 TF/03 Include the error message when we fail to unlink a spool file.
467 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
468 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
469 for maintaining out-of-tree patches for some time.
471 PP/01 Bugzilla 139: Documentation and portability issues.
472 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
473 Handle per-OS dynamic-module compilation flags.
475 PP/02 Let /dev/null have normal permissions.
476 The 4.73 fixes were a little too stringent and complained about the
477 permissions on /dev/null. Exempt it from some checks.
478 Reported by Andreas M. Kirchwitz.
480 PP/03 Report version information for many libraries, including
481 Exim version information for dynamically loaded libraries. Created
482 version.h, now support a version extension string for distributors
483 who patch heavily. Dynamic module ABI change.
485 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
486 privilege escalation vulnerability whereby the Exim run-time user
487 can cause root to append content of the attacker's choosing to
490 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
493 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
494 If dropping privileges for untrusted macros, we disabled normal logging
495 on the basis that it would fail; for the Exim run-time user, this is not
496 the case, and it resulted in successful deliveries going unlogged.
497 Fixed. Reported by Andreas Metzler.
503 PP/01 Date: & Message-Id: revert to normally being appended to a message,
504 only prepend for the Resent-* case. Fixes regression introduced in
505 Exim 4.70 by NM/22 for Bugzilla 607.
507 PP/02 Include check_rfc2047_length in configure.default because we're seeing
508 increasing numbers of administrators be bitten by this.
510 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
512 PP/03 Bugzilla 994: added openssl_options main configuration option.
514 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
516 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
518 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
520 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
521 without a peer certificate, leading to a segfault because of an
522 assumption that peers always have certificates. Be a little more
523 paranoid. Problem reported by Martin Tscholak.
525 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
526 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
527 NB: ClamAV planning to remove STREAM in "middle of 2010".
528 CL also introduces -bmalware, various -d+acl logging additions and
529 more caution in buffer sizes.
531 PP/09 Implemented reverse_ip expansion operator.
533 PP/10 Bugzilla 937: provide a "debug" ACL control.
535 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
537 PP/12 Bugzilla 973: Implement --version.
539 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
541 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
543 PP/15 Bugzilla 816: support multiple condition rules on Routers.
545 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
546 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
547 ignore trailing whitespace.
549 JJ/02 prevent non-panic DKIM error from being sent to paniclog
551 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
554 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
555 Notification from Dr Andrew Aitchison.
557 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
558 ExtendedDetectionInfo response format.
559 Notification from John Horne.
561 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
564 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
565 XSL and documented dependency on system catalogs, with examples of how
568 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
571 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
572 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
573 configuration file which is writeable by the Exim user or group.
575 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
576 of configuration files to cover files specified with the -C option if
577 they are going to be used with root privileges, not just the default
580 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
581 option (effectively making it always true).
583 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
584 files to be used while preserving root privileges.
586 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
587 that rogue child processes cannot use them.
589 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
590 run-time user, instead of root.
592 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
593 Exim run-time user without dropping privileges.
595 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
596 result string, instead of calling string_vformat() twice with the same
599 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
600 for other users. Others should always drop root privileges if they use
601 -C on the command line, even for a whitelisted configure file.
603 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
605 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
611 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
612 $data_path, and $header_path variables; fixed documentation bugs and
615 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
616 exipick to access non-standard spools, including the "frozen" queue
619 NM/01 Bugzilla 965: Support mysql stored procedures.
620 Patch from Alain Williams
622 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
624 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
625 Patch from Andreas Metzler
627 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
628 Patch from Kirill Miazine
630 NM/05 Bugzilla 671: Added umask to procmail example.
632 JJ/03 installed exipick 20100323.0, fixing doc bug
634 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
635 directory. Notification and patch from Dan Rosenberg.
637 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
639 TK/02 Improve log output when DKIM signing operation fails.
641 MH/01 Treat the transport option dkim_domain as a colon separated
642 list, not as a single string, and sign the message with each element,
643 omitting multiple occurences of the same signer.
645 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
646 Bugzilla 985, 986. Patch by Simon Arlott
648 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
649 Patch by Simon Arlott
651 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
652 MBX locking. Notification from Dan Rosenberg.
658 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
660 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
662 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
664 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
666 NM/04 Bugzilla 915: Flag broken perl installation during build.
672 TK/01 Added patch by Johannes Berg that expands the main option
673 "spamd_address" if it starts with a dollar sign.
675 TK/02 Write list of recipients to X-Envelope-Sender header when building
676 the mbox-format spool file for content scanning (suggested by Jakob
679 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
680 (http://www.dcc-servers.net/) support via dccifd. Activated by
681 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
683 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
684 by Mark Daniel Reidel <mr@df.eu>.
686 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
687 When building exim an external PCRE library is now needed -
688 PCRE is a system library on the majority of modern systems.
689 See entry on PCRE_LIBS in EDITME file.
691 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
692 conversation. Added nologin parameter to request.
693 Patch contributed by Kirill Miazine.
695 TF/01 Do not log submission mode rewrites if they do not change the address.
697 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
699 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
700 log files in place. Contributed by Roberto Lima.
702 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
704 TF/03 Bugzilla 615: When checking the local_parts router precondition
705 after a local_part_suffix or local_part_prefix option, Exim now
706 does not use the address's named list lookup cache, since this
707 contains cached lookups for the whole local part.
709 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
710 Robert Millan. Documentation is in experimental-spec.txt.
712 TF/04 Bugzilla 668: Fix parallel build (make -j).
714 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
716 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
717 Patch provided by Jan Srzednicki.
719 TF/05 Leading white space used to be stripped from $spam_report which
720 wrecked the formatting. Now it is preserved.
722 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
723 that they are available at delivery time.
725 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
727 TF/08 TLS error reporting now respects the incoming_interface and
728 incoming_port log selectors.
730 TF/09 Produce a more useful error message if an SMTP transport's hosts
731 setting expands to an empty string.
733 NM/06 Bugzilla 744: EXPN did not work under TLS.
734 Patch provided by Phil Pennock.
736 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
737 Patch provided by Richard Godbee.
739 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
740 acl_smtp_notquit, added index entry.
742 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
743 Patch provided by Eugene Bujak.
745 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
746 accept(). Patch provided by Maxim Dounin.
748 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
749 Patch provided by Phil Pennock.
751 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
753 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
754 Patch provided by Brad "anomie" Jorsch.
756 NM/14 Bugzilla 622: Added timeout setting to transport filter.
757 Patch provided by Dean Brooks.
759 TK/05 Add native DKIM support (does not depend on external libraries).
761 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
762 Patch provided by Graeme Fowler.
764 NM/16 Bugzilla 851: Documentation example syntax fix.
766 NM/17 Changed NOTICE file to remove references to embedded PCRE.
768 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
771 NM/19 Bugzilla 745: TLS version reporting.
772 Patch provided by Phil Pennock.
774 NM/20 Bugzilla 167: bool: condition support.
775 Patch provided by Phil Pennock.
777 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
778 clients. Patch provided by Phil Pennock.
780 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
781 Patch provided by Brad "anomie" Jorsch.
783 NM/23 Bugzilla 687: Fix misparses in eximstats.
784 Patch provided by Heiko Schlittermann.
786 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
787 Patch provided by Heiko Schlittermann.
789 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
790 plus update to original patch.
792 NM/26 Bugzilla 799: Documentation correction for ratelimit.
794 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
795 Patch provided by David Brownlee.
797 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
799 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
801 NM/30 Bugzilla 888: TLS documentation bugfixes.
803 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
805 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
806 Unlike the original bugzilla I have changed all shell scripts in src tree.
808 NM/33 Bugzilla 898: Transport filter timeout fix.
809 Patch by Todd Rinaldo.
811 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
812 Patch by Serge Demonchaux.
814 NM/35 Bugzilla 39: Base64 decode bug fixes.
815 Patch by Jakob Hirsch.
817 NM/36 Bugzilla 909: Correct connect() call in dcc code.
819 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
821 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
823 NM/39 Bugzilla 911: Fixed MakeLinks build script.
829 TK/01 Add preliminary DKIM support. Currently requires a forked version of
830 ALT-N's libdkim that I have put here:
831 http://duncanthrax.net/exim-experimental/
833 Note to Michael Haardt: I had to rename some vars in sieve.c. They
834 were called 'true' and it seems that C99 defines that as a reserved
835 keyword to be used with 'bool' variable types. That means you could
836 not include C99-style headers which use bools without triggering
837 build errors in sieve.c.
839 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
840 as mailq or other aliases. Changed the --help handling significantly
841 to do whats expected. exim_usage() emits usage/help information.
843 SC/01 Added the -bylocaldomain option to eximstats.
845 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
847 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
849 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
855 PH/01 Another patch from the Sieve maintainer.
857 PH/02 When an IPv6 address is converted to a string for single-key lookup
858 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
859 dots are used instead of colons so that keys in lsearch files need not
860 contain colons. This was done some time before quoting was made available
861 in lsearch files. However, iplsearch files do require colons in IPv6 keys
862 (notated using the quote facility) so as to distinguish them from IPv4
863 keys. This meant that lookups for IP addresses in host lists did not work
864 for iplsearch lookups.
866 This has been fixed by arranging for IPv6 addresses to be expressed with
867 colons if the lookup type is iplsearch. This is not incompatible, because
868 previously such lookups could never work.
870 The situation is now rather anomolous, since one *can* have colons in
871 ordinary lsearch keys. However, making the change in all cases is
872 incompatible and would probably break a number of configurations.
874 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
877 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
878 conversion specification without a maximum field width, thereby enabling
879 a rogue spamd server to cause a buffer overflow. While nobody in their
880 right mind would setup Exim to query an untrusted spamd server, an
881 attacker that gains access to a server running spamd could potentially
882 exploit this vulnerability to run arbitrary code as the Exim user.
884 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
885 $primary_hostname instead of what libspf2 thinks the hosts name is.
887 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
888 a directory entry by the name of the lookup key. Previously, if a
889 symlink pointed to a non-existing file or a file in a directory that
890 Exim lacked permissions to read, a lookup for a key matching that
891 symlink would fail. Now it is enough that a matching directory entry
892 exists, symlink or not. (Bugzilla 503.)
894 PH/03 The body_linecount and body_zerocount variables are now exported in the
897 PH/04 Added the $dnslist_matched variable.
899 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
900 This means they are set thereafter only if the connection becomes
903 PH/06 Added the client_condition to authenticators so that some can be skipped
904 by clients under certain conditions.
906 PH/07 The error message for a badly-placed control=no_multiline_responses left
907 "_responses" off the end of the name.
909 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
911 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
912 (without spaces) instead of just copying the configuration text.
914 PH/10 Added the /noupdate option to the ratelimit ACL condition.
916 PH/11 Added $max_received_linelength.
918 PH/12 Added +ignore_defer and +include_defer to host lists.
920 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
921 way in which PCRE > 7.0 is built.
923 PH/14 Implemented queue_only_load_latch.
925 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
926 MAIL command. The effect was to mangle the value on 64-bit systems.
928 PH/16 Another patch from the Sieve maintainer.
930 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
932 PH/18 If a system quota error occurred while trying to create the file for
933 a maildir delivery, the message "Mailbox is full" was not appended to the
934 bounce if the delivery eventually timed out. Change 4.67/27 below applied
935 only to a quota excession during the actual writing of the file.
937 PH/19 It seems that peer DN values may contain newlines (and other non-printing
938 characters?) which causes problems in log lines. The DN values are now
939 passed through string_printing() before being added to log lines.
941 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
942 and InterBase are left for another time.)
944 PH/21 Added message_body_newlines option.
946 PH/22 Guard against possible overflow in moan_check_errorcopy().
948 PH/23 POSIX allows open() to be a macro; guard against that.
950 PH/24 If the recipient of an error message contained an @ in the local part
951 (suitably quoted, of course), incorrect values were put in $domain and
952 $local_part during the evaluation of errors_copy.
958 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
959 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
962 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
963 issue a MAIL command.
965 PH/02 In an ACL statement such as
967 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
969 if a client was not listed at all, or was listed with a value other than
970 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
971 the condition was not true (as it should be), so access was not denied.
972 The bug was that the ! inversion was incorrectly passed on to the second
973 item. This has been fixed.
975 PH/03 Added additional dnslists conditions == and =& which are different from
976 = and & when the dns lookup returns more than one IP address.
978 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
979 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
981 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
982 FSYNC, which compiles an option called disable_fsync that allows for
983 bypassing fsync(). The documentation is heavily laced with warnings.
985 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
987 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
988 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
989 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
990 including adding "make clean"; (3) Added -fPIC when compiling the test
991 dynamically loaded module, to get rid of a warning.
993 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
994 message fails, move_frozen_messages = true and ignore_bounce_errors_after
995 = 0s. The bug is otherwise harmless.
997 PH/07 There was a bug in the dovecot authenticator such that the value of
998 $auth1 could be overwritten, and so not correctly preserved, after a
999 successful authentication. This usually meant that the value preserved by
1000 the server_setid option was incorrect.
1002 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1004 PH/09 Installed PCRE release 7.0.
1006 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1007 run for batched SMTP input. It is now run at the start of every message
1008 in the batch. While fixing this I discovered that the process information
1009 (output by running exiwhat) was not always getting set for -bs and -bS
1010 input. This is fixed, and it now also says "batched" for BSMTP.
1012 PH/11 Added control=no_pipelining.
1014 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1015 patch, slightly modified), and move the expansion of helo_data till after
1016 the connection is made in the smtp transport (so it can use these
1019 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1021 PH/14 Added log_selector = +pid.
1023 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1025 PH/16 Add ${if forany and ${if forall.
1027 PH/17 Added dsn_from option to vary the From: line in DSNs.
1029 PH/18 Flush SMTP output before performing a callout, unless control =
1030 no_callout_flush is set.
1032 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1033 was true (the default) a successful delivery failed to delete the retry
1034 item, thus causing premature timeout of the address. The bug is now
1037 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1039 PH/21 Long custom messages for fakedefer and fakereject are now split up
1040 into multiline reponses in the same way that messages for "deny" and
1041 other ACL rejections are.
1043 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1044 with slight modification.
1046 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1047 draft, changing the syntax and factoring some duplicate code".
1049 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1050 for deliveries of the second and subsequent messages over the same SMTP
1053 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1054 ${reduce, with only minor "tidies".
1056 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1058 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1059 expansion side effects.
1061 PH/27 When a message times out after an over-quota error from an Exim-imposed
1062 quota, the bounce message says "mailbox is full". This message was not
1063 being given when it was a system quota that was exceeded. It now should
1066 MH/03 Made $recipients available in local_scan(). local_scan() already has
1067 better access to the recipient list through recipients_list[], but
1068 $recipients can be useful in postmaster-provided expansion strings.
1070 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1071 in the case of a MAIL command with additional options following the
1072 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1073 were accidentally chopped off.
1075 PH/29 SMTP synchronization checks are implemented when a command is read -
1076 there is a check that no more input is waiting when there shouldn't be
1077 any. However, for some commands, a delay in an ACL can mean that it is
1078 some time before the response is written. In this time, more input might
1079 arrive, invalidly. So now there are extra checks after an ACL has run for
1080 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1081 pipelining has not been advertised.
1083 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1085 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1086 returned in the SMTP response when smtp_return_error_details was false.
1087 This has been fixed.
1089 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1090 instead of the C I/O that was originally supplied, because problems were
1091 reported on Solaris.
1093 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1094 Exim which did not show up earlier: it was assuming that a call to
1095 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1096 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1097 was a macro that became an assignment, so it seemed to work. This has
1098 changed to a proper function call with a void return, hence the compile
1099 error. Exim's code has been fixed.
1101 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1104 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1106 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1108 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1109 because it turns out to be too big or there is a timeout) there is no
1110 "Completed" line in the log. When some messages of this type were
1111 selected by exigrep, they were listed as "not completed". Others were
1112 picked up by some special patterns. I have improved the selection
1113 criteria to be more general.
1115 PH/38 The host_find_failed option in the manualroute router can now be set
1116 to "ignore", to completely ignore a host whose IP address cannot be
1117 found. If all hosts are ignored, the behaviour is controlled by the new
1118 host_all_ignored option.
1120 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1121 homing or because of multiple MX records with /mx) generated more than
1122 one IP address, and the following item turned out to be the local host,
1123 all the secondary addresses of the first item were incorrectly removed
1124 from the list, along with the local host and any following hosts (which
1125 is what is supposed to happen).
1127 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1128 whoever called Exim into the -H file. In the case of the daemon it was
1129 behaving confusingly. When first started, it used values for whoever
1130 started the daemon, but after a SIGHUP it used the Exim user (because it
1131 calls itself on a restart). I have changed the code so that it now always
1134 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1135 message are rejected with the same error (e.g. no authentication or bad
1136 sender address), and a DATA command is nevertheless sent (as can happen
1137 with PIPELINING or a stupid MUA), the error message that was given to the
1138 RCPT commands is included in the rejection of the DATA command. This is
1139 intended to be helpful for MUAs that show only the final error to their
1142 PH/42 Another patch from the Sieve maintainer.
1144 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1145 Eximstats - Fixed some broken HTML links and added missing column headers
1147 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1148 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1150 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1152 PH/43 Yet another patch from the Sieve maintainer.
1154 PH/44 I found a way to check for a TCP/IP connection going away before sending
1155 the response to the final '.' that terminates a message, but only in the
1156 case where the client has not sent further data following the '.'
1157 (unfortunately, this is allowed). However, in many cases there won't be
1158 any further data because there won't be any more messages to send. A call
1159 to select() can be used: if it shows that the input is "ready", there is
1160 either input waiting, or the socket has been closed. An attempt to read
1161 the next input character can distinguish the two cases. Previously, Exim
1162 would have sent an OK response which the client would never have see.
1163 This could lead to message repetition. This fix should cure that, at
1164 least in a lot of common cases.
1166 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1167 advertised in response to EHLO.
1173 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1174 fixed by 4.65/MH/01 (is this a record?) are fixed:
1176 (i) An empty string was always treated as zero by the numeric comparison
1177 operators. This behaviour has been restored.
1179 (ii) It is documented that the numeric comparison operators always treat
1180 their arguments as decimal numbers. This was broken in that numbers
1181 starting with 0 were being interpreted as octal.
1183 While fixing these problems I realized that there was another issue that
1184 hadn't been noticed. Values of message_size_limit (both the global option
1185 and the transport option) were treated as octal if they started with 0.
1186 The documentation was vague. These values are now always treated as
1187 decimal, and I will make that clear in the documentation.
1193 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1194 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1197 MH/01 Don't check that the operands of numeric comparison operators are
1198 integers when their expansion is in "skipping" mode (fixes bug
1199 introduced by 4.64-PH/07).
1201 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1202 child addresses, Exim now panics and dies. Previously, because the count
1203 is held in a short int, deliveries were likely to be lost. As such a
1204 large number of recipients for a single message is ridiculous
1205 (performance will be very, very poor), I have chosen to impose a limit
1206 rather than extend the field.
1212 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1213 leftover -K file (the existence of which was triggered by #402).
1214 While we were at it, introduced process PID as part of the -K
1215 filename. This should rule out race conditions when creating
1218 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1219 processing considerably. Previous code took too long for large mails,
1220 triggering a timeout which in turn triggers #401.
1222 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1223 in the DK code in transports.c. sendfile() is not really portable,
1224 hence the _LINUX specificness.
1226 TF/01 In the add_headers option to the mail command in an Exim filter,
1227 there was a bug that Exim would claim a syntax error in any
1228 header after the first one which had an odd number of characters
1231 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1232 callout verification, Exim cached a "reject" for the entire domain. This
1233 is correct for most verifications, but it is not correct for a recipient
1234 verification with use_sender or use_postmaster set, because in that case
1235 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1236 case of MAIL FROM:<> rejection from other early rejections (e.g.
1237 rejection of HELO). When verifying a recipient using a non-null MAIL
1238 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1239 Whatever the result of the callout, the value of the domain cache is
1240 left unchanged (for any other kind of callout, getting as far as trying
1241 RCPT means that the domain itself is ok).
1243 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1246 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1247 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1248 session, the daemon ignores ECONNECT errors and logs others; it now
1249 ignores EPIPE as well.
1251 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1252 (quoted-printable decoding).
1254 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1255 later the small subsequent patch to fix an introduced bug.
1257 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1259 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1261 PH/08 An error is now given if message_size_limit is specified negative.
1263 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1264 to be given (somewhat) arbitrary names.
1266 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1269 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1270 miscellaneous code fixes
1272 PH/10 Added the log_reject_target ACL modifier to specify where to log
1275 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1276 hostname. This is wrong, because it relates to the incoming message (and
1277 probably the interface on which it is arriving) and not to the outgoing
1278 callout (which could be using a different interface). This has been
1279 changed to use the value of the helo_data option from the smtp transport
1280 instead - this is what is used when a message is actually being sent. If
1281 there is no remote transport (possible with a router that sets up host
1282 addresses), $smtp_active_hostname is used.
1284 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1285 tweaks were necessary in order to get it to work (see also 21 below):
1286 (a) The code assumed that strncpy() returns a negative number on buffer
1287 overflow, which isn't the case. Replaced with Exim's string_format()
1289 (b) There were several signed/unsigned issues. I just did the minimum
1290 hacking in of casts. There is scope for a larger refactoring.
1291 (c) The code used strcasecmp() which is not a standard C function.
1292 Replaced with Exim's strcmpic() function.
1293 (d) The code set only $1; it now sets $auth1 as well.
1294 (e) A simple test gave the error "authentication client didn't specify
1295 service in request". It would seem that Dovecot has changed its
1296 interface. Fortunately there's a specification; I followed it and
1297 changed what the client sends and it appears to be working now.
1299 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1302 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1303 address A is aliased to B and C, where B exists and C does not. Without
1304 -v the output is "A verified" because verification stops after a
1305 successful redirection if more than one address is generated. However,
1306 with -v the child addresses are also verified. Exim was outputting "A
1307 failed to verify" and then showing the successful verification for C,
1308 with its parentage. It now outputs "B failed to verify", showing B's
1309 parentage before showing the successful verification of C.
1311 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1312 look up a TXT record in a specific list after matching in a combined
1315 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1316 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1317 they consult the DNS. I had assumed they would set it the way they
1318 wanted; and indeed my experiments on Linux seem to show that in some
1319 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1320 To be on the safe side, however, I have now made the interface to
1321 host_find_byname() similar to host_find_bydns(), with an argument
1322 containing the DNS resolver options. The host_find_byname() function now
1323 sets these options at its start, just as host_find_bydns() does. The smtp
1324 transport options dns_qualify_single and dns_search_parents are passed to
1325 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1326 of host_find_byname() use the default settings of RES_DEFNAMES
1327 (qualify_single) but not RES_DNSRCH (search_parents).
1329 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1330 spool_read_header() do less string testing, by means of a preliminary
1331 switch on the second character of optional "-foo" lines. (This is
1332 overdue, caused by the large number of possibilities that now exist.
1333 Originally there were few.) While I was there, I also converted the
1334 str(n)cmp tests so they don't re-test the leading "-" and the first
1335 character, in the hope this might squeeze out yet more improvement.
1337 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1338 flag allowing group syntax was set by the header_syntax check but not
1339 turned off, possible causing trouble later; (2) The flag was not being
1340 set at all for the header_verify test, causing "group"-style headers to
1341 be rejected. I have now set it in this case, and also caused header_
1342 verify to ignore an empty address taken from a group. While doing this, I
1343 came across some other cases where the code for allowing group syntax
1344 while scanning a header line wasn't quite right (mostly, not resetting
1345 the flag correctly in the right place). These bugs could have caused
1346 trouble for malformed header lines. I hope it is now all correct.
1348 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1349 with the "reply" argument non-NULL. The code, however (which originally
1350 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1351 but it didn't always do it. This confused somebody who was copying the
1352 code for some other use. I have removed all the tests.
1354 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1355 feature that was used to support insecure browsers during the U.S. crypto
1356 embargo. It requires special client support, and Exim is probably the
1357 only MTA that supported it -- and would never use it because real RSA is
1358 always available. This code has been removed, because it had the bad
1359 effect of slowing Exim down by computing (never used) parameters for the
1360 RSA_EXPORT functionality.
1362 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1363 authenticator to fail if there's a tab character in the incoming data
1364 (there should never be unless someone is messing about, as it's supposed
1365 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1366 if the connection is using TLS or if the remote IP is the same as the
1367 local IP, and the "valid-client-cert option" if a client certificate has
1370 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1371 authenticators. This can be used for authorization after authentication
1372 succeeds. (In the case of plaintext, it servers for both authentication
1375 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1376 if any retry times were supplied.
1378 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1379 connection, where there is no client IP address to check. In this
1380 situation, the verify now always succeeds.
1382 PH/25 Applied John Jetmore's -Mset patch.
1384 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1386 PH/27 In a string expansion for a processed (not raw) header when multiple
1387 headers of the same name were present, leading whitespace was being
1388 removed from all of them, but trailing whitespace was being removed only
1389 from the last one. Now trailing whitespace is removed from each header
1390 before concatenation. Completely empty headers in a concatenation (as
1391 before) are ignored.
1393 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1394 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1396 PH/29 [Removed. This was a change that I later backed out, and forgot to
1397 correct the ChangeLog entry (that I had efficiently created) before
1398 committing the later change.]
1400 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1401 address errors (4xx response to RCPT) over the same connection as other
1402 messages routed to the same hosts. Such deliveries are always "forced",
1403 so retry times are not inspected. This resulted in far too many retries
1404 for the affected addresses. The effect occurred only when there were more
1405 hosts than the hosts_max_try setting in the smtp transport when it had
1406 the 4xx errors. Those hosts that it had tried were not added to the list
1407 of hosts for which the message was waiting, so if all were tried, there
1408 was no problem. Two fixes have been applied:
1410 (i) If there are any address or message errors in an SMTP delivery, none
1411 of the hosts (tried or untried) are now added to the list of hosts
1412 for which the message is waiting, so the message should not be a
1413 candidate for sending over the same connection that was used for a
1414 successful delivery of some other message. This seems entirely
1415 reasonable: after all the message is NOT "waiting for some host".
1416 This is so "obvious" that I'm not sure why it wasn't done
1417 previously. Hope I haven't missed anything, but it can't do any
1418 harm, as the worst effect is to miss an optimization.
1420 (ii) If, despite (i), such a delivery is accidentally attempted, the
1421 routing retry time is respected, so at least it doesn't keep
1422 hammering the server.
1424 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1425 in ${readsocket because some servers need this prod.
1427 PH/32 Added some extra debug output when updating a wait-xxx database.
1429 PH/33 The hint "could be header name not terminated by colon", which has been
1430 given for certain expansion errors for a long time, was not being given
1431 for the ${if def:h_colon_omitted{... case.
1433 PH/34 The spec says: "With one important exception, whenever a domain list is
1434 being scanned, $domain contains the subject domain." There was at least
1435 one case where this was not true.
1437 PH/35 The error "getsockname() failed: connection reset by peer" was being
1438 written to the panic log as well as the main log, but it isn't really
1439 panic-worthy as it just means the connection died rather early on. I have
1440 removed the panic log writing for the ECONNRESET error when getsockname()
1443 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1444 runs only) independently of the message's sender address. This meant
1445 that, if the 4xx error was in fact related to the sender, a different
1446 message to the same recipient with a different sender could confuse
1447 things. In particualar, this can happen when sending to a greylisting
1448 server, but other circumstances could also provoke similar problems.
1449 I have changed the default so that the retry time for these errors is now
1450 based a combination of the sender and recipient addresses. This change
1451 can be overridden by setting address_retry_include_sender=false in the
1454 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1455 remote server are returned as part of bounce messages. This was not
1456 happening for LMTP over a pipe (the lmtp transport), but now it is the
1457 same for both kinds of LMTP.
1459 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1460 in header lines that were in fact CNAMEs. This is no longer the case.
1462 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1463 and queue runs started by the daemon processed all messages. This has
1464 been fixed so that -R and -S can now usefully be given with -q<time>.
1466 PH/40 Import PCRE release 6.7 (fixes some bugs).
1468 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1470 PH/42 Give an error if -q is specified more than once.
1472 PH/43 Renamed the variables $interface_address and $interface_port as
1473 $received_ip_address and $received_port, to make it clear that these
1474 values apply to message reception, and not to the outgoing interface when
1475 a message is delivered. (The old names remain recognized, of course.)
1477 PH/44 There was no timeout on the connect() call when using a Unix domain
1478 socket in the ${readsocket expansion. There now is.
1480 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1481 be meaningful with "accept".
1483 SC/01 Eximstats V1.43
1484 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1486 SC/02 Eximstats V1.44
1487 Use a glob alias rather than an array ref in the generated
1488 parser. This improves both readability and performance.
1490 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1491 Collect SpamAssassin and rejection statistics.
1492 Don't display local sender or destination tables unless
1493 there is data to show.
1494 Added average volumes into the top table text output.
1496 SC/04 Eximstats V1.46
1497 Collect data on the number of addresses (recipients)
1498 as well as the number of messages.
1500 SC/05 Eximstats V1.47
1501 Added 'Message too big' to the list of mail rejection
1502 reasons (thanks to Marco Gaiarin).
1504 SC/06 Eximstats V1.48
1505 Mainlog lines which have GMT offsets and are too short to
1506 have a flag are now skipped.
1508 SC/07 Eximstats V1.49 (Alain Williams)
1509 Added the -emptyok flag.
1511 SC/08 Eximstats V1.50
1512 Fixes for obtaining the IP address from reject messages.
1514 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1515 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1516 whitesspace changes from 4.64-PH/27
1518 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1521 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1522 are found, allow negative numbers in numeric criteria)
1524 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1526 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1527 to match changes made in 4.64-PH/43
1529 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1531 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1532 "domain" is actually an IP address, and force a failure. This locks out
1533 those revolvers/nameservers that support "A-for-A" lookups, in
1534 contravention of the specifications.
1536 PH/48 When a host name was looked up from an IP address, and the subsequent
1537 forward lookup of the name timed out, the host name was left in
1538 $sender_host_name, contrary to the specification.
1540 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1541 restricted to single-key lookups, Exim was not diagnosing an error if
1542 * or *@ was used with a query-style lookup.
1544 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1546 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1547 long ago, but noone interested enough thought of it. Let's just say that
1548 the "1.1" means that there are some new functions that weren't there at
1549 some point in the past.
1551 PH/51 Error processing for expansion failure of helo_data from an smtp
1552 transport during callout processing was broken.
1554 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1555 tested/used via the -bh/-bhc/-bs options.
1557 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1558 bug, fixed in subsequent PCRE releases).
1560 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1561 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1563 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1569 SC/01 Use a glob alias rather than an array ref in eximstats generated
1570 parser. This improves both readability and performance.
1572 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1573 Don't display local sender or destination tables in eximstats unless
1574 there is data to show.
1575 Added average volumes into the eximstats top table text output.
1577 SC/03 Collect data on the number of addresses (recipients) as well
1578 as the number of messages in eximstats.
1580 TF/01 Correct an error in the documentation for the redirect router. Exim
1581 does (usually) call initgroups() when daemonizing.
1583 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1584 with consistent privilege compared to when running as a daemon.
1586 TF/03 Note in the spec that $authenticated_id is not set for local
1587 submissions from trusted users.
1589 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1590 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1592 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1593 by adding some example configuration directives to the default
1594 configuration file. A little bit of work is required to uncomment the
1595 directives and define how usernames and passwords are checked, but
1596 there is now a framework to start from.
1598 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1599 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1600 without this. I don't know how relevant this is to other LDAP libraries.
1602 PH/02 Add the verb name to the "unknown ACL verb" error.
1604 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1606 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1608 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1609 directory not expanded when it should be if an expanded home directory
1610 was set for the address (which is overridden by the transport).
1612 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1615 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1616 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1617 because it is too late at that time, and has no effect.
1619 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1620 security issue with \' (bugzilla #107). I could not use the
1621 PQescapeStringConn() function, because it needs a PGconn value as one of
1624 PH/08 When testing addresses using -bt, indicate those final addresses that
1625 are duplicates that would not cause an additional delivery. At least one
1626 person was confused, thinking that -bt output corresponded to deliveries.
1627 (Suppressing duplicates isn't a good idea as you lose the information
1628 about possibly different redirections that led to the duplicates.)
1630 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1631 systems where poll() doesn't work, in particular OS X.
1633 PH/10 Added more information to debugging output for retry time not reached.
1635 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1636 operations in malware.c.
1638 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1641 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1642 syslog_duplication set false, log lines that would normally be written
1643 both the the main log and to the reject log were not written to syslog at
1646 PH/14 In the default configuration, change the use of "message" in ACL warn
1647 statements to "add_header".
1649 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1650 not followed by a command (e.g. "seen endif").
1652 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1653 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1656 PH/17 Added extra conditions to the default value of delay_warning_condition
1660 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1661 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1662 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1665 The Auto-Submitted: and various List- headers are standardised, whereas I
1666 don't think Precedence: ever was.
1668 PH/18 Refactored debugging code in route_finduser() to show more information,
1669 in particular, the error code if getpwnam() issues one.
1671 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1672 This is apparently needed in addition to the PH/07 change above to avoid
1673 any possible encoding problems.
1675 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1676 but not after initializing Perl.
1678 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1679 output them only if debugging. By default they are written stderr,
1680 apparently, which is not desirable.
1682 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1685 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1688 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1690 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1691 authenticated or an ident call has been made. Suppress the default
1692 values for $authenticated_id and $authenticated_sender (but permit -oMai
1693 and -oMas) when testing with -bh.
1695 PH/24 Re-jigged the order of the tests in the default configuration so that the
1696 tests for valid domains and recipients precede the DNS black list and CSA
1697 tests, on the grounds that those ones are more expensive.
1699 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1700 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1701 This bug exists in every version of Exim that I still have, right back to
1704 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1705 However, an attempt to turn on case-sensitivity in a regex key by
1706 including (?-i) didn't work because the subject string was already
1707 lowercased, and the effects were non-intuitive. It turns out that a
1708 one-line patch can be used to allow (?-i) to work as expected.
1714 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1715 other effects) broke the use of negated acl sub-conditions.
1717 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1720 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1721 "Deny" causes Exim to reject the incoming connection with a 554 error.
1722 Unfortunately, if there is a major crisis, such as a disk failure,
1723 tcp-wrappers gives "deny", whereas what one would like would be some
1724 kind of temporary error. A kludge has been added to help with this.
1725 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1726 554 error is used if errno is still zero or contains ENOENT (which occurs
1727 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1730 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1732 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1733 errors. Otherwise a message that provokes a temporary error (when other
1734 messages do not) can cause a whole host to time out.
1736 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1737 addresses were routed directly to files or pipes from a redirect router.
1738 File deliveries just didn't batch; pipe deliveries might have suffered
1741 PH/06 A failure to get a lock for a hints database would erroneously always say
1742 "Failed to get write lock", even when it was really a read lock.
1744 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1745 of 0600. This has been changed to use the value of the lockfile_mode
1746 option (which defaults to 0600).
1748 PH/08 Applied small patch from the Sieve maintainer.
1750 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1751 folder from quota calculations, a direct delivery into this folder messed
1752 up the contents of the maildirsize file. This was because the regex was
1753 used only to exclude .Trash (or whatever) when the size of the mailbox
1754 was calculated. There was no check that a delivery was happening into an
1755 excluded directory. This bug has been fixed by ignoring all quota
1756 processing for deliveries into excluded directories.
1758 PH/10 Added the maildirfolder_create_regex option to appendfile.
1764 PH/01 The code for finding all the local interface addresses on a FreeBSD
1765 system running IPv6 was broken. This may well have applied to all BSD
1766 systems, as well as to others that have similar system calls. The broken
1767 code found IPv4 interfaces correctly, but gave incorrect values for the
1768 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1769 that it would not match correctly against @[] and not recognize the IPv6
1772 PH/02 The ipliteral router was not recognizing addresses of the form user@
1773 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1775 PH/03 Added disable_ipv6.
1777 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1778 decoded form, because it is most often used to construct To: headers
1779 lines in autoreplies, and the decoded form may well be syntactically
1780 invalid. However, $reply_address has leading white space removed, and all
1781 newlines turned into spaces so that the autoreply transport does not
1784 PH/05 If group was specified without a user on a router, and no group or user
1785 was specified on a transport, the group from the router was ignored.
1787 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1788 for visible compile-time settings that can be used to change these
1789 numbers, for those that want even more. Backwards compatibility with old
1790 spool files has been maintained. However, going back to a previous Exim
1791 release will lost any variables that are in spool files.
1793 PH/07 Two small changes when running in the test harness: increase delay when
1794 passing a TCP/IP connection to a new process, in case the original
1795 process has to generate a bounce, and remove special handling of
1796 127.0.0.2 (sic), which is no longer necessary.
1798 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1799 be the same on different OS.
1801 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1804 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1805 whether --show-vars was specified or not
1807 JJ/02 exipick: Added support for new ACL variable spool format introduced
1810 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1811 syntactically invalid From: or Reply-to: line, and a filter used this to
1812 generate an autoreply, and therefore failed to obtain an address for the
1813 autoreply, Exim could try to deliver to a non-existent relative file
1814 name, causing unrelated and misleading errors. What now happens is that
1815 it logs this as a hard delivery error, but does not attempt to create a
1818 PH/11 The exinext utility has a -C option for testing purposes, but although
1819 the given file was scanned by exinext itself; it wasn't being passed on
1820 when Exim was called.
1822 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1823 an end-of-file indication when reading a command response.
1825 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1826 compiled. In many other places in Exim, IPv6 addresses are always
1827 recognized, so I have changed this. It also means that IPv4 domain
1828 literals of the form [IPV4:n.n.n.n] are now always recognized.
1830 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1831 used if the router is not running as root, for example, when verifying at
1832 ACL time, or when using -bh. The debugging output from this situation was
1833 non-existent - all you got was a failure to exec. I have made two
1836 (a) Failures to set uid/gid, the current directory, or a process leader
1837 in a subprocess such as that created by queryprogram now generate
1838 suitable debugging ouput when -d is set.
1840 (b) The queryprogram router detects when it is not running as root,
1841 outputs suitable debugging information if -d is set, and then runs
1842 the subprocess without attempting to change uid/gid.
1844 PH/15 Minor change to Makefile for building test_host (undocumented testing
1847 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1848 additional section of a DNS packet that returns MX or SRV records.
1849 Instead, it always explicitly searches for A/AAAA records. This avoids
1850 major problems that occur when a DNS server includes only records of one
1851 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1852 fixed another bug: if SRV records were looked up and the corresponding
1853 address records were *not* found in the additional section, the port
1854 values from the SRV records were lost.
1856 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1857 using the correct key (the original address) when searching the retry
1858 rules in order to find which one to use for generating the retry hint.
1860 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1861 adding the default one. Similarly, if it contains a Reply-To: header, the
1862 errors_reply_to option, if set, is not used.
1864 PH/19 When calculating a retry time, Exim used to measure the "time since
1865 failure" by looking at the "first failed" field in the retry record. Now
1866 it does not use this if it is later than than the arrival time of the
1867 message. Instead it uses the arrival time. This makes for better
1868 behaviour in cases where some deliveries succeed, thus re-setting the
1869 "first failed" field. An example is a quota failure for a huge message
1870 when small messages continue to be delivered. Without this change, the
1871 "time since failure" will always be short, possible causing more frequent
1872 delivery attempts for the huge message than are intended.
1873 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1875 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1876 $1, $2, $3) because the numerical variables can be reset during some
1877 expansion items (e.g. "match"), thereby losing the authentication data.
1879 PH/21 Make -bV show the size of off_t variables so that the test suite can
1880 decide whether to run tests for quotas > 2G.
1882 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1883 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1884 filecount value is greater than 2G or if a quota value is greater than 2G
1885 on a system where the size of off_t is not greater than 4, a panic error
1888 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1889 never match. The debug and -bh output now contains an explicit error
1890 message indicating a malformed IPv4 address or mask.
1892 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1893 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1894 PH/23 above applies.
1896 PH/25 Do not write to syslog when running in the test harness. The only
1897 occasion when this arises is a failure to open the main or panic logs
1898 (for which there is an explicit test).
1900 PH/26 Added the /no_tell option to "control=freeze".
1902 PH/27 If a host name lookup failed very early in a connection, for example, if
1903 the IP address matched host_lookup and the reverse lookup yielded a name
1904 that did not have a forward lookup, an error message of the form "no IP
1905 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1906 could be logged. Now it outputs the IP address instead of "NULL".
1908 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1909 allows the sender and the authenticated sender to be set when
1910 submitting a message from within Exim. Since child_open_exim() is
1911 documented for local_scan(), the new function should be too.
1913 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1914 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1915 results in an empty string is now treated as unset.
1917 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1919 PH/31 Added sender_verify_fail logging option.
1921 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1922 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1923 tidied the source and removed it altogether.
1925 PH/33 When a queue run was abandoned because the load average was too high, a
1926 log line was always written; now it is written only if the queue_run log
1927 selector is set. In addition, the log line for abandonment now contains
1928 information about the queue run such as the pid. This is always present
1929 in "start" and "stop" lines but was omitted from the "abandon" line.
1931 PH/34 Omit spaces between a header name and the colon in the error message that
1932 is given when verify = headers_syntax fails (if there are lots of them,
1933 the message gets confusing).
1935 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1936 names, as there are now some PTR records that contain slashes. This check
1937 is only to protect against broken name servers that fall over on strange
1938 characters, so the fact that it applies to all lookups doesn't matter.
1940 PH/36 Now that the new test suite is complete, we can remove some of the
1941 special code in Exim that was needed for the old test suite. For example,
1942 sorting DNS records because real resolvers return them in an arbitrary
1943 order. The new test suite's fake resolver always returns records in the
1946 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1947 bounces) except when queue_only is set, to avoid logging races between
1948 the different processes.
1950 PH/38 Panic-die if .include specifies a non-absolute path.
1952 PH/39 A tweak to the "H" retry rule from its user.
1954 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1955 a label. They prevented compilation on older perls.
1957 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1958 a warning to be raised on newish perls.
1960 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1961 on queue. Changes to match documented behaviour of showing count of
1962 messages matching specified criteria.
1964 PH/40 Changed the default ident timeout from 30s to 5s.
1966 PH/41 Added support for the use of login_cap features, on those BSD systems
1967 that have them, for controlling the resources used by pipe deliveries.
1969 PH/42 The content-scanning code uses fopen() to create files in which to put
1970 message data. Previously it was not paying any attention to the mode of
1971 the files. Exim runs with umask(0) because the rest of the code creates
1972 files with open(), and sets the required mode explicitly. Thus, these
1973 files were ending up world-writeable. This was not a big issue, because,
1974 being within the spool directory, they were not world-accessible. I have
1975 created a function called modefopen, which takes an additional mode
1976 argument. It sets umask(777), creates the file, chmods it to the required
1977 mode, then resets the umask. All the relevant calls to fopen() in the
1978 content scanning code have been changed to use this function.
1980 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1981 to 24 hours. This avoids potential overflow problems when processing G
1982 and H retry rules. I suspect nobody ever tinkers with this value.
1984 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1986 PH/45 When the plaintext authenticator is running as a client, the server's
1987 challenges are checked to ensure they are valid base64 strings. By
1988 default, the authentication attempt is cancelled if an invalid string is
1989 received. Setting client_ignore_invalid_base64 true ignores these errors.
1990 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1991 they are received. Thus, the responses can be made to depend on the
1992 challenges. If an invalid string is ignored, an empty string is placed in
1995 PH/46 Messages that are created by the autoreply transport now contains a
1996 References: header, in accordance with RFCs 2822 and 3834.
1998 PH/47 Added authenticated_sender_force to the smtp transport.
2000 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2002 PH/49 Installed latest patch from the Sieve maintainer.
2004 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2005 was also set, the appendfile transport was unnecessarily scanning a
2006 directory of message files (e.g. for maildir delivery) to find the count
2007 of files (along with the size), even though it did not need this
2008 information. It now does the scan only if it needs to find either the
2009 size of the count of files.
2011 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2013 PH/52 Two bugs concerned with error handling when the smtp transport is
2016 (i) Exim was not creating retry information for temporary errors given
2017 for individual recipients after the DATA command when the smtp transport
2018 was used in LMTP mode. This meant that they could be retried too
2019 frequently, and not timed out correctly.
2021 (ii) Exim was setting the flag that allows error details to be returned
2022 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2023 recipients that were returned after the DATA command.
2025 PH/53 This is related to PH/52, but is more general: for any failing address,
2026 when detailed error information was permitted to be returned to the
2027 sender, but the error was temporary, then after the final timeout, only
2028 "retry timeout exceeded" was returned. Now it returns the full error as
2029 well as "retry timeout exceeded".
2031 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2032 do this, and (what is worse) MTAs that accept it.
2034 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2035 will now be deprecated.
2037 PH/56 New os.c-cygwin from the Cygwin maintainer.
2039 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2040 formats (previously only available in exim formats via -bpr, -bpru,
2041 and -bpra. Now also available in native and exiqgrep formats)
2043 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2044 with very large, slow to parse queues
2046 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2048 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2050 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2051 responses to authentication challenges, though it was showing the
2052 challenges; (ii) I've removed the CR characters from the debug output for
2055 PH/58 Allow for the insertion of a newline as well as a space when a string
2056 is turned into more than one encoded-word during RFC 2047 encoding. The
2057 Sieve code now uses this.
2059 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2060 data_4xx, lost_connection, tls_required.
2062 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2063 message was being sent as an SMTP response.
2065 PH/61 Add -l and -k options to exicyclog.
2067 PH/62 When verifying, if an address was redirected to one new address, so that
2068 verification continued, and the new address failed or deferred after
2069 having set something in $address_data, the value of $address_data was not
2070 passed back to the ACL. This was different to the case when no
2071 redirection occurred. The value is now passed back in both cases.
2073 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2074 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2075 use login_cap.h, so on its own it isn't the distinguishing feature. The
2076 new name refers directly to the setclassresources() function.
2078 PH/65 Added configuration files for NetBSD3.
2080 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2082 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2083 is preferred over IPv4.
2085 PH/68 The bounce_return_message and bounce_return_body options were not being
2086 honoured for bounces generated during the reception of non-SMTP messages.
2087 In particular, this applied to messages rejected by the ACL. This bug has
2088 been fixed. However, if bounce_return_message is true and bounce_return_
2089 body is false, the headers that are returned for a non-SMTP message
2090 include only those that have been read before the error was detected.
2091 (In the case of an ACL rejection, they have all been read.)
2093 PH/69 The HTML version of the specification is now built in a directory called
2094 spec_html instead of spec.html, because the latter looks like a path with
2095 a MIME-type, and this confuses some software.
2097 PH/70 Catch two compiler warnings in sieve.c.
2099 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2100 function verify_get_ident() calls ip_connect() to connect a socket, but
2101 if the "connect()" function timed out, ip_connect() used to close the
2102 socket. However, verify_get_ident() also closes the socket later, and in
2103 between Exim writes to the log, which may get opened at this point. When
2104 the socket was closed in ip_connect(), the log could get the same file
2105 descriptor number as the socket. This naturally causes chaos. The fix is
2106 not to close the socket in ip_connect(); the socket should be closed by
2107 the function that creates it. There was only one place in the code where
2108 this was missing, in the iplookup router, which I don't think anybody now
2109 uses, but I've fixed it anyway.
2111 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2112 well as to direct DNS lookups. Otherwise the handling of names in host
2113 lists is inconsistent and therefore confusing.
2119 PH/01 Two changes to the default runtime configuration:
2121 (1) Move the checks for relay_from_hosts and authenticated clients from
2122 after to before the (commented out) DNS black list checks.
2124 (2) Add control=submission to the relay_from_hosts and authenticated
2125 clients checks, on the grounds that messages accepted by these
2126 statements are most likely to be submissions.
2128 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2130 (1) Generate an error if the third argument for the ${prvs expansion is
2133 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2136 (3) Reset the variables that are obtained from the first argument of
2137 ${prvscheck and used in the second argument before leaving the code,
2138 because their memory is reclaimed, so using them afterwards may do
2141 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2142 one (it's much easier than Tom thought :-).
2144 (5) Because of (4), we can now allow for the use of $prvscheck_result
2145 inside the third argument.
2147 PH/03 For some reason, the default setting of PATH when running a command from
2148 a pipe transport was just "/usr/bin". I have changed it to
2151 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2152 anything to be listed in the output from -bV.
2154 PH/05 When a filter generated an autoreply, the entire To: header line was
2155 quoted in the delivery log line, like this:
2157 => >A.N.Other <ano@some.domain> <original@ddress> ...
2159 This has been changed so that it extracts the operative address. There
2160 may be more than one such address. If so, they are comma-separated, like
2163 => >ano@some.domain,ona@other.domain <original@ddress> ...
2165 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2166 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2167 not being looked up in the rDNS to get a host name, Exim was showing the
2168 IP address twice in Received: lines, even though the IP addresses were
2169 identical. For example:
2171 Received: from [1.2.3.4] (helo=[1.2.3.4])
2173 However, if the real host name was known, it was omitting the HELO data
2174 if it matched the actual IP address. This has been tidied up so that it
2175 doesn't show the same IP address twice.
2177 PH/07 When both +timestamp and +memory debugging was on, the value given by
2178 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2179 called by the debug printing, thereby overwriting the timestamp buffer.
2180 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2182 PH/08 When the original message was included in an autoreply transport, it
2183 always said "this is a copy of the message, including all the headers",
2184 even if body_only or headers_only was set. It now gives an appropriate
2187 PH/09 Applied a patch from the Sieve maintainer which:
2189 o fixes some comments
2190 o adds the (disabled) notify extension core
2191 o adds some debug output for the result of if/elsif tests
2192 o points to the current vacation draft in the documentation
2193 and documents the missing references header update
2197 o fixes a bug in processing the envelope test (when testing
2198 multiple envelope elements, the last element determinted the
2201 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2202 Electronic Mail") by including:
2204 Auto-submitted: auto-generated
2206 in the messages that it generates (bounce messages and others, such as
2207 warnings). In the case of bounce messages for non-SMTP mesages, there was
2208 also a typo: it was using "Auto_submitted" (underscore instead of
2209 hyphen). Since every message generated by Exim is necessarily in response
2210 to another message, thes have all been changed to:
2212 Auto-Submitted: auto-replied
2214 in accordance with these statements in the RFC:
2216 The auto-replied keyword:
2218 - SHOULD be used on messages sent in direct response to another
2219 message by an automatic process,
2221 - MUST NOT be used on manually-generated messages,
2223 - MAY be used on Delivery Status Notifications (DSNs) and Message
2224 Disposition Notifications (MDNs),
2226 - MUST NOT be used on messages generated by automatic or periodic
2227 processes, except for messages which are automatic responses to
2230 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2231 to the default Received: header definition.
2233 PH/12 Added log selector acl_warn_skipped (default on).
2235 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2236 variables because (a) they are in the wrong storage pool and (b) even if
2237 they were copied, it wouldn't work properly because of the caching.
2239 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2240 checking when decoding. Apparently there are clients that generate
2241 overlong encoded strings. Why am I not surprised?
2243 PH/15 If the first argument of "${if match_address" was not empty, but did not
2244 contain an "@" character, Exim crashed. Now it writes a panic log message
2245 and treats the condition as false.
2247 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2249 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2250 extension "envelope-auth". The code is finished and in agreement with
2251 other implementations, but there is no documentation so far and in fact,
2252 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2253 not changing the active code.
2255 Print executed "if" and "elsif" statements when debugging is used. This
2256 helps a great deal to understand what a filter does.
2258 Document more things not specified clearly in RFC3028. I had all this
2259 sorted out, when out of a sudden new issues came to my mind. Oops."
2261 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2264 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2265 canonical form (as documented). However, after a host name lookup from
2266 the IP address, check_host() was doing a simple string comparison with
2267 addresses acquired from the DNS when checking that the found name did
2268 have the original IP as one of its addresses. Since any found IPv6
2269 addresses are likely to be in abbreviated form, the comparison could
2270 fail. Luckily, there already exists a function for doing the comparison
2271 by converting both addresses to binary, so now that is used instead of
2272 the text comparison.
2274 PH/20 There was another similar case to PH/19, when a complete host name was
2275 given in a host list; looking up its IP address could give an abbreviated
2276 form, whereas the current host's name might or might not be abbreviated.
2277 The same fix has been applied.
2283 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2284 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2287 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2288 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2290 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2292 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2293 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2294 Also, for "Auto-Submitted", treat anything other than "no" as
2295 non-personal, in accordance with RFC 3834. (Previously it treated
2296 anything starting "auto-" as non-personal.)
2298 TF/01 The control=submission/name=... option had a problem with syntax
2299 errors if the name included a slash character. The /name= option
2300 now slurps the rest of the string, so it can include any characters
2301 but it must come last in the list of options (after /sender_retain
2304 PH/05 Some modifications to the interface to the fake nameserver for the new
2312 TK/01 Added the "success_on_redirect" address verification option. See
2313 NewStuff for rationale and an example.
2315 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2317 PH/02 Patch to exigrep to allow it to work on syslog lines.
2319 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2320 fread() to skip over the body file's header line, because in Cygwin the
2321 header line is locked and is inaccessible.
2323 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2324 co-exist for some time) to make it clear that it is the Exim ID that is
2325 referenced, not the Message-ID: header line.
2327 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2328 string_format() function, because snprintf() does not exist on all
2331 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2332 ${stat: expansion item.
2334 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2335 protocol synchronization error", to keep the pedants happy.
2337 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2338 well as for IRIX systems, when gcc is being used. See the host.c source
2341 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2343 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2346 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2347 transport and to the smtp transport in LMTP mode.
2349 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2351 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2352 run a filter in a subprocess. This could lead to confusion in subsequent
2353 lookups in the parent process. There should also be a search_tidyup() at
2354 the end of the subprocess.
2356 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2357 only if the host matched helo_try_verify_hosts, which caused the
2358 verification to occur when the EHLO/HELO command was issued. The ACL just
2359 tested the remembered result. Now, if a previous verification attempt has
2360 not happened, "verify = helo" does it there and then.
2362 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2364 TK/03 Fix log output including CR from clamd.
2366 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2367 exist provoked a memory error which could cause a segfault.
2369 PH/15 Installed PCRE 6.2
2371 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2373 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2374 of the problem. Specifically, suggested +O2 rather than +O1 for the
2377 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2379 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2380 record, the port was not correctly logged when the outgoing_port log
2381 selector was set (it logged the transort's default port).
2383 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2384 fallback_hosts, and "hosts" in the smtp transport.
2386 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2387 host errors such as "Connection refused".
2389 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2390 authentication with radiusclient 0.4.9:
2392 - Error returned from rc_read_config was caught wrongly
2393 - Username/password not passed on to radius server due to wrong length.
2395 The presumption is that some radiusclient API changes for 4.51/PH/17
2396 were not taken care of correctly. The code is still untested by me (my
2397 Linux distribution still has 0.3.2 of radiusclient), but it was
2398 contributed by a Radius user.
2400 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2401 expanding the "port" option of the smtp transport.
2403 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2404 while reading a MIME header. Thanks to Tom Hughes for a patch.
2406 PH/24 Include config.h inside local_scan.h so that configuration settings are
2409 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2410 that in an ACL for RCPT (for example), you can examine exactly what was
2413 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2414 commands, but it was not correctly comparing the address with the actual
2415 client host address. Thus, it would show the EHLO address in Received:
2416 header lines when this was not necessary.
2418 PH/27 Added the % operator to ${eval:}.
2420 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2421 it should be ignoring failures (because with -C, for example, it has lost
2422 privilege). It wasn't ignoring creation failures other than "already
2425 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2428 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2429 because an input line was too long, either on its own, or by virtue of
2430 too many continuations, the temporary file was not being removed, and the
2431 return code was incorrect.
2433 PH/31 Missing "BOOL" in function definition in filtertest.c.
2435 PH/32 Applied Sieve patches from the maintainer.
2437 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2439 PH/33 Added "verify = not_blind".
2441 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2442 Local/Makefile (with some defaults set). These are used in built scripts
2443 such as exicyclog, but they have never been used in the exim_install
2444 script (though there are many overriding facilities there). I have
2445 arranged that the exim_install script now takes note of these two
2448 PH/35 Installed configuration files for Dragonfly.
2450 PH/36 When a locally submitted message by a trusted user did not contain a
2451 From: header, and the sender address was obtained from -f or from an SMTP
2452 MAIL command, and the trusted user did not use -F to supply a sender
2453 name, $originator_name was incorrectly used when constructing a From:
2454 header. Furthermore, $originator_name was used for submission mode
2455 messages from external hosts without From: headers in a similar way,
2456 which is clearly wrong.
2458 PH/37 Added control=suppress_local_fixups.
2460 PH/38 When log_selector = +received_sender was set, and the addition of the
2461 sender made the log line's construction buffer exactly full, or one byte
2462 less than full, an overflow happened when the terminating "\n" was
2465 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2466 when the result of a list match is failure because a DNS lookup failed.
2468 PH/40 RM_COMMAND is now used in the building process.
2470 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2471 the "build-* directories that it finds.
2473 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2474 address in a domain literal was a prefix of an interface address.
2476 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2477 when verifying a sender address, unless rewrite_headers is false.
2479 PH/44 Wrote a long comment about why errors_to addresses are verified as
2480 recipients, not senders.
2482 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2483 the ratelimit ACL was added.
2485 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2487 PH/46 Added extra information about PostgreSQL errors to the error string.
2489 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2490 avoiding the need to install special zones in a real server. This is
2491 backwards compatible; if it can't find the fake resolver, it drops back.
2492 Thus, both old and new test suites can be run.
2494 TF/02 Added util/ratelimit.pl
2496 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2497 clock is set back in time.
2499 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2500 Candler <B.Candler@pobox.com>.
2502 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2503 OK for addresses that are the result of redirections.
2505 PH/48 A number of further additions for the benefit of the new test suite,
2506 including a fake gethostbyname() that interfaces to the fake DNS resolver
2509 TF/06 The fix for widen_domains has also been applied to qualify_single and
2510 search_parents which are the other dnslookup options that can cause
2513 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2516 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2518 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2519 service" if it encountered an MX record with an empty target hostname.
2520 The message is now "an MX or SRV record indicated no SMTP service".
2522 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2523 if the DNS of the sending site is misconfigured. This is quite a
2524 common situation. This change restores the behaviour of treating a
2525 helo verification defer as a failure.
2527 PH/51 If self=fail was set on a router, the bounce message did not include the
2528 actual error message.
2534 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2536 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2537 command itself ended in error, the underlying message about the transport
2538 filter timeout was being overwritten with the pipe command error. Now the
2539 underlying error message should be appended to the second error message.
2541 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2543 PH/02 Reduce the amount of output that "make" produces by default. Full output
2544 can still be requested.
2546 PH/03 The warning log line about a condition test deferring for a "warn" verb
2547 was being output only once per connection, rather than after each
2548 occurrence (because it was using the same function as for successful
2549 "warn" verbs). This seems wrong, so I have changed it.
2551 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2552 it should not have, which might have caused a crash in the right
2553 circumstances, but probably never did.
2555 PH/04 Installed a modified version of Tony Finch's patch to make submission
2556 mode fix the return path as well as the Sender: header line, and to
2557 add a /name= option so that you can make the user's friendly name appear
2560 TF/03 Added the control = fakedefer ACL modifier.
2562 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2563 Mark Lowes for thorough testing.
2565 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2567 TK/03 Merged latest SRS patch from Miles Wilton.
2569 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2570 with the definition in sysexits.h (which is #included earlier).
2571 Fortunately, Exim does not actually use EX_OK. The code used to try to
2572 preserve the sysexits.h value, by assumimg that macro definitions were
2573 scanned for macro replacements. I have been disabused of this notion,
2574 so now the code just undefines EX_OK before #including unistd.h.
2576 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2577 in the smtp transport. When a block could not be written in a single
2578 write() function, the timeout was being re-applied to each part-write.
2579 This seems wrong - if the receiver was accepting one byte at a time it
2580 would take for ever. The timeout is now adjusted when this happens. It
2581 doesn't have to be particularly precise.
2583 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2584 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2586 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2587 without a domain if the check to <postmaster@domain> fails.
2589 SC/01 Eximstats: added -xls and the ability to specify output files
2590 (patch written by Frank Heydlauf).
2592 SC/02 Eximstats: use FileHandles for outputing results.
2594 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2596 SC/04 Eximstats: fixed display of large numbers with -nvr option
2598 SC/05 Eximstats: fixed merging of reports with empty tables.
2600 SC/06 Eximstats: added the -include_original_destination flag
2602 SC/07 Eximstats: removed tabs and trailing whitespace.
2604 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2606 TK/06 MBOX spool code: Add real "From " MBOX separator line
2607 so the .eml file is really in mbox format (even though
2608 most programs do not really care). Patch from Alex Miller.
2610 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2611 The latter is generated from $received_to and is only set if the
2612 message has one envelope recipient. SA can use these headers,
2613 obviously out-of-the-box. Patch from Alex Miller.
2615 PH/08 The ${def test on a variable was returning false if the variable's
2616 value was "0", contrary to what the specification has always said!
2617 The result should be true unless the variable is empty.
2619 PH/09 The syntax error of a character other than { following "${if
2620 def:variable_name" (after optional whitespace) was not being diagnosed.
2621 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2622 accidental colon was present, for example, could give incorrect results.
2624 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2625 result is used (not including appendfile, where other changes are about
2628 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2629 This involved changing a lot of size variables from int to off_t. It
2630 should work with maildirs and everything.
2632 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2633 spamd dying while we are connected to it.
2635 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2638 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2639 function for BDB 4.3.
2641 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2643 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2644 the sources for PCRE have been re-arranged and more files are now
2647 PH/15 The code I had for printing potentially long long variables in PH/11
2648 above was not the best (it lost precision). The length of off_t variables
2649 is now inspected at build time, and an appropriate printing format (%ld
2650 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2651 to be "long long int" or "long int". This is needed for the internal
2652 formatting function string_vformat().
2654 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2655 the configuration file to be ":syslog", then the script "guesses" where
2656 the logs files are, rather than using the compiled in default. In our
2657 case the guess is not the same as the compiled default, so the script
2658 suddenly stopped working when I started to use syslog. The patch checks
2659 to see if log_file_path is "". If so, it attempts to read it from exim
2660 with no configuration file to get the compiled in version, before it
2661 falls back to the previous guessing code."
2663 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2664 implementing BATV in an Exim configuration. See NewStuff for the gory
2667 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2668 Makefile that are specific to HP-UX.
2670 PH/18 If the "use_postmaster" option was set for a recipient callout together
2671 with the "random" option, the postmaster address was used as the MAIL
2672 FROM address for the random test, but not for the subsequent recipient
2673 test. It is now used for both.
2675 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2676 patch removes a few documentation additions to RFC 3028, because the
2677 latest draft now contains them. It adds the new en;ascii-case comparator
2678 and a new error check for 8bit text in MIME parts. Comparator and
2679 require names are now matched exactly. I enabled the subaddress
2680 extension, but it is not well tested yet (read: it works for me)."
2682 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2683 rework some of the code of TK/09 above to avoid the hardwired use of
2684 "%lld" and "long long". Replaced the call to snprintf() with a call to
2687 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2688 records point to non-existent hosts", "retry timeout exceeded", and
2689 "retry time not reached for any host after a long failure period".
2691 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2692 experimental DomainKeys support:
2694 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2695 (2) On an error such as an illegally used "control", the wrong name for
2696 the control was given.
2698 These problems did NOT occur unless DomainKeys support was compiled.
2700 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2702 PH/24 Added ${if match_ip condition.
2704 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2705 databases so that it will be absolutely obvious if a crash occurs in the
2706 DB library. This is a regular occurrence (often caused by mis-matched
2709 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2710 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2711 picked up on a user's system that detects such things. There doesn't seem
2712 to be a gcc warning option for this - only an attribute that has to be
2713 put on the function's prototype. It seems that in Fedora Core 4 they have
2714 set this on a number of new functions. No doubt there will be more in due
2717 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2718 specify a transport. However, if an address that was verified by such a
2719 router was the subject of a callout, Exim crashed because it tried to
2720 read the rcpt_include_affixes from the non-existent transport. Now it
2721 just assumes that the setting of that option is false. This bug was
2722 introduced by 4.51/PH/31.
2724 PH/28 Changed -d+all to exclude +memory, because that information is very
2725 rarely of interest, but it makes the output a lot bigger. People tend to
2726 do -d+all out of habit.
2728 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2729 code in os-type was giving problems when libc.so lives in lib64, like on
2732 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2733 aren't the modern standard, and it seems that some systems' include files
2734 don't always have them. Exim was already checking for some of the newer
2735 ones like T_AAAA, and defining it itself. I've added checks for all the
2736 record types that Exim uses.
2738 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2739 not automatically generating a new one, as it is supposed to. This
2740 prevented TLS from working. If the file did exist, but contained invalid
2741 data, a new version was generated, as expected. It was only the case of a
2742 non-existent file that was broken.
2744 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2745 with a change in libdomainkeys > 0.64.
2747 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2748 from DNS. If the selector record carries the flag, it now has
2749 precedence over the domain-wide flag.
2751 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2753 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2754 the use of a port name that isn't defined in /etc/services) occurred, the
2755 message was deferred as in a normal delivery, and thus remained on the
2756 spool, instead of being failed because of the mua_wrapper setting. This
2757 is now fixed, and I tidied up some of the mua_wrapper messages at the
2760 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2761 the messages in a hash of arrays rather than using individual hashes.
2762 This is a bit cleaner and results in dramatic memory savings, albeit
2763 at a slight CPU cost.
2765 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2766 as requested by Marc Sherman.
2768 SC/10 Eximstats: added histograms for user specified patterns as requested
2771 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2773 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2774 fopen() in the content-scanning modules that did not already have it.
2780 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2781 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2783 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2785 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2787 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2788 to test Sieve filters that use "vacation".
2790 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2791 that changes the way the GnuTLS parameters are stored in the cache file.
2792 The new format can be generated externally. For backward compatibility,
2793 if the data in the cache doesn't make sense, Exim assumes it has read an
2794 old-format file, and it generates new data and writes a new file. This
2795 means that you can't go back to an older release without removing the
2798 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2799 work if there are any delivery delays because "one_time" forces the
2800 parent to be marked "delivered", so its unseen clone is never tried
2801 again. For this reason, Exim now forbids the simultaneous setting of
2804 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2805 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2806 entry if you want to know the details. The fix, however, neglected to
2807 consider the case when local delivery batching is involved. The test for
2808 "previously delivered" was not happening when checking to see if an
2809 address could be batched with a previous (undelivered) one; under
2810 certain circumstances this could lead to multiple deliveries to the same
2813 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2814 in its include files, and this causes problems building Exim.
2816 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2817 header_syntax) but Exim was just ignoring anything given after a slash.
2818 In particular, this caused confusion with an attempt to use "verify =
2819 reverse_host_lookup/defer_ok". An error is now given when options are
2820 supplied for verify items that do not have them. (Maybe reverse_host_
2821 lookup should have a defer_ok option, but that's a different point.)
2823 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2824 defined by RFC 821) to 2048, because there were problems with some AUTH
2825 commands, and RFC 1869 says the size should be increased for extended
2826 SMTP commands that take arguments.
2828 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2831 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2832 "unknown" error; now it says that the functionality isn't in the binary.
2834 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2835 an address' error message when a string expansion fails (syntax or
2836 whatever). Otherwise the password may appear in the log. Following change
2837 PH/42 below, there is no longer a chance of it appearing in a bounce
2840 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2842 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2843 of its addresses was ever tried. (Bugzilla bug #2.)
2845 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2846 the result incorrectly in the debug output. (It correctly added a newline
2847 to what was transported.)
2849 TF/01 Added $received_time.
2851 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2852 commented out examples of how to interface to a virus scanner and to
2853 SpamAssassin. Also added commented examples of av_scanner and
2854 spamd_address settings.
2856 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2857 and controls are allowed in which ACLs. There were a couple of minor
2858 errors. Some of the entries in the conditions table (which is a table of
2859 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2860 negation of where the condition IS allowed.
2862 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2864 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2865 header file does not have a version number, so I've had to invent a new
2866 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2867 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2868 radiusclient), but it was contributed by a Radius user.
2870 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2871 files or directories, for OpenSSL.
2873 PH/19 When an Exim process that is running as root has to create an Exim log
2874 file, it does so in a subprocess that runs as exim:exim so as to get the
2875 ownership right at creation (otherwise, other Exim processes might see
2876 the file with the wrong ownership). There was no test for failure of this
2877 fork() call, which would lead to the process getting stuck as it waited
2878 for a non-existent subprocess. Forks do occasionally fail when resources
2879 run out. I reviewed all the other calls to fork(); they all seem to check
2882 PH/20 When checking for unexpected SMTP input at connect time (before writing
2883 the banner), Exim was not dealing correctly with a non-positive return
2884 from the read() function. If the client had disconnected by this time,
2885 the result was a log entry for a synchronization error with an empty
2886 string after "input=" when read() returned zero. If read() returned -1
2887 (an event I could not check), uninitialized data bytes were printed.
2888 There were reports of junk text (parts of files, etc) appearing after
2891 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2893 PH/22 Added support for macro redefinition, and (re)definition in between
2894 driver and ACL definitions.
2896 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2897 forgetting to use the resulting value; it was using the unexpanded value.
2899 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2900 hadn't been configured. The fix is from Juergen Kreileder, who
2901 understands it better than I do:
2903 "Here's what I see happening with three configured cyrus_sasl
2904 authenticators configured (plain, login, cram-md5):
2906 On startup auth_cyrus_sasl_init() gets called for each of these.
2907 This means three calls to sasl_listmech() without a specified mech_list.
2908 => SASL tests which mechs of all available mechs actually work
2909 => three warnings about OTP not working
2910 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2912 With the patch, sasl_listmech() also gets called three times. But now
2913 SASL's mech_list option is set to the server_mech specified in the the
2914 authenticator. Or in other words, the answer from sasl_listmech()
2915 gets limited to just the mech you're testing for (which is different
2917 => the return list contains just 'plain' or 'login', 'cram-md5' or
2918 nothing depending on the value of ob->server_mech.
2920 I've just tested the patch: Authentication still works fine,
2921 unavailable mechs specified in the exim configuration are still
2922 caught, and the auth.log warnings about OTP are gone."
2924 PH/25 When debugging is enabled, the contents of the command line are added
2925 to the debugging output, even when log_selector=+arguments is not
2928 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2929 answer is "GNU", and only if the return is "GNU/something" is the answer
2932 PH/27 $acl_verify_message is now set immediately after the failure of a
2933 verification in an ACL, and so is available in subsequent modifiers. In
2934 particular, the message can be preserved by coding like this:
2936 warn !verify = sender
2937 set acl_m0 = $acl_verify_message
2939 Previously, $acl_verify_message was set only while expanding "message"
2940 and "log_message" when a very denied access.
2942 PH/28 Modified OS/os.c-Linux with
2944 -#ifndef OS_LOAD_AVERAGE
2945 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2947 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2948 nomenclature these days.)
2950 PH/29 Installed patch from the Sieve maintainer that adds the options
2951 sieve_useraddress and sieve_subaddress to the redirect router.
2953 PH/30 In these circumstances:
2954 . Two addresses routed to the same list of hosts;
2955 . First host does not offer TLS;
2956 . First host accepts first address;
2957 . First host gives temporary error to second address;
2958 . Second host offers TLS and a TLS session is established;
2959 . Second host accepts second address.
2960 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2961 and peerdn, if requested) that were in fact used only for the second
2964 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2965 attention to any local part prefix or suffix that was matched by the
2966 router that accepted the address. It now behaves in the same way as it
2967 does for delivery: the affixes are removed from the local part unless
2968 rcpt_include_affixes is set on the transport.
2970 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2971 timeout during the DATA phase of an incoming message.
2973 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2974 applied a patch sanctioned by the Sieve maintainer.
2976 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2977 the uid or gid is negative. A case of a negative gid caused this to be
2978 noticed. The fix allows for either to be negative.
2980 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2981 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2984 PH/36 The change PH/12 above was broken. Fixed it.
2986 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2987 the grounds that routing the same address twice would always produce the
2988 same answer. This might have been true once, but it is certainly no
2989 longer true now. Routing a child address may depend on the previous
2990 routing that produced that child. Some complicated redirection strategies
2991 went wrong when messages had multiple recipients, and made Exim's
2992 behaviour dependent on the order in which the addresses were given.
2994 I have moved the duplicate checking until after the routing is complete.
2995 Exim scans the addresses that are assigned to local and remote
2996 transports, and removes any duplicates. This means that more work will be
2997 done, as duplicates will always all be routed, but duplicates are
2998 presumably rare, so I don't expect this is of any significance.
3000 For deliveries to pipes, files, and autoreplies, the duplicate checking
3001 still happens during the routing process, since they are not going to be
3004 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3005 It corrects a timeout issue with spamd. This is Ian's comment: "The
3006 background is that sometimes spamd either never reads data from a
3007 connection it has accepted, or it never writes response data. The exiscan
3008 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3009 blindly assumes that writes won't block so it may never time out."
3011 PH/39 Allow G after quota size as well as K and M.
3013 PH/40 The value set for $authenticated_id in an authenticator may not contain
3014 binary zeroes or newlines because the value is written to log lines and
3015 to spool files. There was no check on this. Now the value is run through
3016 the string_printing() function so that such characters are converted to
3017 printable escape sequences.
3019 PH/41 $message_linecount is a new variable that contains the total number of
3020 lines in the message. Compare $body_linecount, which is the count for the
3023 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3024 bounce and delay warning messages, except in certain special cases, which
3027 (a) An SMTP error message from a remote host;
3028 (b) A message specified in a :fail: redirection;
3029 (c) A message specified in a "fail" command in a system filter;
3030 (d) A message specified in a FAIL return from the queryprogram router;
3031 (e) A message specified by the cannot_route_message router option.
3033 In these cases only, Exim does include the error details in bounce and
3034 warning messages. There are also a few cases where bland messages such
3035 as "unrouteable address" or "local delivery error" are given.
3037 PH/43 $value is now also set for the "else" part of a ${run expansion.
3039 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3040 being worked on, but at least Exim now implements the latest version to
3043 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3044 process to complete was treated as a delivery failure, a timeout while
3045 writing the message to the pipe was logged, but erroneously treated as a
3046 successful delivery. Such timeouts include transport filter timeouts. For
3047 consistency with the overall process timeout, these timeouts are now
3048 treated as errors, giving rise to delivery failures by default. However,
3049 there is now a new Boolean option for the pipe transport called
3050 timeout_defer, which, if set TRUE, converts the failures into defers for
3051 both kinds of timeout. A transport filter timeout is now identified in
3054 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3055 systems where "make" and "gmake" are different, calling "gmake" at top
3056 level broke things. I've arranged for the value of $(MAKE) to be passed
3057 from the Makefile to this script so that it can call the same version of
3061 A note about Exim versions 4.44 and 4.50
3062 ----------------------------------------
3064 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3065 changes of various kinds. As a consequence, a big documentation update was
3066 needed. This delayed the release for rather longer than seemed good, especially
3067 in the light of a couple of (minor) security issues. Therefore, the changes
3068 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3069 release. So 4.44 and 4.50 are in effect two different branches that both start
3072 I have left the 4.50 change log unchanged; it contains all the changes since
3073 4.43. The change log for 4.44 is below; many of its items are identical to
3074 those for 4.50. This seems to be the most sensible way to preserve the
3075 historical information.
3081 1. Minor wording change to the doc/README.SIEVE file.
3083 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3084 computation of the current number of files was incorrect.
3086 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3087 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3090 4. Give more explanation in the error message when the command for a transport
3091 filter fails to execute.
3093 5. There are several places where Exim runs a non-Exim command in a
3094 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3095 was being done only for the command run by the queryprogram router. It is
3096 now done for all such subprocesses. The other cases are: ${run, transport
3097 filters, and the commands run by the lmtp and pipe transports.
3099 6. Added CONFIGURE_GROUP build-time option.
3101 7. Some older OS have a limit of 256 on the maximum number of file
3102 descriptors. Exim was using setrlimit() to set 1000 as a large value
3103 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3104 systems. I've change it so that if it can't get 1000, it tries for 256.
3106 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3107 was an oversight, and furthermore, ever since the addition of extra
3108 controls (e.g. 4.43/32), the checks on when to allow different forms of
3109 "control" were broken. There should now be diagnostics for all cases when a
3110 control that does not make sense is encountered.
3112 9. Added the /retain_sender option to "control=submission".
3114 10. $recipients is now available in the predata ACL (oversight).
3116 11. Tidy the search cache before the fork to do a delivery from a message
3117 received from the command line. Otherwise the child will trigger a lookup
3118 failure and thereby defer the delivery if it tries to use (for example) a
3119 cached ldap connection that the parent has called unbind on.
3121 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3122 of $address_data from the recipient verification was clobbered by the
3123 sender verification.
3125 13. The value of address_data from a sender verification is now available in
3126 $sender_address_data in subsequent conditions in the ACL statement.
3128 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3130 15. Added a new option "connect=<time>" to callout options, to set a different
3133 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3134 was its contents. (It was OK if the option was not defined at all.)
3136 17. A "Completed" log line is now written for messages that are removed from
3137 the spool by the -Mrm option.
3139 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3140 information about exactly what failed.
3142 19. Added -dd to debug only the daemon process.
3144 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3145 handles timeouts, both on the server side and network timeouts. Renamed the
3146 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3148 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3149 It is now set to "smtps".
3151 22. $host_address is now set to the target address during the checking of
3152 ignore_target_hosts.
3154 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3155 being passed; this would have caused $sender_host_name to have been used if
3156 matching the list had actually called for a host name (not very likely,
3157 since this list is usually IP addresses). A host name is now passed as
3160 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3161 code that specifies a non-restarting handler (typically sigaction() in
3162 modern systems) in an attempt to fix a rare and obscure crash bug.
3164 25. Narrowed the window for a race in the daemon that could cause it to ignore
3165 SIGCHLD signals. This is not a major problem, because they are used only to
3166 wake it up if nothing else does.
3168 26. A malformed maildirsize file could cause Exim to calculate negative values
3169 for the mailbox size or file count. Odd effects could occur as a result.
3170 The maildirsize information is now recalculated if the size or filecount
3173 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3174 support for a long time. Removed HAVE_SYS_VFS_H.
3176 28. Installed the latest version of exipick from John Jetmore.
3178 29. In an address list, if the pattern was not a regular expression, an empty
3179 subject address (from a bounce message) matched only if the pattern was an
3180 empty string. Non-empty patterns were not even tested. This was the wrong
3181 because it is perfectly reasonable to use an empty address as part of a
3182 database query. An empty address is now tested by patterns that are
3183 lookups. However, all the other forms of pattern expect the subject to
3184 contain a local part and a domain, and therefore, for them, an empty
3185 address still always fails if the pattern is not itself empty.
3187 30. Exim went into a mad DNS loop when attempting to do a callout where the
3188 host was specified on an smtp transport, and looking it up yielded more
3189 than one IP address.
3191 31. Re-factored the code for checking spool and log partition space into a
3192 function that finds that data and another that does the check. The former
3193 is then used to implement four new variables: $spool_space, $log_space,
3194 $spool_inodes, and $log_inodes.
3196 32. The RFC2047 encoding function was originally intended for short strings
3197 such as real names; it was not keeping to the 75-character limit for
3198 encoded words that the RFC imposes. It now respects the limit, and
3199 generates multiple encoded words if necessary. To be on the safe side, I
3200 have increased the buffer size for the ${rfc2047: expansion operator from
3203 33. It is now permitted to omit both strings after an "if" condition; if the
3204 condition is true, the result is "true". As before, when the second string
3205 is omitted, a false condition yields an empty string. This makes it less
3206 cumbersome to write custom ACL and router conditions.
3208 34. Failure to deliver a bounce message always caused it to be frozen, even if
3209 there was an errors_to setting on the router. The errors_to setting is now
3212 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3213 canonical form (fully expanded) before being placed in
3214 $sender_host_address.
3216 36. The table in the code that translates DNS record types into text (T_A to
3217 "A" for instance) was missing entries for NS and CNAME. It is just possible
3218 that this could have caused confusion if both these types were looked up
3219 for the same domain, because the text type is used as part of Exim's
3220 per-process caching. But the chance of anyone hitting this buglet seems
3223 37. The dnsdb lookup has been extended in a number of ways.
3225 (1) There is a new type, "zns", which walks up the domain tree until it
3226 finds some nameserver records. It should be used with care.
3228 (2) There is a new type, "mxh", which is like "mx" except that it returns
3229 just the host names, not the priorities.
3231 (3) It is now possible to give a list of domains (or IP addresses) to be
3232 looked up. The behaviour when one of the lookups defers can be
3233 controlled by a keyword.
3235 (4) It is now possible to specify the separator character for use when
3236 multiple records are returned.
3238 38. The dnslists ACL condition has been extended: it is now possible to supply
3239 a list of IP addresses and/or domains to be looked up in a particular DNS
3242 39. Added log_selector=+queue_time_overall.
3244 40. When running the queue in the test harness, wait just a tad after forking a
3245 delivery process, to get repeatability of debugging output.
3247 41. Include certificate and key file names in error message when GnuTLS fails
3248 to set them up, because the GnuTLS error message doesn't include the name
3249 of the failing file when there is a problem reading it.
3251 42. Allow both -bf and -bF in the same test run.
3253 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3255 44. The "Exiscan patch" is now merged into the mainline Exim source.
3257 45. Sometimes the final signoff response after QUIT could fail to get
3258 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3259 < 0 before doing a fflush(). This bug looks as though it goes back to the
3260 introduction of TLS in release 3.20, but "sometimes" must have been rare
3261 because the tests only now provoked it.
3263 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3264 (this can affect the format of dates).
3266 47. exim_tidydb, when checking for the continued existence of a message for
3267 which it has found a message-specific retry record, was not finding
3268 messages that were in split spool directories. Consequently, it was
3269 deleting retry records that should have stayed in existence.
3271 48. Steve fixed some bugs in eximstats.
3273 49. The SPA authentication driver was not abandoning authentication and moving
3274 on to the next authenticator when an expansion was forced to fail,
3275 contradicting the general specification for all authenticators. Instead it
3276 was generating a temporary error. It now behaves as specified.
3278 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3279 (the order specifies the preference for clients). The order is now AES256,
3280 AES128, 3DES, ARCFOUR128.
3282 51. Small patch to Sieve code - explicitly set From: when generating an
3285 52. Exim crashed if a remote delivery caused a very long error message to be
3286 recorded - for instance if somebody sent an entire SpamAssassin report back
3287 as a large number of 550 error lines. This bug was coincidentally fixed by
3288 increasing the size of one of Exim's internal buffers (big_buffer) that
3289 happened as part of the Exiscan merge. However, to be on the safe side, I
3290 have made the code more robust (and fixed the comments that describe what
3293 53. Now that there can be additional text after "Completed" in log lines (if
3294 the queue_time_overall log selector is set), a one-byte patch to exigrep
3295 was needed to allow it to recognize "Completed" as not the last thing in
3298 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3299 patch that reportedly fixes this has been added. I am not expert enough to
3300 create a test for it. This is what the patch creator wrote:
3302 "I found a little strange behaviour of ldap code when working with
3303 Windows 2003 AD Domain, where users was placed in more than one
3304 Organization Units. When I tried to give exim partial DN, the exit code
3305 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3306 But simultaneously result of request was absolutely normal ldap result,
3307 so I produce this patch..."
3309 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3310 REFERENCE, so I have modified the code to exclude the patch when that macro
3313 55. Some experimental protocols are using DNS PTR records for new purposes. The
3314 keys for these records are domain names, not reversed IP addresses. The
3315 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3316 leaves it alone. Component reversal etc. now happens only for IP addresses.
3319 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3321 57. Double the size of the debug message buffer (to 2048) so that more of very
3322 long debug lines gets shown.
3324 58. The exicyclog utility now does better if the number of log files to keep
3325 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3327 59. Two changes related to the smtp_active_hostname option:
3329 (1) $smtp_active_hostname is now available as a variable.
3330 (2) The default for smtp_banner uses $smtp_active_hostname instead
3331 of $primary_hostname.
3333 60. The host_aton() function is supposed to be passed a string that is known
3334 to be a valid IP address. However, in the case of IPv6 addresses, it was
3335 not checking this. This is a hostage to fortune. Exim now panics and dies
3336 if the condition is not met. A case was found where this could be provoked
3337 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3338 components; fortuitously, this particular loophole had already been fixed
3339 by change 4.50/55 above.
3341 If there are any other similar loopholes, the new check in host_aton()
3342 itself should stop them being exploited. The report I received stated that
3343 data on the command line could provoke the exploit when Exim was running as
3344 exim, but did not say which command line option was involved. All I could
3345 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3346 running as the user.
3349 61. There was a buffer overflow vulnerability in the SPA authentication code
3350 (which came originally from the Samba project). I have added a test to the
3351 spa_base64_to_bits() function which I hope fixes it.
3354 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3355 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3357 63. The daemon start-up calls getloadavg() while still root for those OS that
3358 need the first call to be done as root, but it missed one case: when
3359 deliver_queue_load_max is set with deliver_drop_privilege. This is
3360 necessary for the benefit of the queue runner, because there is no re-exec
3361 when deliver_drop_privilege is set.
3363 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3364 This has been fixed.
3366 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3367 was in use, was not putting the data itself into the right store pool;
3368 consequently, it could be overwritten for a subsequent message in the same
3369 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3372 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3374 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3375 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3376 the function were treating the return as a boolean value, which happened to
3377 work because 0=false and not-0=true, but is not correct code.
3379 68. The host_aton() function was not handling scoped IPv6 addresses (those
3380 with, for example, "%eth0" on the end) correctly.
3382 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3383 negated items (that is, ~something) in unsigned ints. Some compilers
3384 apparently mutter when there is no cast.
3386 70. If an address verification called from an ACL failed, and did not produce a
3387 user-specific message (i.e. there was only a "system" message), nothing was
3388 put in $acl_verify_message. In this situation, it now puts the system
3391 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3392 session; change 4.31/43 added the unwanted input to the log line - except
3393 that it did not do this in the start of session case. It now does.
3395 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3396 This is wrong and can cause the other end to generate a synchronization
3397 error if it is another Exim or anything else that does the synchronization
3398 check. A QUIT command is no longer sent after a timeout.
3400 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3401 during host lookups.
3403 74. The defer_ok option of callout verification was not working if it was used
3404 when verifying addresses in header lines, that is, for this case:
3406 verify = header_sender/callout=defer_ok
3408 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3409 those file descriptors could be used for SMTP connections. If anything
3410 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3411 could be sent to the SMTP client, causing chaos. The daemon now opens
3412 stdin, stdout, and stderr to /dev/null when it puts itself into the
3415 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3416 log by default. The user can override this with suitable Perl magic.
3418 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3419 the log message when discard triggers, was not working for the DATA ACL or
3420 for the non-SMTP ACL.
3422 78. Error message wording change in sieve.c.
3424 79. If smtp_accept_max_per_host was set, the number of connections could be
3425 restricted to fewer than expected, because the daemon was trying to set up
3426 a new connection before checking whether the processes handling previous
3427 connections had finished. The check for completed processes is now done
3428 earlier. On busy systems, this bug wouldn't be noticed because something
3429 else would have woken the daemon, and it would have reaped the completed
3432 80. If a message was submitted locally by a user whose login name contained one
3433 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3434 It caused a spool format error. I have fixed the spool reading code. A
3435 related problem was that the "from" clause in the Received: line became
3436 illegal because of the space(s). It is now covered by ${quote_local_part.
3438 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3441 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3443 83. Patch from Sieve maintainer to fix unterminated string problem in
3444 "vacation" handling.
3446 84. Some minor changes to the Linux configuration files to help with other
3447 OS variants using glibc.
3449 85. One more patch for Sieve to update vacation handling to latest spec.
3452 ----------------------------------------------------
3453 See the note above about the 4.44 and 4.50 releases.
3454 ----------------------------------------------------
3460 1. Change 4.43/35 introduced a bug that caused file counts to be
3461 incorrectly computed when quota_filecount was set in an appendfile
3464 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3465 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3468 3. Give more explanation in the error message when the command for a transport
3469 filter fails to execute.
3471 4. There are several places where Exim runs a non-Exim command in a
3472 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3473 was being done only for the command run by the queryprogram router. It is
3474 now done for all such subprocesses. The other cases are: ${run, transport
3475 filters, and the commands run by the lmtp and pipe transports.
3477 5. Some older OS have a limit of 256 on the maximum number of file
3478 descriptors. Exim was using setrlimit() to set 1000 as a large value
3479 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3480 systems. I've change it so that if it can't get 1000, it tries for 256.
3482 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3483 was an oversight, and furthermore, ever since the addition of extra
3484 controls (e.g. 4.43/32), the checks on when to allow different forms of
3485 "control" were broken. There should now be diagnostics for all cases when a
3486 control that does not make sense is encountered.
3488 7. $recipients is now available in the predata ACL (oversight).
3490 8. Tidy the search cache before the fork to do a delivery from a message
3491 received from the command line. Otherwise the child will trigger a lookup
3492 failure and thereby defer the delivery if it tries to use (for example) a
3493 cached ldap connection that the parent has called unbind on.
3495 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3496 of $address_data from the recipient verification was clobbered by the
3497 sender verification.
3499 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3500 was its contents. (It was OK if the option was not defined at all.)
3502 11. A "Completed" log line is now written for messages that are removed from
3503 the spool by the -Mrm option.
3505 12. $host_address is now set to the target address during the checking of
3506 ignore_target_hosts.
3508 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3509 being passed; this would have caused $sender_host_name to have been used if
3510 matching the list had actually called for a host name (not very likely,
3511 since this list is usually IP addresses). A host name is now passed as
3514 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3515 code that specifies a non-restarting handler (typically sigaction() in
3516 modern systems) in an attempt to fix a rare and obscure crash bug.
3518 15. Narrowed the window for a race in the daemon that could cause it to ignore
3519 SIGCHLD signals. This is not a major problem, because they are used only to
3520 wake it up if nothing else does.
3522 16. A malformed maildirsize file could cause Exim to calculate negative values
3523 for the mailbox size or file count. Odd effects could occur as a result.
3524 The maildirsize information is now recalculated if the size or filecount
3527 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3528 support for a long time. Removed HAVE_SYS_VFS_H.
3530 18. Updated exipick to current release from John Jetmore.
3532 19. Allow an empty sender to be matched against a lookup in an address list.
3533 Previously the only cases considered were a regular expression, or an
3536 20. Exim went into a mad DNS lookup loop when doing a callout where the
3537 host was specified on the transport, if the DNS lookup yielded more than
3540 21. The RFC2047 encoding function was originally intended for short strings
3541 such as real names; it was not keeping to the 75-character limit for
3542 encoded words that the RFC imposes. It now respects the limit, and
3543 generates multiple encoded words if necessary. To be on the safe side, I
3544 have increased the buffer size for the ${rfc2047: expansion operator from
3547 22. Failure to deliver a bounce message always caused it to be frozen, even if
3548 there was an errors_to setting on the router. The errors_to setting is now
3551 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3552 canonical form (fully expanded) before being placed in
3553 $sender_host_address.
3555 24. Updated eximstats to version 1.33
3557 25. Include certificate and key file names in error message when GnuTLS fails
3558 to set them up, because the GnuTLS error message doesn't include the name
3559 of the failing file when there is a problem reading it.
3561 26. Expand error message when OpenSSL has problems setting up cert/key files.
3564 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3565 (this can affect the format of dates).
3567 28. exim_tidydb, when checking for the continued existence of a message for
3568 which it has found a message-specific retry record, was not finding
3569 messages that were in split spool directories. Consequently, it was
3570 deleting retry records that should have stayed in existence.
3572 29. eximstats updated to version 1.35
3573 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3574 1.35 - bugfix such that pie charts by volume are generated correctly
3576 30. The SPA authentication driver was not abandoning authentication and moving
3577 on to the next authenticator when an expansion was forced to fail,
3578 contradicting the general specification for all authenticators. Instead it
3579 was generating a temporary error. It now behaves as specified.
3581 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3582 (the order specifies the preference for clients). The order is now AES256,
3583 AES128, 3DES, ARCFOUR128.
3585 31. Small patch to Sieve code - explicitly set From: when generating an
3588 32. Exim crashed if a remote delivery caused a very long error message to be
3589 recorded - for instance if somebody sent an entire SpamAssassin report back
3590 as a large number of 550 error lines. This bug was coincidentally fixed by
3591 increasing the size of one of Exim's internal buffers (big_buffer) that
3592 happened as part of the Exiscan merge. However, to be on the safe side, I
3593 have made the code more robust (and fixed the comments that describe what
3596 33. Some experimental protocols are using DNS PTR records for new purposes. The
3597 keys for these records are domain names, not reversed IP addresses. The
3598 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3599 leaves it alone. Component reversal etc. now happens only for IP addresses.
3602 34. The host_aton() function is supposed to be passed a string that is known
3603 to be a valid IP address. However, in the case of IPv6 addresses, it was
3604 not checking this. This is a hostage to fortune. Exim now panics and dies
3605 if the condition is not met. A case was found where this could be provoked
3606 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3607 components; fortuitously, this particular loophole had already been fixed
3608 by change 4.50/55 or 4.44/33 above.
3610 If there are any other similar loopholes, the new check in host_aton()
3611 itself should stop them being exploited. The report I received stated that
3612 data on the command line could provoke the exploit when Exim was running as
3613 exim, but did not say which command line option was involved. All I could
3614 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3615 running as the user.
3618 35. There was a buffer overflow vulnerability in the SPA authentication code
3619 (which came originally from the Samba project). I have added a test to the
3620 spa_base64_to_bits() function which I hope fixes it.
3623 36. The daemon start-up calls getloadavg() while still root for those OS that
3624 need the first call to be done as root, but it missed one case: when
3625 deliver_queue_load_max is set with deliver_drop_privilege. This is
3626 necessary for the benefit of the queue runner, because there is no re-exec
3627 when deliver_drop_privilege is set.
3629 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3630 was in use, was not putting the data itself into the right store pool;
3631 consequently, it could be overwritten for a subsequent message in the same
3632 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3635 38. Sometimes the final signoff response after QUIT could fail to get
3636 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3637 < 0 before doing a fflush(). This bug looks as though it goes back to the
3638 introduction of TLS in release 3.20, but "sometimes" must have been rare
3639 because the tests only now provoked it.
3645 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3646 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3647 Now it returns an integer. A number of calls were still expecting a T/F
3648 return. Fortuitously, in all cases, the tests worked in OK situations,
3649 which is the norm. However, things would have gone wrong on any write
3650 failures on the smtp file descriptor. This function is used when sending
3651 messages over SMTP and also when doing verify callouts.
3653 2. When Exim is called to do synchronous delivery of a locally submitted
3654 message (the -odf or -odi options), it no longer closes stderr before doing
3657 3. Implemented the mua_wrapper option.
3659 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3661 5. Implemented the functions header_remove(), header_testname(),
3662 header_add_at_position(), and receive_remove_recipient(), and exported them
3665 6. If an ACL "warn" statement specified the addition of headers, Exim already
3666 inserted X-ACL-Warn: at the start if there was no header name. However, it
3667 was not making this test for the second and subsequent header lines if
3668 there were newlines in the string. This meant that an invalid header could
3669 be inserted if Exim was badly configured.
3671 7. Allow an ACL "warn" statement to add header lines at the start or after all
3672 the Received: headers, as well as at the end.
3674 8. Added the rcpt_4xx retry error code.
3676 9. Added postmaster_mailfrom=xxx to callout verification option.
3678 10. Added mailfrom=xxxx to the callout verification option, for verify=
3681 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3682 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3684 12. Inserted some casts to stop certain compilers warning when using pointer
3685 differences as field lengths or precisions in printf-type calls (mostly
3686 affecting debugging statements).
3688 13. Added optional readline() support for -be (dynamically loaded).
3690 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3691 same clock tick as a message's arrival, so that its received time was the
3692 same as the "first fail" time on the retry record, and that message
3693 remained on the queue past the ultimate address timeout, every queue runner
3694 would try a delivery (because it was past the ultimate address timeout) but
3695 after another failure, the ultimate address timeout, which should have then
3696 bounced the address, did not kick in. This was a "< instead of <=" error;
3697 in most cases the first failure would have been in the next clock tick
3698 after the received time, and all would be well.
3700 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3701 being recognized when the domain list was tested by the match_domain
3702 condition in an expansion string.
3704 16. Added the ${str2b64: operator.
3706 17. Exim was always calling setrlimit() to set a large limit for the number of
3707 processes, without checking whether the existing limit was already
3708 adequate. (It did check for the limit on file descriptors.) Furthermore,
3709 errors from getrlimit() and setrlimit() were being ignored. Now they are
3710 logged to the main and panic logs, but Exim does carry on, to try to do its
3711 job under whatever limits there are.
3713 18. Imported PCRE 5.0.
3715 19. Trivial typo in log message " temporarily refused connection" (the leading
3718 20. If the log selector return_path_on_delivery was set and an address was
3719 redirected to /dev/null, the delivery process crashed because it assumed
3720 that a return path would always be set for a "successful" delivery. In this
3721 case, the whole delivery is bypassed as an optimization, and therefore no
3724 21. Internal re-arrangement: the function for sending a challenge and reading
3725 a response while authentication was assuming a zero-terminated challenge
3726 string. It's now changed to take a pointer and a length, to allow for
3727 binary data in such strings.
3729 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3731 23. Exim was not respecting finduser_retries when seeking the login of the
3732 uid under which it was called; it was always trying 10 times. (The default
3733 setting of finduser_retries is zero.) Also, it was sleeping after the final
3734 failure, which is pointless.
3736 24. Implemented tls_on_connect_ports.
3738 25. Implemented acl_smtp_predata.
3740 26. If the domain in control=submission is set empty, Exim assumes that the
3741 authenticated id is a complete email address when it generates From: or
3742 Sender: header lines.
3744 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3745 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3746 chown and chgrp in /bin and hostname in /usr/bin.
3748 28. Exim was keeping the "process log" file open after each use, just as it
3749 does for the main log. This opens the possibility of it remaining open for
3750 long periods when the USR1 signal hits a daemon. Occasional processlog
3751 errors were reported, that could have been caused by this. Anyway, it seems
3752 much more sensible not to leave this file open at all, so that is what now
3755 29. The long-running daemon process does not normally write to the log once it
3756 has entered its main loop, and it closes the log before doing so. This is
3757 so that log files can straightforwardly be renamed and moved. However,
3758 there are a couple of unusual error situations where the daemon does write
3759 log entries, and I had neglected to close the log afterwards.
3761 30. The text of an SMTP error response that was received during a remote
3762 delivery was being truncated at 512 bytes. This is too short for some of
3763 the long messages that one sometimes sees. I've increased the limit to
3766 31. It is now possible to make retry rules that apply only when a message has a
3767 specific sender, in particular, an empty sender.
3769 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3770 it possible to be selective about when SMTP synchronization is enforced.
3772 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3774 32. Implemented hosts_connection_nolog.
3776 33. Added an ACL for QUIT.
3778 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3781 35. Added mailbox_size and mailbox_filecount to appendfile.
3783 36. Added control = no_multiline_responses to ACLs.
3785 37. There was a bug in the logic of the code that waits for the clock to tick
3786 in the case where the clock went backwards by a substantial amount such
3787 that the microsecond fraction of "now" was more than the microsecond
3788 fraction of "then" (but the whole seconds number was less).
3790 38. Added support for the libradius Radius client library this is found on
3791 FreeBSD (previously only the radiusclient library was supported).
3797 1. When certain lookups returned multiple values in the form name=value, the
3798 quoting of the values was not always being done properly. Specifically:
3799 (a) If the value started with a double quote, but contained no whitespace,
3801 (b) If the value contained whitespace other than a space character (i.e.
3802 tabs or newlines or carriage returns) it was not quoted.
3803 This fix has been applied to the mysql and pgsql lookups by writing a
3804 separate quoting function and calling it from the lookup code. The fix
3805 should probably also be applied to nisplus, ibase and oracle lookups, but
3806 since I cannot test any of those, I have not disturbed their existing code.
3808 2. A hit in the callout cache for a specific address caused a log line with no
3809 reason for rejecting RCPT. Now it says "Previous (cached) callout
3810 verification failure".
3812 3. There was an off-by-one bug in the queryprogram router. An over-long
3813 return line was truncated at 256 instead of 255 characters, thereby
3814 overflowing its buffer with the terminating zero. As well as fixing this, I
3815 have increased the buffer size to 1024 (and made a note to document this).
3817 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3818 when Exim is waiting for an SMTP response from a remote server, Exim
3819 restarts its select() call on the socket, thereby resetting its timeout.
3820 This is not a problem when such interrupts are rare. Somebody set up a cron
3821 job to run exiwhat every 2 minutes, which is less than the normal select()
3822 timeout (5 or 10 minutes). This meant that the select() timeout never
3823 kicked in because it was always reset. I have fixed this by comparing the
3824 time when an interrupt arrives with the time at the start of the first call
3825 to select(). If more time than the timeout has elapsed, the interrupt is
3826 treated as a timeout.
3828 5. Some internal re-factoring in preparation for the addition of Sieve
3829 extensions (by MH). In particular, the "personal" test is moved to a
3830 separate function, and given an option for scanning Cc: and Bcc: (which is
3831 not set for Exim filters).
3833 6. When Exim created an email address using the login of the caller as the
3834 local part (e.g. when creating a From: or Sender: header line), it was not
3835 quoting the local part when it contained special characters such as @.
3837 7. Installed new OpenBSD configuration files.
3839 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3840 try to make them clearer.
3842 9. Callout options, other than the timeout value, were being ignored when
3843 verifying sender addresses in header lines. For example, when using
3845 verify = header_sender/callout=no_cache
3847 the cache was (incorrectly) being used.
3849 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3850 only the Cygwin environment.
3852 11. When return_path_on_delivery was set as a log selector, if different remote
3853 addresses in the same message used different return paths and parallel
3854 remote delivery occurred, the wrong values would sometimes be logged.
3855 (Whenever a remote delivery process finished, the return path value from
3856 the most recently started remote delivery process was logged.)
3858 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3859 lines when AUTH and/or TLS are in use. This is the "received protocol"
3860 field. Exim used to use "asmtp" for authenticated SMTP, without any
3861 indication (in the protocol name) for TLS use. Now it follows the RFC and
3862 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3863 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3864 names appear in log lines as well as in Received: header lines.
3866 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3867 extensions, and comparison tests, and to fix some bugs.
3869 14. Changes to the "personal" filter test:
3871 (1) The test was buggy in that it was just doing the equivalent of
3872 "contains" tests on header lines. For example, if a user's address was
3873 anne@some.where, the "personal" test would incorrectly be true for
3875 To: susanne@some.where
3877 This test is now done by extracting each address from the header in turn,
3878 and checking the entire address. Other tests that are part of "personal"
3879 are now done using regular expressions (for example, to check local parts
3880 of addresses in From: header lines).
3882 (2) The list of non-personal local parts in From: addresses has been
3883 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3884 taken from the Sieve specification recommendations.
3886 (3) If the message contains any header line starting with "List-" it is
3887 treated as non-personal.
3889 (4) The test for "circular" in the Subject: header line has been removed
3890 because it now seems ill-conceived.
3892 15. Minor typos in src/EDITME comments corrected.
3894 16. Installed latest exipick from John Jetmore.
3896 17. If headers_add on a router specified a text string that was too long for
3897 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3898 of string_sprintf() is now avoided.
3900 18. $message_body_size was not set (it was always zero) when running the DATA
3901 ACL and the local_scan() function.
3903 19. For the "mail" command in an Exim filter, no default was being set for
3904 the once_repeat time, causing a random time value to be used if "once" was
3905 specified. (If the value happened to be <= 0, no repeat happened.) The
3906 default is now 0s, meaning "never repeat". The "vacation" command was OK
3907 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3908 (I found it when inspecting the code).
3910 20. There is now an overall timeout for performing a callout verification. It
3911 defaults to 4 times the callout timeout, which applies to individual SMTP
3912 commands during the callout. The overall timeout applies when there is more
3913 than one host that can be tried. The timeout is checked before trying the
3914 next host. This prevents very long delays if there are a large number of
3915 hosts and all are timing out (e.g. when the network connections are timing
3916 out). The value of the overall timeout can be changed by specifying an
3917 additional sub-option for "callout", called "maxwait". For example:
3919 verify = sender/callout=5s,maxwait=20s
3921 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3922 to the end before writing, but this should make it even safer).
3924 22. Exim was forgetting that it had advertised PIPELINING for the second and
3925 subsequent messages on an SMTP connection. It was also not resetting its
3926 memory on STARTTLS and an internal HELO.
3928 23. When Exim logs an SMTP synchronization error within a session, it now
3929 records whether PIPELINING has been advertised or not.
3931 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3932 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3933 rather than long int.
3935 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3937 26. Added the never_mail option to autoreply.
3943 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3944 crash if the getsockname() call failed; this can happen if a connection is
3945 closed very soon after it is established. The problem was simply in the
3946 order in which certain operations were done, causing Exim to try to write
3947 to the SMTP stream before it had set up the file descriptor. The bug has
3948 been fixed by making things happen in the correct order.
3954 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3955 before the connection was closed, thus losing the rejection response.
3957 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3958 some early Solaris releases, but causes trouble in current releases where
3959 socklen_t is defined.
3961 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3964 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3967 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3968 that releasing the top of it at the end releases what was used for sub-
3969 expansions (unless the block got too big). However, discard this block if
3970 the first thing is a variable or header, so that we can use its block when
3971 it is dynamic (useful for very large $message_headers, for example).
3973 6. Lookups now cache *every* query, not just the most recent. A new, separate
3974 store pool is used for this. It can be recovered when all lookup caches are
3975 flushed. Lookups now release memory at the end of their result strings.
3976 This has involved some general refactoring of the lookup sources.
3978 7. Some code has been added to the store_xxx() functions to reduce the amount
3979 of flapping under certain conditions.
3981 8. log_incoming_interface used to affect only the <= reception log lines. Now
3982 it causes the local interface and port to be added to several more SMTP log
3983 lines, for example "SMTP connection from", and rejection lines.
3985 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3987 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3989 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3990 could be overwritten at the end of the current message (or the start of a
3991 new message if it was set in a HELO ACL). The value is now preserved for
3992 the duration of the SMTP connection.
3994 12. If a transport had a headers_rewrite setting, and a matching header line
3995 contained an unqualified address, that address was qualified, even if it
3996 did not match any rewriting rules. The underlying bug was that the values
3997 of the flags that permit the existence of unqualified sender and recipient
3998 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3999 non-local messages, and by -bnq for local messages) were not being
4000 preserved with the message after it was received.
4002 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4003 "next input=" as part of the text comprising the host identity instead of
4004 the correct text. The code was using the same buffer for two different
4005 strings. However, depending on which order the printing function evaluated
4006 its arguments, the bug did not always show up. Under Linux, for example, my
4007 test suite worked just fine.
4009 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4010 doesn't work with some older versions of Perl. It has been changed to "my",
4011 which in any case is probably the better facility to use.
4013 15. A really picky compiler found some instances of statements for creating
4014 error messages that either had too many or two few arguments for the format
4017 16. The size of the buffer for calls to the DNS resolver has been increased
4018 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4019 for addresses that have a lot of PTR records. This alleviates a problem; it
4020 does not fully solve it.
4022 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4023 buffer now truncates the list and logs the incident, which is the same
4024 action as happens when Exim is looking up a host name and its aliases.
4025 Previously in this situation something unpredictable would happen;
4026 sometimes it was "internal error: store_reset failed".
4028 18. If a server dropped the connection unexpectedly when an Exim client was
4029 using GnuTLS and trying to read a response, the client delivery process
4030 crashed while trying to generate an error log message.
4032 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4033 string, for example:
4035 warn message = H1: something\nH2: something
4037 the text was added as a single header line from Exim's point of view
4038 though it ended up OK in the delivered message. However, searching for the
4039 second and subsequent header lines using $h_h2: did not work. This has been
4040 fixed. Similarly, if a system filter added multiple headers in this way,
4041 the routers could not see them.
4043 20. Expanded the error message when iplsearch is called with an invalid key to
4044 suggest using net-iplsearch in a host list.
4046 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4047 ACLs are no longer actually imposed (and a message to that effect is
4050 22. If a "gecos" field in a passwd entry contained escaped characters, in
4051 particular, if it contained a \" sequence, Exim got it wrong when building
4052 a From: or a Sender: header from that name. A second bug also caused
4053 incorrect handling when an unquoted " was present following a character
4054 that needed quoting.
4056 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4057 was not being matched caselessly.
4059 24. Arranged for all hyphens in the exim.8 source to be escaped with
4062 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4063 properly. Recipient callout cache records were still being keyed to include
4064 the sender, even when use_sender was set false. This led to far more
4065 callouts that were necessary. The sender is no longer included in the key
4066 when use_sender is false.
4068 26. Added "control = submission" modifier to ACLs.
4070 27. Added the ${base62d: operator to decode base 62 numbers.
4072 28. dnsdb lookups can now access SRV records.
4074 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4075 the configuration file.
4077 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4078 (-v) mode. This makes the output for a verbose queue run more intelligible.
4080 31. Added a use_postmaster feature to recipient callouts.
4082 32. Added the $body_zerocount variable, containing the number of binary zero
4083 bytes in the message body.
4085 33. The time of last modification of the "new" subdirectory is now used as the
4086 "mailbox time last read" when there is a quota error for a maildir
4089 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4091 35. Added +ignore_unknown as a special item in host lists.
4093 36. Code for decoding IPv6 addresses in host lists is now included, even if
4094 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4095 address was recognized as an IP address, but was then not correctly decoded
4096 into binary, causing unexpected and incorrect effects when compared with
4103 1. Very minor rewording of debugging text in manualroute to say "list of
4104 hosts" instead of "hostlist".
4106 2. If verify=header_syntax was set, and a header line with an unqualified
4107 address (no domain) and a large number of spaces between the end of the
4108 name and the colon was received, the reception process suffered a buffer
4109 overflow, and (when I tested it) crashed. This was caused by some obsolete
4110 code that should have been removed. The fix is to remove it!
4112 3. When running in the test harness, delay a bit after writing a bounce
4113 message to get a bit more predictability in the log output.
4115 4. Added a call to search_tidyup() just before forking a reception process. In
4116 theory, someone could use a lookup in the expansion of smtp_accept_max_
4117 per_host which, without the tidyup, could leave open a database connection.
4119 5. Added the variables $recipient_data and $sender_data which get set from a
4120 lookup success in an ACL "recipients" or "senders" condition, or a router
4121 "senders" option, similar to $domain_data and $local_part_data.
4123 6. Moved the writing of debug_print from before to after the "senders" test
4126 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4127 problems for message scanning, either using a data ACL, or using
4128 local_scan() because the Received: header was not generated till after they
4129 were called (in order to set the time as the time of reception completion).
4130 I have revised the way this works. The header is now generated after the
4131 body is received, but before the ACL or local_scan() are called. After they
4132 are run, the timestamp in the header is updated.
4138 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4139 before starting a queue runner without re-exec. This happened only when
4140 deliver_drop_privilege was set or when the Exim user was set to root. The
4141 effect of the bug was that timeouts during subsequent deliveries caused
4142 crashes instead of being properly handled. The handler is now left at its
4143 default (and expected) setting.
4145 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4146 message, again when deliver_drop_privilege is set or Exim is run as root.
4147 The bug described in (1) was not present in this case, but the tidying up
4148 of the other signals was missing. I have made the two cases consistent.
4150 3. The ignore_target_hosts setting on a manualroute router was being ignored
4151 for hosts that were looked up using the /MX notation.
4153 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4156 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4157 operated on the sender address. After changing the $sender_address to <>
4158 for the sender address verify, Exim was re-instated it as the original
4159 (before rewriting) address, but remembering that it had rewritten it, so it
4160 wasn't rewriting it again. This bug also had the effect of breaking the
4161 sender address verification caching when the sender address was rewritten.
4163 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4164 This has been changed so that if the ip literal address matches
4165 ignore_target_hosts, the router declines.
4167 7. Added expansion conditions match_domain, match_address, and match_local_
4168 part (NOT match_host).
4170 8. The placeholder for the Received: header didn't have a length field set.
4172 9. Added code to Exim itself and to exim_lock to test for a specific race
4173 condition that could lead to file corruption when using MBX delivery. The
4174 issue is with the lockfile that is created in /tmp. If this file is removed
4175 after a process has opened it but before that process has acquired a lock,
4176 there is the potential for a second process to recreate the file and also
4177 acquire a lock. This could lead to two Exim processes writing to the file
4178 at the same time. The added code performs the same test as UW imapd; it
4179 checks after acquiring the lock that its file descriptor still refers to
4180 the same named file.
4182 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4183 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4186 11. Added the smtp_active_hostname option. If used, this will typically be made
4187 to depend on the incoming interface address. Because $interface_address is
4188 not set up until the daemon has forked a reception process, error responses
4189 that can happen earlier (such as "too many connections") no longer contain
4192 12. If an expansion in a condition on a "warn" statement fails because a lookup
4193 defers, the "warn" statement is abandoned, and the next ACL statement is
4194 processed. Previously this caused the whole ACL to be aborted.
4196 13. Added the iplsearch lookup type.
4198 14. Added ident_timeout as a log selector.
4200 15. Added tls_certificate_verified as a log selector.
4202 16. Added a global option tls_require_ciphers (compare the smtp transport
4203 option of the same name). This controls incoming TLS connections.
4205 17. I finally figured out how to make tls_require_ciphers do a similar thing
4206 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4207 before starting the TLS session.
4209 18. Tabs are now shown as \t in -bP output.
4211 19. If the log selector return_path_on_delivery was set, Exim crashed when
4212 bouncing a message because it had too many Received: header lines.
4214 20. If two routers both had headers_remove settings, and the first one included
4215 a superfluous trailing colon, the final name in the first list and the
4216 first name in the second list were incorrectly joined into one item (with a
4217 colon in the middle).
4223 1. Added -C and -D options to the exinext utility, mainly to make it easier
4224 to include in the automated testing, but these could be helpful when
4225 multiple configurations are in use.
4227 2. The exinext utility was not formatting the output nicely when there was
4228 an alternate port involved in the retry record key, nor when there was a
4229 message id as well (for retries that were specific to a specific message
4230 and a specific host). It was also confused by IPv6 addresses, because of
4231 the additional colons they contain. I have fixed the IPv4 problem, and
4232 patched it up to do a reasonable job for IPv6.
4234 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4235 delivery, the log line now contains "pipelined" if PIPELINING was used.
4237 4. An SMTP transport process used to panic and die if the bind() call to set
4238 an explicit outgoing interface failed. This has been changed; it is now
4239 treated in the same way as a connect() failure.
4241 5. A reference to $sender_host_name in the part of a conditional expansion
4242 that was being skipped was still causing a DNS lookup. This no longer
4245 6. The def: expansion condition was not recognizing references to header lines
4246 that used bh_ and bheader_.
4248 7. Added the _cache feature to named lists.
4250 8. The code for checking quota_filecount in the appendfile transport was
4251 allowing one more file than it should have been.
4253 9. For compatibility with Sendmail, the command line option
4261 and sets the incoming protocol and host name (for trusted callers). The
4262 host name and its colon can be omitted when only the protocol is to be set.
4263 Note the Exim already has two private options, -pd and -ps, that refer to
4264 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4265 "s", but I don't think that's a major issue.
4267 10. A number of refactoring changes to the code, none of which should affect
4270 (a) The number of logging options was getting close to filling up the
4271 32-bit word that was used as a bit map. I have split them into two classes:
4272 those that are passed in the argument to log_write(), and those that are
4273 only ever tested independently outside of that function. These are now in
4274 separate 32-bit words, so there is plenty of room for expansion again.
4275 There is no change in the user interface or the logging behaviour.
4277 (b) When building, for example, log lines, the code previously used a
4278 macro that called string_cat() twice, in order to add two strings. This is
4279 not really sufficiently general. Furthermore, there was one instance where
4280 it was actually wrong because one of the argument was used twice, and in
4281 one call a function was used. (As it happened, calling the function twice
4282 did not affect the overall behaviour.) The macro has been replaced by a
4283 function that can join an arbitrary number of extra strings onto a growing
4286 (c) The code for expansion conditions now uses a table and a binary chop
4287 instead of a serial search (which was left over from when there were very
4288 few conditions). Also, it now recognizes conditions like "pam" even when
4289 the relevant support is not compiled in: a suitably worded error message is
4290 given if an attempt is made to use such a condition.
4292 11. Added ${time_interval:xxxxx}.
4294 12. A bug was causing one of the ddress fields not to be passed back correctly
4295 from remote delivery subprocesses. The field in question was not being
4296 subsequently used, so this caused to problems in practice.
4298 13. Added new log selectors queue_time and deliver_time.
4300 14. Might have fixed a bug in maildirsizefile handling that threw up
4301 "unexpected character" debug warnings, and recalculated the data
4302 unnecessarily. In any case, I expanded the warning message to give more
4305 15. Added the message "Restricted characters in address" to the statements in
4306 the default ACL that block characters like @ and % in local parts.
4308 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4309 Three changes have been made:
4311 (a) There was a serious bug; a negative response to MAIL caused the whole
4312 recipient domain to be cached as invalid, thereby blocking all messages
4313 to all local parts at the same domain, from all senders. This bug has
4314 been fixed. The domain is no longer cached after a negative response to
4315 MAIL if the sender used is not empty.
4317 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4320 (c) A new callout option, "use_sender" has been added for people who want
4321 the modified behaviour.
4327 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4330 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4331 indeed breaks things for older releases.
4333 3. Added additional logging to the case where there is a problem reading data
4334 from a filter that is running in a subprocess using a pipe, in order to
4335 try to track down a specific problem.
4337 4. Testing facility fudge: when running in the test harness and attempting
4338 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4339 getting "No route to host". Convert this to a timeout.
4341 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4344 6. Some OS don't have socklen_t but use size_t instead. This affects the
4345 fifth argument of getsockopt() amongst other things. This is now
4346 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4347 can be set for individual OS. I have set it for SunOS5, OSF1, and
4348 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4349 some earlier ones do not.
4351 7. Change 4.30/15 was not doing the test caselessly.
4353 8. The standard form for an IPv6 address literal was being rejected by address
4354 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4355 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4356 this, as well as the form without the "IPv6" on the front (but only when
4357 address literals are enabled, of course).
4359 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4361 10. Exim crashed if a message with an empty sender address specified by -f
4362 encountered a router with an errors_to setting. This could be provoked only
4363 by a command such as
4367 where an empty string was supplied; "<>" did not hit this bug.
4369 11. Installed PCRE release 4.5.
4371 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4372 remained set. It is now erased.
4374 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4375 times from message ids (which are base 36 rather than the normal 62).
4377 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4378 were being counted as actual protocol errors, and logged if the log
4379 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4380 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4381 and DATA following a set of rejected RCPTs do not count as protocol errors.
4382 In other words, Exim assumes they were pipelined, though this may not
4383 actually be the case. Of course, in all cases the client gets an
4384 appropriate error code.
4386 15. If a lookup fails in an ACL condition, a message about the failure may
4387 be available; it is used if testing the ACL cannot continue, because most
4388 such messages specify what the cause of the deferral is. However, some
4389 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4390 that caused an old message to be retained and used if a later statement
4391 caused a defer, replacing the real cause of the deferral.
4393 16. If an IP address had so many PTR records that the DNS lookup buffer
4394 was not large enough to hold them, Exim could crash while trying to process
4395 the truncated data. It now detects and logs this case.
4397 17. Further to 4.21/58, another change has been made: if (and only if) the
4398 first line of a message (the first header line) ends with CRLF, a bare LF
4399 in a subsequent header line has a space inserted after it, so as not to
4400 terminate the header.
4402 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4403 unnecessarily, used atoi() instead of strtol(), and didn't check the
4404 termination when getting file sizes from file names by regex.
4406 19. Completely re-implemented the support for maildirsize files, in the light
4407 of a number of problems with the previous contributed implementation
4408 (4.30/29). In particular:
4410 . If the quota is zero, the maildirsize file is maintained, but no quota is
4413 . If the maildir directory does not exist, it is created before any attempt
4414 to write a maildirsize file.
4416 . The quota value in the file is just a cache; if the quota is changed in
4417 the transport, the new value overrides.
4419 . A regular expression is available for excluding directories from the
4422 20. The autoreply transport checks the characters in options that define the
4423 message's headers; it allows continued headers, but it was checking with
4424 isspace() after an embedded newline instead of explicitly looking for a
4427 21. If all the "regular" hosts to which an address was routed had passed their
4428 expiry times, and had not reached their retry times, the address was
4429 bounced, even if fallback hosts were defined. Now Exim should go on to try
4432 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4433 equivalent code in the SMTP transport. Some hosts send humungous responses
4434 to HELO/EHLO, more than 1024 it seems.
4436 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4437 gives compiler warnings in some environments. I've now done it "properly",
4440 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4441 (because of problems with the built-in one) was declared to return uschar *
4442 instead of char *, causing compiler failure.
4444 25. Fixed a file descriptor leak when processing alias/forward files.
4446 26. Fixed a minor format string issue in dbfn.c.
4448 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4450 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4451 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4454 29. When a sender address is verified, it is cached, to save repeating the test
4455 when there is more than one recipient in a message. However, when the
4456 verification involves a callout, it is possible for different callout
4457 options to be set for different recipients. It is too complicated to keep
4458 track of this in the cache, so now Exim always runs a verification when a
4459 callout is required, relying on the callout cache for the optimization.
4460 The overhead is duplication of the address routing, but this should not be
4463 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4464 to be verified with callout=postmaster, and the main callout worked but the
4465 postmaster check failed, the verification correctly failed. However, if a
4466 subsequent RCPT command asked for sender verification *without* the
4467 postmaster check, incorrect caching caused this verification also to fail,
4470 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4471 it was not caching the DNS options (qualify_single, search_parents) that
4472 were used when the lookup failed. A subsequent lookup with different
4473 options therefore always gave the same answer, though there were cases
4474 where it should not have. (Example: a "domains = !$mx_any" option on a
4475 dnslookup router: the "domains" option is always processed without any
4476 widening, but the router might have qualify_single set.) Now Exim uses the
4477 cached value only when the same options are set.
4479 32. Added John Jetmore's "exipick" utility to the distribution.
4481 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4482 than a timeout (e.g. a certificate is required, and is not provided), an
4483 Exim server now closes the connection immediately. Previously it waited for
4484 the client to close - but if the client is SSL, it seems that they each
4485 wait for each other, leading to a delay before one of them times out.
4487 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4488 maintained 0.8.x compatibility because I don't think many are using it, and
4489 it is clearly obsolete.
4491 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4494 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4495 client certificate was expired. A simple patch fixes this, though I don't
4496 understand the full logic of why the verify callback is called multiple
4499 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4500 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4501 which causes problems with some clients (such as the Certicom SSL Plus
4502 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4503 disables the coutermeasure allowing Eudora to connect."
4505 38. Exim was not checking that a write() to a log file succeeded. This could
4506 lead to Bad Things if a log got too big, in particular if it hit a file
4507 size limit. Exim now panics and dies if it cannot write to a log file, just
4508 as it does if it cannot open a log file.
4510 39. Modified OS/Makefile-Linux so that it now contains
4512 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4514 The two -D definitions ensure that Exim is compiled with large file
4515 support, which makes it possible to handle log files that are bigger than
4518 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4519 instance) a domain was checked against a named list that involved a lookup,
4520 causing $domain_data to be set, then another domain was checked against the
4521 same list, then the first domain was re-checked, the value of $domain_data
4522 after the final check could be wrong. In particular, if the second check
4523 failed, it could be set empty. This bug probably also applied to
4526 41. The strip_trailing_dot option was not being applied to the address given
4527 with the -f command-line option.
4529 42. The code for reading a message's header from the spool was incrementing
4530 $received_count, but never initializing it. This meant that the value was
4531 incorrect (doubled) while delivering a message in the same process in which
4532 it was received. In the most common configuration of Exim, this never
4533 happens - a fresh exec is done - but it can happen when
4534 deliver_drop_privilege is set.
4536 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4537 it now includes up to 150 characters of the unexpected data in the log
4540 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4541 and building data strings. The size of both of these buffers was 10 000
4542 bytes - far larger than anybody would *ever* want, thought I. Needless to
4543 say, somebody hit the limit. I have increased the maximum line length to
4544 20 000 and the maximum data length of concatenated lines to 100 000. I have
4545 also fixed two bugs, because there was no checking on these buffers. Tsk,
4546 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4547 buffer is too small.
4549 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4550 lsearch lookups. Now it does.
4552 46. When parsing a route_list item in a manualroute router, a fixed-length
4553 buffer was used for the list of hosts. I made this 1024 bytes long,
4554 thinking that nobody would ever have a list of hosts that long. Wrong.
4555 Somebody had a whole pile of complicated expansion conditions, and the
4556 string was silently truncated, leading to an expansion error. It turns out
4557 that it is easier to change to an unlimited length (owing to other changes
4558 that have happened since this code was originally written) than to build
4559 structure for giving a limitation error. The length of the item that
4560 expands into the list of hosts is now unlimited.
4562 47. The lsearch lookup could not handle data where the length of text line was
4563 more than 4095 characters. Such lines were truncated, leading to shortened
4564 data being returned. It should now handle lines of any length.
4566 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4567 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4570 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4571 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4572 no longer happens in comment lines. A list of replacements is now placed
4573 at the head of all of the source files, except those whose only change is
4574 to replace PERL_COMMAND in the very first #! line.
4576 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4577 messages on the queue, with a bottom-up merge sort, using code contributed
4578 by Michael Haardt. This should make operations like -bp somewhat faster on
4579 large queues. It won't affect queue runners, except when queue_run_in_order
4582 51. Installed eximstats 1.31 in the distribution.
4584 52. Added support for SRV lookups to the dnslookup router.
4586 53. If an ACL referred to $message_body or $message_body_end, the value was not
4587 reset for any messages that followed in the same SMTP session.
4589 54. The store-handling optimization for building very long strings was not
4590 differentiating between the different store pools. I don't think this
4591 actually made any difference in practice, but I've tidied it.
4593 55. While running the routers to verify a sender address, $sender_address
4594 was still set to the sender address. This is wrong, because when routing to
4595 send a bounce to the sender, it would be empty. Therefore, I have changed
4596 it so that, while verifying a sender address, $sender_address is set to <>.
4597 (There is no change to what happens when verifying a recipient address.)
4599 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4600 target A or AAAA records (if not already returned) without resetting the
4601 qualify_single or search_parents options of the DNS resolver. These are
4602 inappropriate in this case because the targets of MX and SRV records must
4603 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4604 target that, when qualified, matched something in the local domain. These
4605 two options are now turned off when doing these lookups.
4607 57. It seems that at least some releases of Reiserfs (which does not have the
4608 concept of a fixed number of inodes) returns zero and not -1 for the
4609 number of available inodes. This interacted badly with check_spool_inodes,
4610 which assumed that -1 was the "no such thing" setting. What I have done is
4611 to check that the total number of inodes is greater than zero before doing
4612 the test of how many are available.
4614 58. When a "warn" ACL statement has a log_message modifier, the message is
4615 remembered, and not repeated. This is to avoid a lot of repetition when a
4616 message has many recipients that cause the same warning to be written.
4617 Howewer, Exim was preserving the list of already written lines for an
4618 entire SMTP session, which doesn't seem right. The memory is now reset if a
4619 new message is started.
4621 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4622 debugging output unless log_rewrite was also set.
4624 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4625 of (int)(handle) when we know that handle contains (void *)(-1).
4627 61. The Exim daemon panic-logs an error return when it closes the incoming
4628 connection. However "connection reset by peer" seems to be common, and
4629 isn't really an error worthy of noting specially, so that particular error
4632 62. When Exim is trying to find all the local interfaces, it used to panic and
4633 die if the ioctl to get the interface flags failed. However, it seems that
4634 on at least one OS (Solaris 9) it is possible to have an interface that is
4635 included in the list of interfaces, but for which you get a failure error
4636 for this call. This happens when the interface is not "plumbed" into a
4637 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4638 failure of the "get flags" call assumes that the interface is down.
4640 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4641 makes life easier for people who are doing arithmetic on fields extracted
4642 from dates, where you often get leading zeros that should not be
4643 interpreted as octal.
4645 64. Added qualify_domain to the redirect router, to override the global
4648 65. If a pathologically long header line contained very many addresses (the
4649 report of this problem mentioned 10 000) and each of them was rewritten,
4650 Exim could use up a very large amount of memory. (It kept on making new
4651 copies of the header line as it rewrote, and never released the old ones.)
4652 At the expense of a bit more processing, the header rewriting function has
4653 been changed so that it no longer eats memory in this way.
4655 66. The generation of the Received: header has been moved from the time that a
4656 message starts to be received, to the time that it finishes. The timestamp
4657 in the Received: header should now be very close to that of the <= log
4658 line. There are two side-effects of this change:
4660 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4661 logged header lines no longer include the local Received: line, because
4662 it has not yet been created. The same applies to a copy of the message
4663 that is returned to a non-SMTP sender when a message is rejected.
4665 (b) When a filter file is tested using -bf, no additional Received: header
4666 is added to the test message. After some thought, I decided that this
4669 This change does not affect the value of $received_for. It is still set
4670 after address rewriting, but before local_scan() is called.
4672 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4674 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4675 gave an unhelpful panic error message, and a defer error. I have managed to
4676 change this behaviour so that it now rejects any supplied certificate,
4677 which seems right, as the list of acceptable certificates is empty.
4679 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4680 gave an unhelpful defer error. I have not managed to make this reject any
4681 supplied certificates, but the error message it gives is "no certificate
4682 supplied", which is not helpful.
4684 70. exigrep's output now also includes lines that are not associated with any
4685 message, but which match the given pattern. Implemented by a patch from
4686 Martin Sluka, which also tidied up the Perl a bit.
4688 71. Recipient callout verification, like sender verification, was using <> in
4689 the MAIL FROM command. This isn't really the right thing, since the actual
4690 sender may affect whether the remote host accepts the recipient or not. I
4691 have changed it to use the actual sender in the callout; this means that
4692 the cache record is now keyed on a recipient/sender pair, not just the
4693 recipient address. There doesn't seem to be a real danger of callout loops,
4694 since a callout by the remote host to check the sender would use <>.
4695 [SEE ABOVE: changed after hitting problems.]
4697 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4698 temporary errors. However, in the case of such a code being given after
4699 the end of a data transmission (i.e. after ".") Exim was failing to write
4700 a retry record for the message. (Yes, there was some broken host that was
4701 actually sending 8xx at this point.)
4703 73. An unknown lookup type in a host list could cause Exim to panic-die when
4704 the list was checked. (An example that provoked this was putting <; in the
4705 middle of a list instead of at the start.) If this happened during a DATA
4706 ACL check, a -D file could be left lying around. This kind of configuration
4707 error no longer causes Exim to die; instead it causes a defer errror. The
4708 incident is still logged to the main and panic logs.
4710 74. Buglet left over from Exim 3 conversion. The message "too many messages
4711 in one connection" was written to the rejectlog but not the mainlog, except
4712 when address rewriting (yes!) was being logged.
4714 75. Added write_rejectlog option.
4716 76. When a system filter was run not as root (that is, when system_filter_user
4717 was set), the values of the $n variables were not being returned to the
4718 main process; thus, they were not subsequently available in the $sn
4721 77. Added +return_path_on_delivery log selector.
4723 78. A connection timeout was being treated differently from recipients deferred
4724 when testing hosts_max_try with a message that was older than the host's
4725 retry timeout. (The host should not be counted, thus allowing all hosts to
4726 be tried at least once before bouncing.) This may have been the cause of an
4727 occasionally reported bug whereby a message would remain on the queue
4728 longer than the retry timeout, but would be bounced if a delivery was
4729 forced. I say "may" because I never totally pinned down the problem;
4730 setting up timeout/retry tests is difficult. See also the next item.
4732 79. The ultimate address timeout was not being applied to errors that involved
4733 a combination of host plus message (for example, a timeout on a MAIL
4734 command). When an address resolved to a number of possible hosts, and they
4735 were not all tried for each delivery (e.g. because of hosts_max_try), a
4736 message could remain on the queue longer than the retry timeout.
4738 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4741 81. Fixed an obscure SMTP outgoing bug which required at least the following
4742 conditions: (a) there was another message waiting for the same server;
4743 (b) the server returned 5xx to all RCPT commands in the first message so
4744 that the message was not completed; (c) the server dropped the connection
4745 or gave a negative response to the RSET that Exim sends to abort the
4746 transaction. The observed case was a dropped connection after DATA that had
4747 been sent in pipelining mode. That is, the server had advertised PIPELINING
4748 but was not implementing it correctly. The effect of the bug was incorrect
4749 behaviour, such as trying another host, and this could lead to a crash.
4755 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4756 and daemon.c were passed as pointers to ints; they should have been
4757 pointers to socklen_t variables (which are typically unsigned ints).
4759 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4762 3. Fixed a really odd bug that affected only the testing scheme; patching a
4763 certain fixed string in the binary changed the value of another string that
4764 happened to be identical to the end of the original first string.
4766 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4767 name", it returns that address as the IP address. On some operating
4768 systems (e.g. Solaris), it also passes back the IP address string as the
4769 "host name". However, on others (e.g. Linux), it passes back an empty
4770 string. Exim wasn't checking for this, and was changing the host name to an
4771 empty string, assuming it had been canonicized.
4773 5. Although rare, it is permitted to have more than one PTR record for a given
4774 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4775 all the names associated with an address, because they do in Solaris.
4776 However, it seems that they do not in Linux for data that comes from the
4777 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4778 I found this out when I moved to a new Linux workstation and tried to run
4779 the Exim test suite.
4781 To get round this problem I have changed the code so that it now does its
4782 own call to the DNS to look up PTR records when searching for a host name.
4783 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4784 addresses that are only in /etc/hosts are still found.
4786 This behaviour is, however, controlled by an option called host_lookup_
4787 order, which defaults to "bydns:byaddr". If people want to use the other
4788 order, or indeed, just use one or the other means of lookup, they can
4789 specify it in this variable.
4791 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4792 some operating systems, this comes back from gethostbyaddr() as an empty
4793 string, and this is what Exim used to test for. However, it seems that in
4794 other systems, "." is yielded. Exim now tests for this case too.
4796 7. The values of check_spool_space and check_log_space are now held internally
4797 as a number of kilobytes instead of an absolute number of bytes. If a
4798 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4799 kilobyte. This means that much larger values can be stored.
4801 8. Exim monitor: an attempt to get the action menu when not actually pointing
4802 at a message produces an empty menu entitled "No message selected". This
4803 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4804 no entries in it ("Shell widget menu has zero width and/or height"). So I
4805 have added a single, blank menu entry in this case.
4807 9. Added ${quote_local_part.
4809 10. MIME decoding is now applied to the contents of Subject: header lines when
4812 11. Now that a reference to $sender_host_address automatically causes a reverse
4813 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4814 host lookup before query-style lookups in lists that might use this
4815 variable. This has therefore been abolished, and the "net-" prefix is no
4816 longer necessary for query-style lookups.
4818 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4819 have been a typo for LFLAGS, so it has been changed.
4821 13. The install script calls Exim with "-C /dev/null" in order to find the
4822 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4823 to be output. Howeve, since Exim outputs its version number before the
4824 error, it didn't break the script. It just looked ugly. I fixed this by
4825 always allowing "-C /dev/null" if the caller is root.
4827 14. Ignore overlarge ACL variable number when reading spool file - insurance
4828 against a later release with more variables having written the file.
4830 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4831 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4832 this, as well as the form without the "IPv6" on the front.
4834 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4835 OS/Makefile-Darwin file.
4837 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4838 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4840 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4841 variable. However, some OS use ints for inodes. I've added cast to long int
4842 to get rid of the compiler warning.
4844 19. I had forgotten to lock out "/../" in configuration file names when
4845 ALT_CONFIG_PREFIX was set.
4847 20. Routers used for verification do not need to specify transports. However,
4848 if such a router generated a host list, and callout was configured, Exim
4849 crashed, because it could not find a port number from the (non-existent)
4850 transport. It now assumes port 25 in this circumstance.
4852 21. Added the -t option to exigrep.
4854 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4855 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4856 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4857 them is set, LOOKUP_LSEARCH is forced.
4859 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4861 24. Added sender and host information to the "rejected by local_scan()" log
4862 line; previously there was no indication of these.
4864 25. Added .include_if_exists.
4866 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4867 Linux. It turns out that not all file systems support this. Apparently some
4868 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4869 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4870 not supported on the file descriptor, is now ignored when Exim is trying to
4871 sync a directory. This applies only to Linux.
4873 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4875 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4876 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4879 29. Added support for maildirsize files from supplied patch (modified a bit).
4881 30. The use of :fail: followed by an empty string could lead Exim to respond to
4882 sender verification failures with (e.g.):
4884 550 Verification failed for <xxx>
4885 550 Sender verify failed
4887 where the first response line was missing the '-' that indicates it is not
4888 the final line of the response.
4890 31. The loop for finding the name of the user that called Exim had a hardwired
4891 limit of 10; it now uses the value of finduser_retries, which is used for
4892 all other user lookups.
4894 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4897 33. Exim was neglecting to zero errno before one call of strtol() when
4898 expanding a string and expecting an integer value. On some systems this
4899 resulted in spurious "integer overflow" errors. Also, it was casting the
4900 result into an int without checking.
4902 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4903 did not work. The code looks as if it has *never* worked, though it appears
4904 to have been documented since at least releast 1.62. I have made it work.
4906 35. The "timeout_DNS" error in retry rules, also documented since at least
4907 1.62, also never worked. As it isn't clear exactly what this means, and
4908 clearly it isn't a major issue, I have abolished the feature by treating it
4909 as "timeout", and writing a warning to the main and panic logs.
4911 36. The display of retry rules for -brt wasn't always showing the error code
4914 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4915 timeout_connect_A, timeout_connect_MX.
4917 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4918 to the empty sender.
4920 39. The daemon was not analysing the content of -oX till after it had closed
4921 stderr and disconnected from the controlling terminal. This meant that any
4922 syntax errors were only noted on the panic log, and the return code from
4923 the command was 0. By re-arranging the code a little, I've made the
4924 decoding happen first, so such errors now appear on stderr, and the return
4925 code is 1. However, the actual setting up of the sockets still happens in
4926 the disconnected process, so errors there are still only recorded on the
4929 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4930 connections (as happens on some IP stacks) was logged at start up time as
4931 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4932 it from "IPv6 and IPv4", which means that two separate sockets are being
4935 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4936 says whether AF_INET or AF_INET6 was passed as an argument.
4938 42. Exiwhat output was messed up when time zones were included in log
4941 43. Exiwhat now gives more information about the daemon's listening ports,
4942 and whether -tls-on-connect was used.
4944 44. The "port" option of the smtp transport is now expanded.
4946 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4947 silently ignored. Now an error message is written to the main and panic
4950 46. There's a new ACL modifier called "logwrite" which writes to a log file
4951 as soon as it is encountered.
4953 47. Added $local_user_uid and $local_user_gid at routing time.
4955 48. Exim crashed when trying to verify a sender address that was being
4958 49. Exim was recognizing only a space character after ".include". It now also
4959 recognizes a tab character.
4961 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4962 extracting the relevant information from the specification. The man page no
4963 longer contains scrambled data for the -d option, and I've added a section
4964 at the front about calling Exim under different names.
4966 51. Added "extra_headers" argument to the "mail" command in filter files.
4968 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4971 53. Installed eximstats 1.29.
4973 54. Added transport_filter_timeout as a generic transport option.
4975 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4976 Cc: header lines. This was required by RFC 822, but it not required by RFC
4979 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4980 incoming messages that did not have them. Now it does so only if the
4981 message originates locally, that is, if there is no associated remote host
4982 address. When Resent- header lines are present, this applies to the Resent-
4983 lines rather than the non-Resent- lines.
4985 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4986 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4988 58. Messages for configuration errors now include the name of the main
4989 configuration file - useful now that there may be more than one file in a
4990 list (.included file names were always shown).
4992 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4993 for those rare installations that do not start the daemon as root or run it
4994 setuid root. I've cut out the call to initgroups() if the daemon is not
4997 60. The Exim user and group can now be bound into the binary as text strings
4998 that are looked up at the start of Exim's processing.
5000 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5002 62. Added $mailstore_basename variable.
5004 63. Installed patch to sieve.c from Michael Haardt.
5006 64. When Exim failed to open the panic log after failing to open the main log,
5007 the original message it was trying to log was written to stderr and debug
5008 output, but if they were not available (the usual case in production), it
5009 was lost. Now it is written to syslog before the two lines that record the
5010 failures to open the logs.
5012 65. Users' Exim filters run in subprocesses under the user's uid. It is
5013 possible for a "deliver" command or an alias in a "personal" command to
5014 provoke an address rewrite. If logging of address rewriting is configured,
5015 this fails because the process is not running as root or exim. There may be
5016 a better way of dealing with this, but for the moment (because 4.30 needs
5017 to be released), I have disabled address rewrite logging when running a
5018 filter in a non-root, non-exim process.
5024 1. The buildconfig auxiliary program wasn't quoting the value set for
5025 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5026 not defined. This bug was masked in 4.22 by the effect that was fixed in
5029 2. Some messages that were rejected after a message id was allocated were
5030 shown as "incomplete" by exigrep. It no longer does this for messages that
5031 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5033 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5034 have allow_domain_literals set, the ID did not get logged in the <= line.
5035 Domain literals are now always recognized in Message-ID: header lines.
5037 4. The first argument for a ${extract expansion item is the key name or field
5038 number. Leading and trailing spaces in this item were not being ignored,
5039 causing some misleading effects.
5041 5. When deliver_drop_privilege was set, single queue runner processes started
5042 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5043 same command in the process it spins off) were not dropping privilege.
5045 6. When the daemon running as "exim" started a queue runner, it always
5046 re-executed Exim in the spun-off process. This is a waste of effort when
5047 deliver_drop_privilege is set. The new process now just calls the
5048 queue-runner function directly.
5054 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5057 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5058 ignored. Though the use of -oP was forcing the writing of a pid file, it
5059 was always written to the default place.
5061 3. If the message "no IP address found for host xxxx" is generated during
5062 incoming verification, it is now followed by identification of the incoming
5063 connection (so you can more easily find what provoked it).
5065 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5067 5. Added some features to "harden" Exim a bit more against certain attacks:
5069 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5070 be put in Local/Makefile. This is like the never_users runtime option,
5071 but it cannot be overridden. The default setting is "root".
5073 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5074 prefix string with which any file named in a -C command line option
5077 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5078 is retained for -C and -D only if the caller of Exim is root. Without
5079 it, the exim user may also use -C and -D and retain privilege.
5081 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5082 command line option is disabled.
5084 6. Macro names set by the -D option must start with an upper case letter, just
5085 like macro names defined in the configuration file.
5087 7. Added "dereference=" facility to LDAP.
5089 8. Two instances of the typo "uknown" in the source files are fixed.
5091 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5092 the Configure-Makefile script screwed up while processing it.
5094 10. Incorporated PCRE 4.4.
5096 11. The SMTP synchronization check was not operating right at the start of an
5097 SMTP session. For example, it could not catch a HELO sent before the client
5098 waited for the greeting. There is now a check for outstanding input at the
5099 point when the greeting is written. Because of the duplex, asynchronous
5100 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5101 way, but not yet received, when the check is performed.
5103 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5104 on TCP/IP sockets, because this apparently causes some broken clients to
5107 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5108 unchanged) from the Cygwin maintainer.
5110 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5111 was supported instead of testing for mbx. Effectively a typo.
5113 15. The spa authenticator server code was not checking that the input it
5114 received was valid base64.
5116 16. The debug output line for the "set" modifier in ACLs was not showing the
5117 name of the variable that was being set.
5119 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5121 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5122 lookup on its own. Something else was needed to trigger the lookup. For
5123 example, a match in host_lookup or the need for a host name in a host list.
5124 Now, if $sender_host_name is referenced and the host name has not yet been
5125 looked up, a lookup is performed. If the lookup fails, the variable remains
5126 empty, and $host_lookup_failed is set to "1".
5128 19. Added "eqi" as a case-independent comparison operator.
5130 20. The saslauthd authentication condition could segfault if neither service
5131 nor realm was specified.
5133 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5134 error message that was logged was misleading, and incoming SMTP
5135 connections were dropped. The message is now more accurate, and temporary
5136 errors are given to SMTP connections.
5138 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5139 (except RSET) with a 421 error, until QUIT is received. However, it was
5140 failing to send a response to QUIT.
5142 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5143 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5144 ACL is now run after the helo_try_verify_hosts code.
5146 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5147 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5148 case-independent, so other case variants are also recognized). Apparently
5149 some systems use these upper case variants.
5151 25. If more than two messages were waiting for the same host, and a transport
5152 filter was specified for the transport, Exim sent two messages over the
5153 same TCP/IP connection, and then failed with "socket operation on non-
5154 socket" when it tried to send the third.
5156 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5158 27. The extern definition of crypt16() in expand.c was not being excluded when
5159 the OS had its own crypt16() function.
5161 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5162 as a preferred synonym for return_size_limit, both as an option and as an
5165 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5167 30. Changed the default configuration ACL to relax the local part checking rule
5168 for addresses that are not in any local domains. For these addresses,
5169 slashes and pipe symbols are allowed within local parts, but the sequence
5170 /../ is explicitly forbidden.
5172 31. SPA server authentication was not clearing the challenge buffer before
5175 32. log_message in a "warn" ACL statement was writing to the reject log as
5176 well as to the main log, which contradicts the documentation and doesn't
5177 seem right (because no rejection is happening). So I have stopped it.
5179 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5180 However, I am unable to do any testing of this.
5182 34. Fixed an infelicity in the appendfile transport. When checking directories
5183 for a mailbox, to see if any needed to be created, it was accidentally
5184 using path names with one or more superfluous leading slashes; tracing
5185 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5187 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5188 contents are added to the log line that is written for every discarded
5189 recipient. (Previously a log_message setting was ignored.)
5191 36. The ${quote: operator now quotes the string if it is empty.
5193 37. The install script runs exim in order to find its version number. If for
5194 some reason other than non-existence or emptiness, which it checks, it
5195 could not run './exim', it was installing it with an empty version number,
5196 i.e. as "exim-". This error state is now caught, and the installation is
5199 38. An argument was missing from the function that creates an error message
5200 when Exim fails to connect to the socket for saslauthd authentication.
5201 This could cause Exim to crash, or give a corrupted message.
5203 39. Added isip, isip4, and isip6 to ${if conditions.
5205 40. The ACL variables $acl_xx are now saved with the message, and can be
5206 accessed later in routers, transports, and filters.
5208 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5209 strings in the file are not string-expanded.
5211 42. If a MAIL command specified a SIZE value that was too large to fit into an
5212 int variable, the check against message_size_limit failed. Such values are
5213 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5214 day this will have to be increased, but I don't think I want to be around
5215 when emails are that large.
5222 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5223 iconv() is not standard in FreeBSD.
5225 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5226 IPv6 enabled. The observed symptom was a segmentation fault on return from
5227 the function os_common_find_running_interfaces() in src/os.c.
5229 3. In the check_special_case() function in daemon.c I had used "errno" as an
5230 argument name, which causes warnings on some systems. This was basically a
5231 typo, since it was named "eno" in the comments!
5233 4. The code that waits for the clock to tick (at a resolution of some fraction
5234 of a second) so as to ensure message-id uniqueness was always waiting for
5235 at least one whole tick, when it could have waited for less. [This is
5236 almost certainly not relevant at current processor speeds, where it is
5237 unlikely to ever wait at all. But we try to future-proof.]
5239 5. The function that sleeps for a time interval that includes fractions of a
5240 second contained a race. It did not block SIGALRM between setting the
5241 timer, and suspending (a couple of lines later). If the interval was short
5242 and the sigsuspend() was delayed until after it had expired, the suspension
5243 never ended. On busy systems this could lead to processes getting stuck for
5246 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5247 process, before it forks any delivery processes. The open lookup caching
5248 mechanism meant that the open file or database connection was passed into
5249 the delivery process. The problem was that delivery processes always tidy
5250 up cached lookup data. This could cause a problem for the next delivery
5251 process started by the queue runner, because the external queue runner
5252 process does not know about the closure. So the next delivery process
5253 still has data in the lookup cache. In the case of a file lookup, there was
5254 no problem because closing a file descriptor in a subprocess doesn't affect
5255 the parent. However, if the lookup was caching a connection to a database,
5256 the connection was closed, and the second delivery process was likely to
5257 see errors such as "PGSQL: query failed: server closed the connection
5258 unexpectedly". The problem has been fixed by closing all cached lookups
5259 in a queue runner before running a delivery process.
5261 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5262 seem to have the "const" qualifier which it has on other OS. I've
5265 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5266 *of the same type* (client or server) with the same public name that an
5267 error should be diagnosed.
5269 9. When Exim looked up a host name for an IP address, but failed to find the
5270 original IP address when looking up the host name (a safety check), it
5271 output the message "<ip address> does not match any IP for NULL", which was
5272 confusing, to say the least. The bug was that the host name should have
5273 appeared instead of "NULL".
5275 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5276 user that is built into the binary, and the -C or -D options is used, root
5277 privilege is dropped before the configuration file is read. In addition,
5278 logging is switched to stderr instead of the normal log files. If the
5279 configuration then re-defines the Exim user, the unprivileged environment
5280 is probably not what is expected, so Exim logs a panic warning message (but
5283 However, if deliver_drop_privilege is set, the unprivileged state may well
5284 be exactly what is intended, so the warning has been cut out in that case,
5285 and Exim is allowed to try to write to its normal log files.
5291 1. smtp_return_error_details was not giving details for temporary sender
5292 or receiver verification errors.
5294 2. Diagnose a configuration error if two authenticators have the same public
5297 3. Exim used not to create the message log file for a message until the first
5298 delivery attempt. This could be confusing when incoming messages were held
5299 for policy or load reasons. The message log file is now created at the time
5300 the message is received, and an initial "Received" line is written to it.
5302 4. The automatically generated man page for command line options had a minor
5303 bug that caused no ill effects; however, a more serious problem was that
5304 the procedure for building the man page automatically didn't always
5305 operate. Consequently, release 4.20 contains an out-of-date version. This
5306 shouldn't happen again.
5308 5. When building Exim with embedded Perl support, the script that builds the
5309 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5310 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5312 6. The freeze_tell option was not being used for messages that were frozen on
5313 arrival, either by an ACL or by local_scan().
5315 7. Added the smtp_incomplete_transaction log selector.
5317 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5318 was accepting AUTH without a new EHLO.
5320 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5321 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5324 10. Logging of TCP/IP connections (when configured) now happens in the main
5325 daemon process instead of the child process, so that the TCP/IP connection
5326 count is more accurate (but it can never be perfect).
5328 11. The use of "drop" in a nested ACL was not being handled correctly in the
5329 outer ACL. Now, if condition failure induced by the nested "drop" causes
5330 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5331 or "require"), the connection is dropped.
5333 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5334 that yield "discard" can now be used with an "accept" or a "discard" verb,
5335 but an error is generated for any others (because I can't see a useful way
5336 to define what should happen).
5338 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5339 are now processed in the same way as lines in the Exim configuration file.
5340 In particular, continuation lines are supported.
5342 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5344 15. Added -ti meaning -t -i.
5346 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5347 domains, and warn by logging if others are found.
5349 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5350 returned from GIFCONF when Exim is trying to find the list of interfaces on
5351 a host. The code in os.c has been modified to copy each ifreq to an aligned
5352 structure in all cases.
5354 Also, in some cases, the returned ifreq's were being copied to a 'struct
5355 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5356 means the last couple of bytes of an IPv6 address could be chopped if the
5357 ifreq contained only a normal sockaddr (14 bytes storage).
5359 18. Named domain lists were not supported in the hosts_treat_as_local option.
5360 An entry such as +xxxx was not recognized, and was treated as a literal
5363 19. Ensure that header lines added by a DATA ACL are included in the reject log
5364 if the ACL subsequently rejects the message.
5366 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5367 MD5 (which is deprecated).
5369 21. When testing a filter file using -bf, Exim was writing a message when it
5370 took the sender from a "From " line in the message, but it was not doing so
5371 when it took $return_path from a Return-Path: header line. It now does.
5373 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5374 with a valid header line field name (a series of printing characters
5375 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5377 23. Changed "disc" in the source to "disk" to conform to the documentation and
5378 the book and for uniformity.
5380 24. Ignore Sendmail's -Ooption=value command line item.
5382 25. When execve() failed while trying to run a command in a pipe transport,
5383 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5384 could be confused with a return value of 69 from the command itself. This
5385 has been changed to 127, the value the shell returns if it is asked to run
5386 a non-existent command. The wording for the related log line suggests a
5387 non-existent command as the problem.
5389 26. If received_header_text expands to an empty string, do not add a Received:
5390 header line to the message. (Well, it adds a token one on the spool, but
5391 marks it "old" so that it doesn't get used or transmitted.)
5393 27. Installed eximstats 1.28 (addition of -nt option).
5395 28. There was no check for failure on the call to getsockname() in the daemon
5396 code. This can fail if there is a shortage of resources on the system, with
5397 ENOMEM, for example. A temporary error is now given on failure.
5399 29. Contrary to the C standard, it seems that in some environments, the
5400 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5401 program. Exim now does this explicitly; it affects the formatting of
5402 timestamps using strftime().
5404 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5405 complaints. I've now initialized all the variables, to avoid this.
5407 32. Header lines added by a system filter were not being "seen" during
5408 transport-time rewrites.
5410 33. The info_callback() function passed to OpenSSL is set up with type void
5411 (*)(SSL *, int, int), as described somewhere. However, when calling the
5412 function (actually a macro) that sets it up, the type void(*)() is
5413 expected. I've put in a cast to prevent warnings from picky compilers.
5415 34. If a DNS black list lookup found a CNAME record, but there were no A
5416 records associated with the domain it pointed at, Exim crashed.
5418 35. If a DNS black list lookup returned more than one A record, Exim ignored
5419 all but the first. It now scans all returned addresses if a particular IP
5420 value is being sought. In this situation, the contents of the
5421 $dnslist_value variable are a list of all the addresses, separated by a
5424 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5425 to accept a host name and all its aliases if the forward lookup for any of
5426 them yielded the IP address of the incoming connection. Now it accepts only
5427 those names whose forward lookup yields the correct IP address. Any other
5428 names are discarded. This closes a loophole whereby a rogue DNS
5429 administrator could create reverse DNS records to break through a
5430 wildcarded host restriction in an ACL.
5432 37. If a user filter or a system filter that ran in a subprocess used any of
5433 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5434 the wrong values were passed to the pipe command ($thisaddress had the
5435 value of $0, $0 had the value of $1, etc). This bug was introduced by
5436 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5438 38. Improved the line breaking for long SMTP error messages from ACLs.
5439 Previously, if there was no break point between 40 and 75 characters, Exim
5440 left the rest of the message alone. Two changes have been made: (a) I've
5441 reduced the minimum length to 35 characters; (b) if it can't find a break
5442 point between 35 and 75 characters, it looks ahead and uses the first one
5443 that it finds. This may give the occasional overlong line, but at least the
5444 remaining text gets split now.
5446 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5447 file descriptors might be low, and that setting 1000 would always raise it.
5448 It turns out that in some environments, the limit is already over 1000 and
5449 that lowering it causes trouble. So now Exim takes care not to decrease it.
5451 40. When delivering a message, the value of $return_path is set to $sender_
5452 address at the start of routing (routers may change the value). By an
5453 oversight, this default was not being set up when an address was tested by
5454 -bt or -bv, which affected the outcome if any router or filter referred to
5457 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5458 log only when "message" or "log_message" are set. However, if one of the
5459 conditions was an address verification, or a call to a nested ACL, the
5460 messages generated by the underlying test were being passed through. This
5461 no longer happens. The underlying message is available in $acl_verify_
5462 message for both "message" and "log_message" expansions, so it can be
5463 passed through if needed.
5465 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5466 new expansion $bh_ to give the encoded byte string without charset
5467 translation. Translation happens only if iconv() is available; HAVE_ICONV
5468 indicates this at build time. HEADERS_CHARSET gives the charset to
5469 translate to; headers_charset can change it in the configuration, and
5470 "headers charset" can change it in an individual filter file.
5472 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5473 that creates RFC 2047 encoded "words" labels them as that charset instead
5474 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5475 expansion operator; (ii) when Exim creates a From: line for a local
5476 message; (iii) when a header line is rewritten to include a "phrase" part.
5478 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5479 buggy, causing it to skip the first lines of messages whose message ID
5480 ended in 'D'. This would not have bitten before Exim release 4.14, because
5481 message IDs were unlikely to end in 'D' before then. The effect was to have
5482 incorrect size information for certain domains.
5484 45. #include "config.h" was missing at the start of the crypt16.c module. This
5485 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5488 46. If there was a timeout during a "random" callout check, Exim treated it as
5489 a failure of the random address, and carried on sending RSET and the real
5490 address. If the delay was just some slowness somewhere, the response to the
5491 original RCPT would be taken as a response to RSET and so on, causing
5492 mayhem of various kinds.
5494 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5495 when I implemented it. It didn't allow for the fact that some option values
5496 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5497 the right test for positive values.
5499 48. Domain names in DNS records are case-independent. Exim always looks them up
5500 in lower case. Some resolvers return domain names in exactly the case they
5501 appear in the zone file, that is, they may contain uppercase letters. Not
5502 all resolvers do this - some return always lower case. Exim was treating a
5503 change of case by a resolver as a change of domain, similar to a widening
5504 of a domain abbreviation. This triggered its re-routing code and so it was
5505 trying to route what was effectively the same domain again. This normally
5506 caused routing to fail (because the router wouldn't handle the domain
5507 twice). Now Exim checks for this case specially, and just changes the
5508 casing of the domain that it ultimately uses when it transmits the message
5511 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5514 50. If a filter generated a file delivery with a non-absolute name (possible if
5515 no home directory exists for the router), the forbid_file option was not
5518 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5519 the existing equality matching.
5521 52. Exim was using ints instead of ino_t variables in some places where it was
5522 dealing with inode numbers.
5524 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5525 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5526 called TMPDIR, and if it finds it is different, it changes its value.
5528 54. The smtp_printf() function is now made available to local_scan() so
5529 additional output lines can be written before returning. There is also an
5530 smtp_fflush() function to enable the detection of a dropped connection.
5531 The variables smtp_input and smtp_batched_input are exported to
5534 55. Changed the default runtime configuration: the message "Unknown user"
5535 has been removed from the ACL, and instead placed on the localuser router,
5536 using the cannot_route_message feature. This means that any verification
5537 failures that generate their own messages won't get overridden. Similarly,
5538 the "Unrouteable address" message that was in the ACL for unverifiable
5539 relay addresses has also been removed.
5541 56. Added hosts_avoid_esmtp to the smtp transport.
5543 57. The exicyclog script was not checking for the esoteric option
5544 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5545 will work only if exicyclog is run under the appropriate euid.
5547 58. Following a discussion on the list, the rules by which Exim recognises line
5548 endings on incoming messages have been changed. The -dropcr and drop_cr
5549 options are now no-ops, retained only for backwards compatibility. The
5550 following line terminators are recognized: LF CRLF CR. However, special
5551 processing applies to CR:
5553 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5554 nor a local message in the state where . is a terminator.
5556 (ii) If a bare CR is encountered in a header line, an extra space is added
5557 after the line terminator so as not to end the header. The reasoning
5558 behind this is that bare CRs in header lines are most likely either
5559 to be mistakes, or people trying to play silly games.
5561 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5562 was being incorrectly given as 18 bytes larger than it should have been.
5563 This is a VOB (very old bug).
5565 60. This may never have affected anything current, but just in case it has:
5566 When the local host is found other than at the start of a list of hosts,
5567 the local host, those with the same MX, and any that follow, are discarded.
5568 When the list in question was part of a longer list of hosts, the following
5569 hosts (not currently being processed) were also being discarded. This no
5570 longer happens. I'm not sure if this situation could ever has previously
5573 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5576 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5577 header. This is something that is recommended in a new Internet Draft, and
5578 is something that is documented as being done by Sendmail. There are two
5579 possible values. For messages generated by the autoreply transport, Exim
5582 Auto-Submitted: auto-replied
5584 whereas for all other generated messages (e.g. bounces) it adds
5586 Auto-Submitted: auto-generated
5588 63. The "personal" condition in filters now includes a test for the
5589 Auto-Submitted: header. If it contains the string "auto-" the message it
5590 not considered personal.
5592 64. Added rcpt_include_affixes as a generic transport option.
5594 65. Added queue_only_override (default true).
5596 66. Added the syslog_duplication option.
5598 67. If what should have been the first header line of a message consisted of
5599 a space followed by a colon, Exim was mis-interpreting it as a header line.
5600 It isn't of course - it is syntactically invalid and should therefore be
5601 treated as the start of the message body. The misbehaviour could have
5602 caused a number of strange effects, including loss of data in subsequent
5603 header lines, and spool format errors.
5605 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5606 client host had authenticated. This control can now be exercised by an ACL
5607 for more flexibility.
5609 69. By default, callouts do not happen when testing with -bh. There is now a
5610 variant, -bhc, which does actually run the callout code, including
5611 consulting and updating the callout cache.
5613 70. Added support for saslauthd authentication, courtesy of Alexander
5616 71. If statvfs() failed on the spool or log directories while checking their
5617 size for availability, Exim confusingly gave the error "space shortage".
5618 Furthermore, in debugging mode it crashed with a floating point exception.
5619 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5620 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5621 serious problem, Exim now writes to the main and panic logs when this
5622 happens, with details of the failure. It then refuses to accept the
5623 incoming message, giving the message "spool directory problem" or "log
5624 directory problem" with a 421 code for SMTP messages.
5626 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5627 0, 1, and 2 exist, because some OS complain for execs without them (see
5628 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5629 descriptors. However, the code omitted to check that the open succeeded,
5630 causing mysterious errors if for some reason the permissions on /dev/null
5631 got screwed. Now Exim writes a message to the main and panic logs, and
5632 bombs out if it can't open /dev/null.
5634 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5635 interact so that it is all more flexible. It is supposed to remain
5636 backwards compatible. Also added extra_local_interfaces.
5638 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5639 to bomb out with an assertion failure - to the client this appears as a
5640 connection drop. This problem occurs in the part of the code that was taken
5641 from the Samba project. Fortunately, the assertion is in a very simple
5642 function, so I have fixed this by reproducing the function inline in the
5643 one place where it is called, and arranging for authentication to fail
5644 instead of killing the process with assert().
5646 75. The SPA client code was not working when the server requested OEM rather
5647 than Unicode encoding.
5649 76. Added code to make require_files with a specific uid setting more usable in
5650 the case where statting the file as root fails - usually a non-root-mounted
5651 NFS file system. When this happens and the failure is EACCES, Exim now
5652 forks a subprocess and does the per-uid checking as the relevant uid.
5654 77. Added process_log_path.
5656 78. If log_file_path was not explicitly set, a setting of check_log_space or
5657 check_log_inodes was ignored.
5659 79. If a space check for the spool or log partitions fails, the incident is now
5660 logged. Of course, in the latter case the data may get lost...
5662 80. Added the %p formatting code to string_format() so that it can be used to
5663 print addresses in debug_print(). Adjusted all the address printing in the
5664 debugging in store.c to use %p rather than %d.
5666 81. There was a concern that a line of code in smtp_in.c could overflow a
5667 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5668 initially expressed, the concern was not well-founded, because trailing
5669 spaces are removed early. However, if the trailing spaces were followed by
5670 a NULL, they did not get removed, so the overflow was possible. Two fixes
5673 (a) I re-wrote the offending code in a cleaner fashion.
5674 (b) If an incoming SMTP command contains a NULL character, it is rejected
5677 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5678 runs initgroups(), so that if the Exim user is in any additional groups,
5679 they will be used during message reception.
5685 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob