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.
40 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
42 NM/01 Bugzilla 1197 - Spec typo
43 Bugzilla 1196 - Spec examples corrections
45 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
47 PP/09 Add gnutls_enable_pkcs11 option.
49 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
50 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
56 PP/01 Handle short writes when writing local log-files.
57 In practice, only affects FreeBSD (8 onwards).
58 Bugzilla 1053, with thanks to Dmitry Isaikin.
60 NM/01 Bugzilla 949 - Documentation tweak
62 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
65 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
67 PP/02 Implemented gsasl authenticator.
69 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
71 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
72 `pkg-config foo` for cflags/libs.
74 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
75 with rest of GSASL and with heimdal_gssapi.
77 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
78 `pkg-config foo` for cflags/libs for the TLS implementation.
80 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
81 properties get this fed in as external SSF. A number of robustness
82 and debugging improvements to the cyrus_sasl authenticator.
84 PP/08 cyrus_sasl server now expands the server_realm option.
86 PP/09 Bugzilla 1214 - Log authentication information in reject log.
87 Patch by Jeremy Harris.
89 PP/10 Added dbmjz lookup type.
91 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
93 PP/12 MAIL args handles TAB as well as SP, for better interop with
94 non-compliant senders.
95 Analysis and variant patch by Todd Lyons.
97 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
98 Bug report from Lars Müller <lars@samba.org> (via SUSE),
99 Patch from Dirk Mueller <dmueller@suse.com>
101 PP/13 tls_peerdn now print-escaped for spool files.
102 Observed some $tls_peerdn in wild which contained \n, which resulted
103 in spool file corruption.
105 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
106 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
107 or write after TLS renegotiation, which otherwise led to messages
110 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
111 as a tracking header (ie: a signed header comes before the signature).
112 Patch from Wolfgang Breyha.
114 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
115 comma-sep list; embedded commas doubled.
117 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
119 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
121 Report and patch from Dmitry Banschikov.
123 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
124 Removed SSL_clear() after SSL_new() which led to protocol negotiation
125 failures. We appear to now support TLS1.1+ with Exim.
127 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
128 lets Exim select keys and certificates based upon TLS SNI from client.
129 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
130 before an outbound SMTP session. New log_selector, +tls_sni.
132 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
133 NULL dereference. Report and patch from Alun Jones.
135 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
136 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
137 Not seeing resolver debug output on NetBSD, but suspect this is a
138 resolver implementation change.
140 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
141 Left warnings. Added "eximon gdb" invocation mode.
143 PP/21 Defaulting "accept_8bitmime" to true, not false.
145 PP/22 Added -bw for inetd wait mode support.
147 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
148 locate the relevant includes and libraries. Made this the default.
150 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
151 Bugzilla 1246, report and most of solution from Tomasz Kusy.
153 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
154 This may cause build issues on older platforms.
156 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
157 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
158 gnutls_require_mac & gnutls_require_protocols (no longer supported).
159 Added SNI support via GnuTLS too.
160 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
162 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
164 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
165 Applied second patch from Janne, implementing suggestion to default
166 multiple-strings-in-record handling to match SPF spec.
168 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
170 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
171 read-only, out of scope).
172 Patch from Wolfgang Breyha, report from Stuart Northfield.
174 PP/29 Fix three issues highlighted by clang analyser static analysis.
175 Only crash-plausible issue would require the Cambridge-specific
176 iplookup router and a misconfiguration.
177 Report from Marcin Mirosław.
179 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
181 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
182 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
183 As part of this, removing so much warning spew let me fix some minor
184 real issues in debug logging.
186 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
187 assignment on my part. Fixed.
189 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
190 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
191 Janne Snabb (who went above and beyond: thank you).
193 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
194 string otherwise requires a connection and a bunch more work and it's
195 relatively easy to get wrong. Should also expose TLS library linkage
198 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
199 64-bit ${eval} (JH/03).
201 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
202 GNU libc to support some of the 64-bit stuff, should not lead to
203 conflicts. Defined before os.h is pulled in, so if a given platform
204 needs to override this, it can.
206 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
207 protection layer was required, which is not implemented.
208 Bugzilla 1254, patch from Wolfgang Breyha.
210 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
211 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
212 tls_dhparam take prime identifiers. Also unbreak combination of
213 OpenSSL+DH_params+TLSSNI.
215 PP/39 Disable SSLv2 by default in OpenSSL support.
221 PP/01 Solaris build fix for Oracle's LDAP libraries.
222 Bugzilla 1109, patch from Stephen Usher.
224 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
226 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
229 TF/02 Fix a couple more cases where we did not log the error message
230 when unlink() failed. See also change 4.74-TF/03.
232 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
233 lock up or crash if it happened to be inside a call to libc when it
234 got a SIGUSR1 from exiwhat.
236 The SIGUSR1 handler appends the current process status to the process
237 log which is later printed by exiwhat. It used to use the general
238 purpose logging code to do this, but several functions it calls are
239 not safe for signals.
241 The new output code in the SIGUSR1 handler is specific to the process
242 log, and simple enough that it's easy to inspect for signal safety.
243 Removing some special cases also simplifies the general logging code.
244 Removing the spurious timestamps from the process log simplifies
247 TF/04 Improved ratelimit ACL condition.
249 The /noupdate option has been deprecated in favour of /readonly which
250 has clearer semantics. The /leaky, /strict, and /readonly update modes
251 are mutually exclusive. The update mode is no longer included in the
252 database key; it just determines when the database is updated. (This
253 means that when you upgrde Exim will forget old rate measurements.)
255 Exim now checks that the per_* options are used with an update mode that
256 makes sense for the current ACL. For example, when Exim is processing a
257 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
258 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
259 must specify per_mail/readonly. If you omit the update mode it defaults to
260 /leaky where that makes sense (as before) or /readonly where required.
262 The /noupdate option is now undocumented but still supported for
263 backwards compatibility. It is equivalent to /readonly except that in
264 ACLs where /readonly is required you may specify /leaky/noupdate or
265 /strict/noupdate which are treated the same as /readonly.
267 A useful new feature is the /count= option. This is a generalization
268 of the per_byte option, so that you can measure the throughput of other
269 aggregate values. For example, the per_byte option is now equivalent
270 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
272 The per_rcpt option has been generalized using the /count= mechanism
273 (though it's more complicated than the per_byte equivalence). When it is
274 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
275 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
276 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
277 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
278 Note that using per_rcpt with a non-readonly update mode in more than
279 one ACL will cause the recipients to be double-counted. (The per_mail
280 and per_byte options don't have this problem.)
282 The handling of very low rates has changed slightly. If the computed rate
283 is less than the event's count (usually one) then this event is the first
284 after a long gap. In this case the rate is set to the same as this event's
285 count, so that the first message of a spam run is counted properly.
287 The major new feature is a mechanism for counting the rate of unique
288 events. The new per_addr option counts the number of different
289 recipients that someone has sent messages to in the last time period. It
290 behaves like per_rcpt if all the recipient addresses are different, but
291 duplicate recipient addresses do not increase the measured rate. Like
292 the /count= option this is a general mechanism, so the per_addr option
293 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
294 example, measure the rate that a client uses different sender addresses
295 with the options per_mail/unique=$sender_address. There are further
296 details in the main documentation.
298 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
300 TF/06 Removed a few PCRE remnants.
302 TF/07 Automatically extract Exim's version number from tags in the git
303 repository when doing development or release builds.
305 PP/02 Raise smtp_cmd_buffer_size to 16kB.
306 Bugzilla 879. Patch from Paul Fisher.
308 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
309 Heavily based on revision 40f9a89a from Simon Arlott's tree.
312 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
314 PP/05 Variable $av_failed, true if the AV scanner deferred.
315 Bugzilla 1078. Patch from John Horne.
317 PP/06 Stop make process more reliably on build failure.
318 Bugzilla 1087. Patch from Heiko Schlittermann.
320 PP/07 Make maildir_use_size_file an _expandable_ boolean.
321 Bugzilla 1089. Patch from Heiko Schlittermann.
323 PP/08 Handle ${run} returning more data than OS pipe buffer size.
324 Bugzilla 1131. Patch from Holger Weiß.
326 PP/09 Handle IPv6 addresses with SPF.
327 Bugzilla 860. Patch from Wolfgang Breyha.
329 PP/10 GnuTLS: support TLS 1.2 & 1.1.
331 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
334 PP/11 match_* no longer expand right-hand-side by default.
335 New compile-time build option, EXPAND_LISTMATCH_RHS.
336 New expansion conditions, "inlist", "inlisti".
338 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
340 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
342 PP/14 fix log_write() format string regression from TF/03.
343 Bugzilla 1152. Patch from Dmitry Isaikin.
349 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
351 PP/02 Harmonised TLS library version reporting; only show if debugging.
352 Layout now matches that introduced for other libraries in 4.74 PP/03.
354 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
356 PP/04 New "dns_use_edns0" global option.
358 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
361 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
362 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
364 TK/01 Updated PolarSSL code to 0.14.2.
365 Bugzilla 1097. Patch from Andreas Metzler.
367 PP/07 Catch divide-by-zero in ${eval:...}.
370 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
373 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
374 format-string attack -- SECURITY: remote arbitrary code execution.
376 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
377 time unintentionally subject to list matching rules, letting the header
378 cause arbitrary Exim lookups (of items which can occur in lists, *not*
379 arbitrary string expansion). This allowed for information disclosure.
381 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
382 INT_MIN/-1 -- value coerced to INT_MAX.
388 NM/01 Workround for PCRE version dependency in version reporting
391 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
392 This fixes portability to compilers other than gcc, notably
393 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
395 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
396 makefiles for portability to HP-UX and POSIX correctness.
398 PP/01 Permit LOOKUP_foo enabling on the make command-line.
399 Also via indented variable definition in the Makefile.
400 (Debugging by Oliver Heesakkers).
402 PP/02 Restore caching of spamd results with expanded spamd_address.
403 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
405 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
406 Improves build reliability. Fix from: Frank Elsner
408 NM/02 Fix wide character breakage in the rfc2047 coding
409 Fixes bug 1064. Patch from Andrey N. Oktyabrski
411 NM/03 Allow underscore in dnslist lookups
412 Fixes bug 1026. Patch from Graeme Fowler
414 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
415 Code patches from Adam Ciarcinski of NetBSD.
417 NM/04 Fixed exiqgrep to cope with mailq missing size issue
420 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
421 is logged, to avoid truncation. Patch from John Horne.
423 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
424 Patch from Jakob Hirsch.
426 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
427 SQL string expansion failure details.
428 Patch from Andrey Oktyabrski.
430 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
431 Patch from Simon Arlott.
433 PP/09 New lookups functionality failed to compile on old gcc which rejects
434 extern declarations in function scope.
435 Patch from Oliver Fleischmann
437 PP/10 Use sig_atomic_t for flags set from signal handlers.
438 Check getgroups() return and improve debugging.
439 Fixed developed for diagnosis in bug 927 (which turned out to be
442 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
443 Patch from Mark Zealey.
445 PP/12 Bugzilla 1056: Improved spamd server selection.
446 Patch from Mark Zealey.
448 PP/13 Bugzilla 1086: Deal with maildir quota file races.
449 Based on patch from Heiko Schlittermann.
451 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
452 Patch from Uwe Doering, sign-off by Michael Haardt.
454 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
455 variable declaration deep within a block. Bug and patch from
458 PP/15 lookups-Makefile IRIX compatibilty coercion.
460 PP/16 Make DISABLE_DKIM build knob functional.
462 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
463 Patch by Simon Arlott
465 TF/03 Fix valgrind.h portability to C89 compilers that do not support
466 variable argument macros. Our copy now differs from upstream.
472 TF/01 Failure to get a lock on a hints database can have serious
473 consequences so log it to the panic log.
475 TF/02 Log LMTP confirmation messages in the same way as SMTP,
476 controlled using the smtp_confirmation log selector.
478 TF/03 Include the error message when we fail to unlink a spool file.
480 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
481 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
482 for maintaining out-of-tree patches for some time.
484 PP/01 Bugzilla 139: Documentation and portability issues.
485 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
486 Handle per-OS dynamic-module compilation flags.
488 PP/02 Let /dev/null have normal permissions.
489 The 4.73 fixes were a little too stringent and complained about the
490 permissions on /dev/null. Exempt it from some checks.
491 Reported by Andreas M. Kirchwitz.
493 PP/03 Report version information for many libraries, including
494 Exim version information for dynamically loaded libraries. Created
495 version.h, now support a version extension string for distributors
496 who patch heavily. Dynamic module ABI change.
498 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
499 privilege escalation vulnerability whereby the Exim run-time user
500 can cause root to append content of the attacker's choosing to
503 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
506 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
507 If dropping privileges for untrusted macros, we disabled normal logging
508 on the basis that it would fail; for the Exim run-time user, this is not
509 the case, and it resulted in successful deliveries going unlogged.
510 Fixed. Reported by Andreas Metzler.
516 PP/01 Date: & Message-Id: revert to normally being appended to a message,
517 only prepend for the Resent-* case. Fixes regression introduced in
518 Exim 4.70 by NM/22 for Bugzilla 607.
520 PP/02 Include check_rfc2047_length in configure.default because we're seeing
521 increasing numbers of administrators be bitten by this.
523 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
525 PP/03 Bugzilla 994: added openssl_options main configuration option.
527 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
529 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
531 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
533 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
534 without a peer certificate, leading to a segfault because of an
535 assumption that peers always have certificates. Be a little more
536 paranoid. Problem reported by Martin Tscholak.
538 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
539 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
540 NB: ClamAV planning to remove STREAM in "middle of 2010".
541 CL also introduces -bmalware, various -d+acl logging additions and
542 more caution in buffer sizes.
544 PP/09 Implemented reverse_ip expansion operator.
546 PP/10 Bugzilla 937: provide a "debug" ACL control.
548 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
550 PP/12 Bugzilla 973: Implement --version.
552 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
554 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
556 PP/15 Bugzilla 816: support multiple condition rules on Routers.
558 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
559 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
560 ignore trailing whitespace.
562 JJ/02 prevent non-panic DKIM error from being sent to paniclog
564 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
567 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
568 Notification from Dr Andrew Aitchison.
570 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
571 ExtendedDetectionInfo response format.
572 Notification from John Horne.
574 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
577 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
578 XSL and documented dependency on system catalogs, with examples of how
581 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
584 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
585 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
586 configuration file which is writeable by the Exim user or group.
588 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
589 of configuration files to cover files specified with the -C option if
590 they are going to be used with root privileges, not just the default
593 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
594 option (effectively making it always true).
596 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
597 files to be used while preserving root privileges.
599 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
600 that rogue child processes cannot use them.
602 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
603 run-time user, instead of root.
605 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
606 Exim run-time user without dropping privileges.
608 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
609 result string, instead of calling string_vformat() twice with the same
612 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
613 for other users. Others should always drop root privileges if they use
614 -C on the command line, even for a whitelisted configure file.
616 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
618 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
624 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
625 $data_path, and $header_path variables; fixed documentation bugs and
628 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
629 exipick to access non-standard spools, including the "frozen" queue
632 NM/01 Bugzilla 965: Support mysql stored procedures.
633 Patch from Alain Williams
635 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
637 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
638 Patch from Andreas Metzler
640 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
641 Patch from Kirill Miazine
643 NM/05 Bugzilla 671: Added umask to procmail example.
645 JJ/03 installed exipick 20100323.0, fixing doc bug
647 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
648 directory. Notification and patch from Dan Rosenberg.
650 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
652 TK/02 Improve log output when DKIM signing operation fails.
654 MH/01 Treat the transport option dkim_domain as a colon separated
655 list, not as a single string, and sign the message with each element,
656 omitting multiple occurences of the same signer.
658 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
659 Bugzilla 985, 986. Patch by Simon Arlott
661 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
662 Patch by Simon Arlott
664 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
665 MBX locking. Notification from Dan Rosenberg.
671 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
673 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
675 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
677 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
679 NM/04 Bugzilla 915: Flag broken perl installation during build.
685 TK/01 Added patch by Johannes Berg that expands the main option
686 "spamd_address" if it starts with a dollar sign.
688 TK/02 Write list of recipients to X-Envelope-Sender header when building
689 the mbox-format spool file for content scanning (suggested by Jakob
692 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
693 (http://www.dcc-servers.net/) support via dccifd. Activated by
694 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
696 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
697 by Mark Daniel Reidel <mr@df.eu>.
699 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
700 When building exim an external PCRE library is now needed -
701 PCRE is a system library on the majority of modern systems.
702 See entry on PCRE_LIBS in EDITME file.
704 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
705 conversation. Added nologin parameter to request.
706 Patch contributed by Kirill Miazine.
708 TF/01 Do not log submission mode rewrites if they do not change the address.
710 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
712 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
713 log files in place. Contributed by Roberto Lima.
715 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
717 TF/03 Bugzilla 615: When checking the local_parts router precondition
718 after a local_part_suffix or local_part_prefix option, Exim now
719 does not use the address's named list lookup cache, since this
720 contains cached lookups for the whole local part.
722 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
723 Robert Millan. Documentation is in experimental-spec.txt.
725 TF/04 Bugzilla 668: Fix parallel build (make -j).
727 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
729 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
730 Patch provided by Jan Srzednicki.
732 TF/05 Leading white space used to be stripped from $spam_report which
733 wrecked the formatting. Now it is preserved.
735 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
736 that they are available at delivery time.
738 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
740 TF/08 TLS error reporting now respects the incoming_interface and
741 incoming_port log selectors.
743 TF/09 Produce a more useful error message if an SMTP transport's hosts
744 setting expands to an empty string.
746 NM/06 Bugzilla 744: EXPN did not work under TLS.
747 Patch provided by Phil Pennock.
749 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
750 Patch provided by Richard Godbee.
752 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
753 acl_smtp_notquit, added index entry.
755 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
756 Patch provided by Eugene Bujak.
758 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
759 accept(). Patch provided by Maxim Dounin.
761 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
762 Patch provided by Phil Pennock.
764 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
766 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
767 Patch provided by Brad "anomie" Jorsch.
769 NM/14 Bugzilla 622: Added timeout setting to transport filter.
770 Patch provided by Dean Brooks.
772 TK/05 Add native DKIM support (does not depend on external libraries).
774 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
775 Patch provided by Graeme Fowler.
777 NM/16 Bugzilla 851: Documentation example syntax fix.
779 NM/17 Changed NOTICE file to remove references to embedded PCRE.
781 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
784 NM/19 Bugzilla 745: TLS version reporting.
785 Patch provided by Phil Pennock.
787 NM/20 Bugzilla 167: bool: condition support.
788 Patch provided by Phil Pennock.
790 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
791 clients. Patch provided by Phil Pennock.
793 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
794 Patch provided by Brad "anomie" Jorsch.
796 NM/23 Bugzilla 687: Fix misparses in eximstats.
797 Patch provided by Heiko Schlittermann.
799 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
800 Patch provided by Heiko Schlittermann.
802 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
803 plus update to original patch.
805 NM/26 Bugzilla 799: Documentation correction for ratelimit.
807 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
808 Patch provided by David Brownlee.
810 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
812 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
814 NM/30 Bugzilla 888: TLS documentation bugfixes.
816 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
818 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
819 Unlike the original bugzilla I have changed all shell scripts in src tree.
821 NM/33 Bugzilla 898: Transport filter timeout fix.
822 Patch by Todd Rinaldo.
824 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
825 Patch by Serge Demonchaux.
827 NM/35 Bugzilla 39: Base64 decode bug fixes.
828 Patch by Jakob Hirsch.
830 NM/36 Bugzilla 909: Correct connect() call in dcc code.
832 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
834 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
836 NM/39 Bugzilla 911: Fixed MakeLinks build script.
842 TK/01 Add preliminary DKIM support. Currently requires a forked version of
843 ALT-N's libdkim that I have put here:
844 http://duncanthrax.net/exim-experimental/
846 Note to Michael Haardt: I had to rename some vars in sieve.c. They
847 were called 'true' and it seems that C99 defines that as a reserved
848 keyword to be used with 'bool' variable types. That means you could
849 not include C99-style headers which use bools without triggering
850 build errors in sieve.c.
852 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
853 as mailq or other aliases. Changed the --help handling significantly
854 to do whats expected. exim_usage() emits usage/help information.
856 SC/01 Added the -bylocaldomain option to eximstats.
858 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
860 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
862 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
868 PH/01 Another patch from the Sieve maintainer.
870 PH/02 When an IPv6 address is converted to a string for single-key lookup
871 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
872 dots are used instead of colons so that keys in lsearch files need not
873 contain colons. This was done some time before quoting was made available
874 in lsearch files. However, iplsearch files do require colons in IPv6 keys
875 (notated using the quote facility) so as to distinguish them from IPv4
876 keys. This meant that lookups for IP addresses in host lists did not work
877 for iplsearch lookups.
879 This has been fixed by arranging for IPv6 addresses to be expressed with
880 colons if the lookup type is iplsearch. This is not incompatible, because
881 previously such lookups could never work.
883 The situation is now rather anomolous, since one *can* have colons in
884 ordinary lsearch keys. However, making the change in all cases is
885 incompatible and would probably break a number of configurations.
887 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
890 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
891 conversion specification without a maximum field width, thereby enabling
892 a rogue spamd server to cause a buffer overflow. While nobody in their
893 right mind would setup Exim to query an untrusted spamd server, an
894 attacker that gains access to a server running spamd could potentially
895 exploit this vulnerability to run arbitrary code as the Exim user.
897 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
898 $primary_hostname instead of what libspf2 thinks the hosts name is.
900 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
901 a directory entry by the name of the lookup key. Previously, if a
902 symlink pointed to a non-existing file or a file in a directory that
903 Exim lacked permissions to read, a lookup for a key matching that
904 symlink would fail. Now it is enough that a matching directory entry
905 exists, symlink or not. (Bugzilla 503.)
907 PH/03 The body_linecount and body_zerocount variables are now exported in the
910 PH/04 Added the $dnslist_matched variable.
912 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
913 This means they are set thereafter only if the connection becomes
916 PH/06 Added the client_condition to authenticators so that some can be skipped
917 by clients under certain conditions.
919 PH/07 The error message for a badly-placed control=no_multiline_responses left
920 "_responses" off the end of the name.
922 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
924 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
925 (without spaces) instead of just copying the configuration text.
927 PH/10 Added the /noupdate option to the ratelimit ACL condition.
929 PH/11 Added $max_received_linelength.
931 PH/12 Added +ignore_defer and +include_defer to host lists.
933 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
934 way in which PCRE > 7.0 is built.
936 PH/14 Implemented queue_only_load_latch.
938 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
939 MAIL command. The effect was to mangle the value on 64-bit systems.
941 PH/16 Another patch from the Sieve maintainer.
943 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
945 PH/18 If a system quota error occurred while trying to create the file for
946 a maildir delivery, the message "Mailbox is full" was not appended to the
947 bounce if the delivery eventually timed out. Change 4.67/27 below applied
948 only to a quota excession during the actual writing of the file.
950 PH/19 It seems that peer DN values may contain newlines (and other non-printing
951 characters?) which causes problems in log lines. The DN values are now
952 passed through string_printing() before being added to log lines.
954 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
955 and InterBase are left for another time.)
957 PH/21 Added message_body_newlines option.
959 PH/22 Guard against possible overflow in moan_check_errorcopy().
961 PH/23 POSIX allows open() to be a macro; guard against that.
963 PH/24 If the recipient of an error message contained an @ in the local part
964 (suitably quoted, of course), incorrect values were put in $domain and
965 $local_part during the evaluation of errors_copy.
971 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
972 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
975 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
976 issue a MAIL command.
978 PH/02 In an ACL statement such as
980 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
982 if a client was not listed at all, or was listed with a value other than
983 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
984 the condition was not true (as it should be), so access was not denied.
985 The bug was that the ! inversion was incorrectly passed on to the second
986 item. This has been fixed.
988 PH/03 Added additional dnslists conditions == and =& which are different from
989 = and & when the dns lookup returns more than one IP address.
991 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
992 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
994 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
995 FSYNC, which compiles an option called disable_fsync that allows for
996 bypassing fsync(). The documentation is heavily laced with warnings.
998 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1000 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1001 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1002 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1003 including adding "make clean"; (3) Added -fPIC when compiling the test
1004 dynamically loaded module, to get rid of a warning.
1006 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1007 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1008 = 0s. The bug is otherwise harmless.
1010 PH/07 There was a bug in the dovecot authenticator such that the value of
1011 $auth1 could be overwritten, and so not correctly preserved, after a
1012 successful authentication. This usually meant that the value preserved by
1013 the server_setid option was incorrect.
1015 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1017 PH/09 Installed PCRE release 7.0.
1019 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1020 run for batched SMTP input. It is now run at the start of every message
1021 in the batch. While fixing this I discovered that the process information
1022 (output by running exiwhat) was not always getting set for -bs and -bS
1023 input. This is fixed, and it now also says "batched" for BSMTP.
1025 PH/11 Added control=no_pipelining.
1027 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1028 patch, slightly modified), and move the expansion of helo_data till after
1029 the connection is made in the smtp transport (so it can use these
1032 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1034 PH/14 Added log_selector = +pid.
1036 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1038 PH/16 Add ${if forany and ${if forall.
1040 PH/17 Added dsn_from option to vary the From: line in DSNs.
1042 PH/18 Flush SMTP output before performing a callout, unless control =
1043 no_callout_flush is set.
1045 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1046 was true (the default) a successful delivery failed to delete the retry
1047 item, thus causing premature timeout of the address. The bug is now
1050 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1052 PH/21 Long custom messages for fakedefer and fakereject are now split up
1053 into multiline reponses in the same way that messages for "deny" and
1054 other ACL rejections are.
1056 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1057 with slight modification.
1059 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1060 draft, changing the syntax and factoring some duplicate code".
1062 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1063 for deliveries of the second and subsequent messages over the same SMTP
1066 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1067 ${reduce, with only minor "tidies".
1069 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1071 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1072 expansion side effects.
1074 PH/27 When a message times out after an over-quota error from an Exim-imposed
1075 quota, the bounce message says "mailbox is full". This message was not
1076 being given when it was a system quota that was exceeded. It now should
1079 MH/03 Made $recipients available in local_scan(). local_scan() already has
1080 better access to the recipient list through recipients_list[], but
1081 $recipients can be useful in postmaster-provided expansion strings.
1083 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1084 in the case of a MAIL command with additional options following the
1085 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1086 were accidentally chopped off.
1088 PH/29 SMTP synchronization checks are implemented when a command is read -
1089 there is a check that no more input is waiting when there shouldn't be
1090 any. However, for some commands, a delay in an ACL can mean that it is
1091 some time before the response is written. In this time, more input might
1092 arrive, invalidly. So now there are extra checks after an ACL has run for
1093 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1094 pipelining has not been advertised.
1096 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1098 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1099 returned in the SMTP response when smtp_return_error_details was false.
1100 This has been fixed.
1102 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1103 instead of the C I/O that was originally supplied, because problems were
1104 reported on Solaris.
1106 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1107 Exim which did not show up earlier: it was assuming that a call to
1108 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1109 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1110 was a macro that became an assignment, so it seemed to work. This has
1111 changed to a proper function call with a void return, hence the compile
1112 error. Exim's code has been fixed.
1114 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1117 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1119 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1121 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1122 because it turns out to be too big or there is a timeout) there is no
1123 "Completed" line in the log. When some messages of this type were
1124 selected by exigrep, they were listed as "not completed". Others were
1125 picked up by some special patterns. I have improved the selection
1126 criteria to be more general.
1128 PH/38 The host_find_failed option in the manualroute router can now be set
1129 to "ignore", to completely ignore a host whose IP address cannot be
1130 found. If all hosts are ignored, the behaviour is controlled by the new
1131 host_all_ignored option.
1133 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1134 homing or because of multiple MX records with /mx) generated more than
1135 one IP address, and the following item turned out to be the local host,
1136 all the secondary addresses of the first item were incorrectly removed
1137 from the list, along with the local host and any following hosts (which
1138 is what is supposed to happen).
1140 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1141 whoever called Exim into the -H file. In the case of the daemon it was
1142 behaving confusingly. When first started, it used values for whoever
1143 started the daemon, but after a SIGHUP it used the Exim user (because it
1144 calls itself on a restart). I have changed the code so that it now always
1147 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1148 message are rejected with the same error (e.g. no authentication or bad
1149 sender address), and a DATA command is nevertheless sent (as can happen
1150 with PIPELINING or a stupid MUA), the error message that was given to the
1151 RCPT commands is included in the rejection of the DATA command. This is
1152 intended to be helpful for MUAs that show only the final error to their
1155 PH/42 Another patch from the Sieve maintainer.
1157 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1158 Eximstats - Fixed some broken HTML links and added missing column headers
1160 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1161 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1163 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1165 PH/43 Yet another patch from the Sieve maintainer.
1167 PH/44 I found a way to check for a TCP/IP connection going away before sending
1168 the response to the final '.' that terminates a message, but only in the
1169 case where the client has not sent further data following the '.'
1170 (unfortunately, this is allowed). However, in many cases there won't be
1171 any further data because there won't be any more messages to send. A call
1172 to select() can be used: if it shows that the input is "ready", there is
1173 either input waiting, or the socket has been closed. An attempt to read
1174 the next input character can distinguish the two cases. Previously, Exim
1175 would have sent an OK response which the client would never have see.
1176 This could lead to message repetition. This fix should cure that, at
1177 least in a lot of common cases.
1179 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1180 advertised in response to EHLO.
1186 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1187 fixed by 4.65/MH/01 (is this a record?) are fixed:
1189 (i) An empty string was always treated as zero by the numeric comparison
1190 operators. This behaviour has been restored.
1192 (ii) It is documented that the numeric comparison operators always treat
1193 their arguments as decimal numbers. This was broken in that numbers
1194 starting with 0 were being interpreted as octal.
1196 While fixing these problems I realized that there was another issue that
1197 hadn't been noticed. Values of message_size_limit (both the global option
1198 and the transport option) were treated as octal if they started with 0.
1199 The documentation was vague. These values are now always treated as
1200 decimal, and I will make that clear in the documentation.
1206 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1207 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1210 MH/01 Don't check that the operands of numeric comparison operators are
1211 integers when their expansion is in "skipping" mode (fixes bug
1212 introduced by 4.64-PH/07).
1214 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1215 child addresses, Exim now panics and dies. Previously, because the count
1216 is held in a short int, deliveries were likely to be lost. As such a
1217 large number of recipients for a single message is ridiculous
1218 (performance will be very, very poor), I have chosen to impose a limit
1219 rather than extend the field.
1225 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1226 leftover -K file (the existence of which was triggered by #402).
1227 While we were at it, introduced process PID as part of the -K
1228 filename. This should rule out race conditions when creating
1231 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1232 processing considerably. Previous code took too long for large mails,
1233 triggering a timeout which in turn triggers #401.
1235 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1236 in the DK code in transports.c. sendfile() is not really portable,
1237 hence the _LINUX specificness.
1239 TF/01 In the add_headers option to the mail command in an Exim filter,
1240 there was a bug that Exim would claim a syntax error in any
1241 header after the first one which had an odd number of characters
1244 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1245 callout verification, Exim cached a "reject" for the entire domain. This
1246 is correct for most verifications, but it is not correct for a recipient
1247 verification with use_sender or use_postmaster set, because in that case
1248 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1249 case of MAIL FROM:<> rejection from other early rejections (e.g.
1250 rejection of HELO). When verifying a recipient using a non-null MAIL
1251 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1252 Whatever the result of the callout, the value of the domain cache is
1253 left unchanged (for any other kind of callout, getting as far as trying
1254 RCPT means that the domain itself is ok).
1256 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1259 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1260 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1261 session, the daemon ignores ECONNECT errors and logs others; it now
1262 ignores EPIPE as well.
1264 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1265 (quoted-printable decoding).
1267 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1268 later the small subsequent patch to fix an introduced bug.
1270 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1272 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1274 PH/08 An error is now given if message_size_limit is specified negative.
1276 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1277 to be given (somewhat) arbitrary names.
1279 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1282 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1283 miscellaneous code fixes
1285 PH/10 Added the log_reject_target ACL modifier to specify where to log
1288 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1289 hostname. This is wrong, because it relates to the incoming message (and
1290 probably the interface on which it is arriving) and not to the outgoing
1291 callout (which could be using a different interface). This has been
1292 changed to use the value of the helo_data option from the smtp transport
1293 instead - this is what is used when a message is actually being sent. If
1294 there is no remote transport (possible with a router that sets up host
1295 addresses), $smtp_active_hostname is used.
1297 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1298 tweaks were necessary in order to get it to work (see also 21 below):
1299 (a) The code assumed that strncpy() returns a negative number on buffer
1300 overflow, which isn't the case. Replaced with Exim's string_format()
1302 (b) There were several signed/unsigned issues. I just did the minimum
1303 hacking in of casts. There is scope for a larger refactoring.
1304 (c) The code used strcasecmp() which is not a standard C function.
1305 Replaced with Exim's strcmpic() function.
1306 (d) The code set only $1; it now sets $auth1 as well.
1307 (e) A simple test gave the error "authentication client didn't specify
1308 service in request". It would seem that Dovecot has changed its
1309 interface. Fortunately there's a specification; I followed it and
1310 changed what the client sends and it appears to be working now.
1312 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1315 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1316 address A is aliased to B and C, where B exists and C does not. Without
1317 -v the output is "A verified" because verification stops after a
1318 successful redirection if more than one address is generated. However,
1319 with -v the child addresses are also verified. Exim was outputting "A
1320 failed to verify" and then showing the successful verification for C,
1321 with its parentage. It now outputs "B failed to verify", showing B's
1322 parentage before showing the successful verification of C.
1324 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1325 look up a TXT record in a specific list after matching in a combined
1328 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1329 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1330 they consult the DNS. I had assumed they would set it the way they
1331 wanted; and indeed my experiments on Linux seem to show that in some
1332 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1333 To be on the safe side, however, I have now made the interface to
1334 host_find_byname() similar to host_find_bydns(), with an argument
1335 containing the DNS resolver options. The host_find_byname() function now
1336 sets these options at its start, just as host_find_bydns() does. The smtp
1337 transport options dns_qualify_single and dns_search_parents are passed to
1338 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1339 of host_find_byname() use the default settings of RES_DEFNAMES
1340 (qualify_single) but not RES_DNSRCH (search_parents).
1342 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1343 spool_read_header() do less string testing, by means of a preliminary
1344 switch on the second character of optional "-foo" lines. (This is
1345 overdue, caused by the large number of possibilities that now exist.
1346 Originally there were few.) While I was there, I also converted the
1347 str(n)cmp tests so they don't re-test the leading "-" and the first
1348 character, in the hope this might squeeze out yet more improvement.
1350 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1351 flag allowing group syntax was set by the header_syntax check but not
1352 turned off, possible causing trouble later; (2) The flag was not being
1353 set at all for the header_verify test, causing "group"-style headers to
1354 be rejected. I have now set it in this case, and also caused header_
1355 verify to ignore an empty address taken from a group. While doing this, I
1356 came across some other cases where the code for allowing group syntax
1357 while scanning a header line wasn't quite right (mostly, not resetting
1358 the flag correctly in the right place). These bugs could have caused
1359 trouble for malformed header lines. I hope it is now all correct.
1361 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1362 with the "reply" argument non-NULL. The code, however (which originally
1363 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1364 but it didn't always do it. This confused somebody who was copying the
1365 code for some other use. I have removed all the tests.
1367 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1368 feature that was used to support insecure browsers during the U.S. crypto
1369 embargo. It requires special client support, and Exim is probably the
1370 only MTA that supported it -- and would never use it because real RSA is
1371 always available. This code has been removed, because it had the bad
1372 effect of slowing Exim down by computing (never used) parameters for the
1373 RSA_EXPORT functionality.
1375 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1376 authenticator to fail if there's a tab character in the incoming data
1377 (there should never be unless someone is messing about, as it's supposed
1378 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1379 if the connection is using TLS or if the remote IP is the same as the
1380 local IP, and the "valid-client-cert option" if a client certificate has
1383 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1384 authenticators. This can be used for authorization after authentication
1385 succeeds. (In the case of plaintext, it servers for both authentication
1388 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1389 if any retry times were supplied.
1391 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1392 connection, where there is no client IP address to check. In this
1393 situation, the verify now always succeeds.
1395 PH/25 Applied John Jetmore's -Mset patch.
1397 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1399 PH/27 In a string expansion for a processed (not raw) header when multiple
1400 headers of the same name were present, leading whitespace was being
1401 removed from all of them, but trailing whitespace was being removed only
1402 from the last one. Now trailing whitespace is removed from each header
1403 before concatenation. Completely empty headers in a concatenation (as
1404 before) are ignored.
1406 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1407 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1409 PH/29 [Removed. This was a change that I later backed out, and forgot to
1410 correct the ChangeLog entry (that I had efficiently created) before
1411 committing the later change.]
1413 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1414 address errors (4xx response to RCPT) over the same connection as other
1415 messages routed to the same hosts. Such deliveries are always "forced",
1416 so retry times are not inspected. This resulted in far too many retries
1417 for the affected addresses. The effect occurred only when there were more
1418 hosts than the hosts_max_try setting in the smtp transport when it had
1419 the 4xx errors. Those hosts that it had tried were not added to the list
1420 of hosts for which the message was waiting, so if all were tried, there
1421 was no problem. Two fixes have been applied:
1423 (i) If there are any address or message errors in an SMTP delivery, none
1424 of the hosts (tried or untried) are now added to the list of hosts
1425 for which the message is waiting, so the message should not be a
1426 candidate for sending over the same connection that was used for a
1427 successful delivery of some other message. This seems entirely
1428 reasonable: after all the message is NOT "waiting for some host".
1429 This is so "obvious" that I'm not sure why it wasn't done
1430 previously. Hope I haven't missed anything, but it can't do any
1431 harm, as the worst effect is to miss an optimization.
1433 (ii) If, despite (i), such a delivery is accidentally attempted, the
1434 routing retry time is respected, so at least it doesn't keep
1435 hammering the server.
1437 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1438 in ${readsocket because some servers need this prod.
1440 PH/32 Added some extra debug output when updating a wait-xxx database.
1442 PH/33 The hint "could be header name not terminated by colon", which has been
1443 given for certain expansion errors for a long time, was not being given
1444 for the ${if def:h_colon_omitted{... case.
1446 PH/34 The spec says: "With one important exception, whenever a domain list is
1447 being scanned, $domain contains the subject domain." There was at least
1448 one case where this was not true.
1450 PH/35 The error "getsockname() failed: connection reset by peer" was being
1451 written to the panic log as well as the main log, but it isn't really
1452 panic-worthy as it just means the connection died rather early on. I have
1453 removed the panic log writing for the ECONNRESET error when getsockname()
1456 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1457 runs only) independently of the message's sender address. This meant
1458 that, if the 4xx error was in fact related to the sender, a different
1459 message to the same recipient with a different sender could confuse
1460 things. In particualar, this can happen when sending to a greylisting
1461 server, but other circumstances could also provoke similar problems.
1462 I have changed the default so that the retry time for these errors is now
1463 based a combination of the sender and recipient addresses. This change
1464 can be overridden by setting address_retry_include_sender=false in the
1467 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1468 remote server are returned as part of bounce messages. This was not
1469 happening for LMTP over a pipe (the lmtp transport), but now it is the
1470 same for both kinds of LMTP.
1472 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1473 in header lines that were in fact CNAMEs. This is no longer the case.
1475 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1476 and queue runs started by the daemon processed all messages. This has
1477 been fixed so that -R and -S can now usefully be given with -q<time>.
1479 PH/40 Import PCRE release 6.7 (fixes some bugs).
1481 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1483 PH/42 Give an error if -q is specified more than once.
1485 PH/43 Renamed the variables $interface_address and $interface_port as
1486 $received_ip_address and $received_port, to make it clear that these
1487 values apply to message reception, and not to the outgoing interface when
1488 a message is delivered. (The old names remain recognized, of course.)
1490 PH/44 There was no timeout on the connect() call when using a Unix domain
1491 socket in the ${readsocket expansion. There now is.
1493 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1494 be meaningful with "accept".
1496 SC/01 Eximstats V1.43
1497 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1499 SC/02 Eximstats V1.44
1500 Use a glob alias rather than an array ref in the generated
1501 parser. This improves both readability and performance.
1503 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1504 Collect SpamAssassin and rejection statistics.
1505 Don't display local sender or destination tables unless
1506 there is data to show.
1507 Added average volumes into the top table text output.
1509 SC/04 Eximstats V1.46
1510 Collect data on the number of addresses (recipients)
1511 as well as the number of messages.
1513 SC/05 Eximstats V1.47
1514 Added 'Message too big' to the list of mail rejection
1515 reasons (thanks to Marco Gaiarin).
1517 SC/06 Eximstats V1.48
1518 Mainlog lines which have GMT offsets and are too short to
1519 have a flag are now skipped.
1521 SC/07 Eximstats V1.49 (Alain Williams)
1522 Added the -emptyok flag.
1524 SC/08 Eximstats V1.50
1525 Fixes for obtaining the IP address from reject messages.
1527 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1528 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1529 whitesspace changes from 4.64-PH/27
1531 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1534 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1535 are found, allow negative numbers in numeric criteria)
1537 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1539 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1540 to match changes made in 4.64-PH/43
1542 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1544 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1545 "domain" is actually an IP address, and force a failure. This locks out
1546 those revolvers/nameservers that support "A-for-A" lookups, in
1547 contravention of the specifications.
1549 PH/48 When a host name was looked up from an IP address, and the subsequent
1550 forward lookup of the name timed out, the host name was left in
1551 $sender_host_name, contrary to the specification.
1553 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1554 restricted to single-key lookups, Exim was not diagnosing an error if
1555 * or *@ was used with a query-style lookup.
1557 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1559 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1560 long ago, but noone interested enough thought of it. Let's just say that
1561 the "1.1" means that there are some new functions that weren't there at
1562 some point in the past.
1564 PH/51 Error processing for expansion failure of helo_data from an smtp
1565 transport during callout processing was broken.
1567 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1568 tested/used via the -bh/-bhc/-bs options.
1570 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1571 bug, fixed in subsequent PCRE releases).
1573 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1574 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1576 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1582 SC/01 Use a glob alias rather than an array ref in eximstats generated
1583 parser. This improves both readability and performance.
1585 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1586 Don't display local sender or destination tables in eximstats unless
1587 there is data to show.
1588 Added average volumes into the eximstats top table text output.
1590 SC/03 Collect data on the number of addresses (recipients) as well
1591 as the number of messages in eximstats.
1593 TF/01 Correct an error in the documentation for the redirect router. Exim
1594 does (usually) call initgroups() when daemonizing.
1596 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1597 with consistent privilege compared to when running as a daemon.
1599 TF/03 Note in the spec that $authenticated_id is not set for local
1600 submissions from trusted users.
1602 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1603 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1605 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1606 by adding some example configuration directives to the default
1607 configuration file. A little bit of work is required to uncomment the
1608 directives and define how usernames and passwords are checked, but
1609 there is now a framework to start from.
1611 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1612 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1613 without this. I don't know how relevant this is to other LDAP libraries.
1615 PH/02 Add the verb name to the "unknown ACL verb" error.
1617 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1619 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1621 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1622 directory not expanded when it should be if an expanded home directory
1623 was set for the address (which is overridden by the transport).
1625 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1628 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1629 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1630 because it is too late at that time, and has no effect.
1632 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1633 security issue with \' (bugzilla #107). I could not use the
1634 PQescapeStringConn() function, because it needs a PGconn value as one of
1637 PH/08 When testing addresses using -bt, indicate those final addresses that
1638 are duplicates that would not cause an additional delivery. At least one
1639 person was confused, thinking that -bt output corresponded to deliveries.
1640 (Suppressing duplicates isn't a good idea as you lose the information
1641 about possibly different redirections that led to the duplicates.)
1643 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1644 systems where poll() doesn't work, in particular OS X.
1646 PH/10 Added more information to debugging output for retry time not reached.
1648 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1649 operations in malware.c.
1651 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1654 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1655 syslog_duplication set false, log lines that would normally be written
1656 both the the main log and to the reject log were not written to syslog at
1659 PH/14 In the default configuration, change the use of "message" in ACL warn
1660 statements to "add_header".
1662 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1663 not followed by a command (e.g. "seen endif").
1665 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1666 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1669 PH/17 Added extra conditions to the default value of delay_warning_condition
1673 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1674 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1675 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1678 The Auto-Submitted: and various List- headers are standardised, whereas I
1679 don't think Precedence: ever was.
1681 PH/18 Refactored debugging code in route_finduser() to show more information,
1682 in particular, the error code if getpwnam() issues one.
1684 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1685 This is apparently needed in addition to the PH/07 change above to avoid
1686 any possible encoding problems.
1688 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1689 but not after initializing Perl.
1691 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1692 output them only if debugging. By default they are written stderr,
1693 apparently, which is not desirable.
1695 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1698 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1701 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1703 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1704 authenticated or an ident call has been made. Suppress the default
1705 values for $authenticated_id and $authenticated_sender (but permit -oMai
1706 and -oMas) when testing with -bh.
1708 PH/24 Re-jigged the order of the tests in the default configuration so that the
1709 tests for valid domains and recipients precede the DNS black list and CSA
1710 tests, on the grounds that those ones are more expensive.
1712 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1713 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1714 This bug exists in every version of Exim that I still have, right back to
1717 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1718 However, an attempt to turn on case-sensitivity in a regex key by
1719 including (?-i) didn't work because the subject string was already
1720 lowercased, and the effects were non-intuitive. It turns out that a
1721 one-line patch can be used to allow (?-i) to work as expected.
1727 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1728 other effects) broke the use of negated acl sub-conditions.
1730 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1733 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1734 "Deny" causes Exim to reject the incoming connection with a 554 error.
1735 Unfortunately, if there is a major crisis, such as a disk failure,
1736 tcp-wrappers gives "deny", whereas what one would like would be some
1737 kind of temporary error. A kludge has been added to help with this.
1738 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1739 554 error is used if errno is still zero or contains ENOENT (which occurs
1740 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1743 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1745 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1746 errors. Otherwise a message that provokes a temporary error (when other
1747 messages do not) can cause a whole host to time out.
1749 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1750 addresses were routed directly to files or pipes from a redirect router.
1751 File deliveries just didn't batch; pipe deliveries might have suffered
1754 PH/06 A failure to get a lock for a hints database would erroneously always say
1755 "Failed to get write lock", even when it was really a read lock.
1757 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1758 of 0600. This has been changed to use the value of the lockfile_mode
1759 option (which defaults to 0600).
1761 PH/08 Applied small patch from the Sieve maintainer.
1763 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1764 folder from quota calculations, a direct delivery into this folder messed
1765 up the contents of the maildirsize file. This was because the regex was
1766 used only to exclude .Trash (or whatever) when the size of the mailbox
1767 was calculated. There was no check that a delivery was happening into an
1768 excluded directory. This bug has been fixed by ignoring all quota
1769 processing for deliveries into excluded directories.
1771 PH/10 Added the maildirfolder_create_regex option to appendfile.
1777 PH/01 The code for finding all the local interface addresses on a FreeBSD
1778 system running IPv6 was broken. This may well have applied to all BSD
1779 systems, as well as to others that have similar system calls. The broken
1780 code found IPv4 interfaces correctly, but gave incorrect values for the
1781 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1782 that it would not match correctly against @[] and not recognize the IPv6
1785 PH/02 The ipliteral router was not recognizing addresses of the form user@
1786 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1788 PH/03 Added disable_ipv6.
1790 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1791 decoded form, because it is most often used to construct To: headers
1792 lines in autoreplies, and the decoded form may well be syntactically
1793 invalid. However, $reply_address has leading white space removed, and all
1794 newlines turned into spaces so that the autoreply transport does not
1797 PH/05 If group was specified without a user on a router, and no group or user
1798 was specified on a transport, the group from the router was ignored.
1800 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1801 for visible compile-time settings that can be used to change these
1802 numbers, for those that want even more. Backwards compatibility with old
1803 spool files has been maintained. However, going back to a previous Exim
1804 release will lost any variables that are in spool files.
1806 PH/07 Two small changes when running in the test harness: increase delay when
1807 passing a TCP/IP connection to a new process, in case the original
1808 process has to generate a bounce, and remove special handling of
1809 127.0.0.2 (sic), which is no longer necessary.
1811 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1812 be the same on different OS.
1814 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1817 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1818 whether --show-vars was specified or not
1820 JJ/02 exipick: Added support for new ACL variable spool format introduced
1823 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1824 syntactically invalid From: or Reply-to: line, and a filter used this to
1825 generate an autoreply, and therefore failed to obtain an address for the
1826 autoreply, Exim could try to deliver to a non-existent relative file
1827 name, causing unrelated and misleading errors. What now happens is that
1828 it logs this as a hard delivery error, but does not attempt to create a
1831 PH/11 The exinext utility has a -C option for testing purposes, but although
1832 the given file was scanned by exinext itself; it wasn't being passed on
1833 when Exim was called.
1835 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1836 an end-of-file indication when reading a command response.
1838 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1839 compiled. In many other places in Exim, IPv6 addresses are always
1840 recognized, so I have changed this. It also means that IPv4 domain
1841 literals of the form [IPV4:n.n.n.n] are now always recognized.
1843 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1844 used if the router is not running as root, for example, when verifying at
1845 ACL time, or when using -bh. The debugging output from this situation was
1846 non-existent - all you got was a failure to exec. I have made two
1849 (a) Failures to set uid/gid, the current directory, or a process leader
1850 in a subprocess such as that created by queryprogram now generate
1851 suitable debugging ouput when -d is set.
1853 (b) The queryprogram router detects when it is not running as root,
1854 outputs suitable debugging information if -d is set, and then runs
1855 the subprocess without attempting to change uid/gid.
1857 PH/15 Minor change to Makefile for building test_host (undocumented testing
1860 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1861 additional section of a DNS packet that returns MX or SRV records.
1862 Instead, it always explicitly searches for A/AAAA records. This avoids
1863 major problems that occur when a DNS server includes only records of one
1864 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1865 fixed another bug: if SRV records were looked up and the corresponding
1866 address records were *not* found in the additional section, the port
1867 values from the SRV records were lost.
1869 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1870 using the correct key (the original address) when searching the retry
1871 rules in order to find which one to use for generating the retry hint.
1873 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1874 adding the default one. Similarly, if it contains a Reply-To: header, the
1875 errors_reply_to option, if set, is not used.
1877 PH/19 When calculating a retry time, Exim used to measure the "time since
1878 failure" by looking at the "first failed" field in the retry record. Now
1879 it does not use this if it is later than than the arrival time of the
1880 message. Instead it uses the arrival time. This makes for better
1881 behaviour in cases where some deliveries succeed, thus re-setting the
1882 "first failed" field. An example is a quota failure for a huge message
1883 when small messages continue to be delivered. Without this change, the
1884 "time since failure" will always be short, possible causing more frequent
1885 delivery attempts for the huge message than are intended.
1886 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1888 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1889 $1, $2, $3) because the numerical variables can be reset during some
1890 expansion items (e.g. "match"), thereby losing the authentication data.
1892 PH/21 Make -bV show the size of off_t variables so that the test suite can
1893 decide whether to run tests for quotas > 2G.
1895 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1896 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1897 filecount value is greater than 2G or if a quota value is greater than 2G
1898 on a system where the size of off_t is not greater than 4, a panic error
1901 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1902 never match. The debug and -bh output now contains an explicit error
1903 message indicating a malformed IPv4 address or mask.
1905 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1906 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1907 PH/23 above applies.
1909 PH/25 Do not write to syslog when running in the test harness. The only
1910 occasion when this arises is a failure to open the main or panic logs
1911 (for which there is an explicit test).
1913 PH/26 Added the /no_tell option to "control=freeze".
1915 PH/27 If a host name lookup failed very early in a connection, for example, if
1916 the IP address matched host_lookup and the reverse lookup yielded a name
1917 that did not have a forward lookup, an error message of the form "no IP
1918 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1919 could be logged. Now it outputs the IP address instead of "NULL".
1921 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1922 allows the sender and the authenticated sender to be set when
1923 submitting a message from within Exim. Since child_open_exim() is
1924 documented for local_scan(), the new function should be too.
1926 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1927 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1928 results in an empty string is now treated as unset.
1930 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1932 PH/31 Added sender_verify_fail logging option.
1934 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1935 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1936 tidied the source and removed it altogether.
1938 PH/33 When a queue run was abandoned because the load average was too high, a
1939 log line was always written; now it is written only if the queue_run log
1940 selector is set. In addition, the log line for abandonment now contains
1941 information about the queue run such as the pid. This is always present
1942 in "start" and "stop" lines but was omitted from the "abandon" line.
1944 PH/34 Omit spaces between a header name and the colon in the error message that
1945 is given when verify = headers_syntax fails (if there are lots of them,
1946 the message gets confusing).
1948 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1949 names, as there are now some PTR records that contain slashes. This check
1950 is only to protect against broken name servers that fall over on strange
1951 characters, so the fact that it applies to all lookups doesn't matter.
1953 PH/36 Now that the new test suite is complete, we can remove some of the
1954 special code in Exim that was needed for the old test suite. For example,
1955 sorting DNS records because real resolvers return them in an arbitrary
1956 order. The new test suite's fake resolver always returns records in the
1959 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1960 bounces) except when queue_only is set, to avoid logging races between
1961 the different processes.
1963 PH/38 Panic-die if .include specifies a non-absolute path.
1965 PH/39 A tweak to the "H" retry rule from its user.
1967 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1968 a label. They prevented compilation on older perls.
1970 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1971 a warning to be raised on newish perls.
1973 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1974 on queue. Changes to match documented behaviour of showing count of
1975 messages matching specified criteria.
1977 PH/40 Changed the default ident timeout from 30s to 5s.
1979 PH/41 Added support for the use of login_cap features, on those BSD systems
1980 that have them, for controlling the resources used by pipe deliveries.
1982 PH/42 The content-scanning code uses fopen() to create files in which to put
1983 message data. Previously it was not paying any attention to the mode of
1984 the files. Exim runs with umask(0) because the rest of the code creates
1985 files with open(), and sets the required mode explicitly. Thus, these
1986 files were ending up world-writeable. This was not a big issue, because,
1987 being within the spool directory, they were not world-accessible. I have
1988 created a function called modefopen, which takes an additional mode
1989 argument. It sets umask(777), creates the file, chmods it to the required
1990 mode, then resets the umask. All the relevant calls to fopen() in the
1991 content scanning code have been changed to use this function.
1993 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1994 to 24 hours. This avoids potential overflow problems when processing G
1995 and H retry rules. I suspect nobody ever tinkers with this value.
1997 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1999 PH/45 When the plaintext authenticator is running as a client, the server's
2000 challenges are checked to ensure they are valid base64 strings. By
2001 default, the authentication attempt is cancelled if an invalid string is
2002 received. Setting client_ignore_invalid_base64 true ignores these errors.
2003 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2004 they are received. Thus, the responses can be made to depend on the
2005 challenges. If an invalid string is ignored, an empty string is placed in
2008 PH/46 Messages that are created by the autoreply transport now contains a
2009 References: header, in accordance with RFCs 2822 and 3834.
2011 PH/47 Added authenticated_sender_force to the smtp transport.
2013 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2015 PH/49 Installed latest patch from the Sieve maintainer.
2017 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2018 was also set, the appendfile transport was unnecessarily scanning a
2019 directory of message files (e.g. for maildir delivery) to find the count
2020 of files (along with the size), even though it did not need this
2021 information. It now does the scan only if it needs to find either the
2022 size of the count of files.
2024 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2026 PH/52 Two bugs concerned with error handling when the smtp transport is
2029 (i) Exim was not creating retry information for temporary errors given
2030 for individual recipients after the DATA command when the smtp transport
2031 was used in LMTP mode. This meant that they could be retried too
2032 frequently, and not timed out correctly.
2034 (ii) Exim was setting the flag that allows error details to be returned
2035 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2036 recipients that were returned after the DATA command.
2038 PH/53 This is related to PH/52, but is more general: for any failing address,
2039 when detailed error information was permitted to be returned to the
2040 sender, but the error was temporary, then after the final timeout, only
2041 "retry timeout exceeded" was returned. Now it returns the full error as
2042 well as "retry timeout exceeded".
2044 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2045 do this, and (what is worse) MTAs that accept it.
2047 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2048 will now be deprecated.
2050 PH/56 New os.c-cygwin from the Cygwin maintainer.
2052 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2053 formats (previously only available in exim formats via -bpr, -bpru,
2054 and -bpra. Now also available in native and exiqgrep formats)
2056 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2057 with very large, slow to parse queues
2059 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2061 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2063 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2064 responses to authentication challenges, though it was showing the
2065 challenges; (ii) I've removed the CR characters from the debug output for
2068 PH/58 Allow for the insertion of a newline as well as a space when a string
2069 is turned into more than one encoded-word during RFC 2047 encoding. The
2070 Sieve code now uses this.
2072 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2073 data_4xx, lost_connection, tls_required.
2075 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2076 message was being sent as an SMTP response.
2078 PH/61 Add -l and -k options to exicyclog.
2080 PH/62 When verifying, if an address was redirected to one new address, so that
2081 verification continued, and the new address failed or deferred after
2082 having set something in $address_data, the value of $address_data was not
2083 passed back to the ACL. This was different to the case when no
2084 redirection occurred. The value is now passed back in both cases.
2086 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2087 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2088 use login_cap.h, so on its own it isn't the distinguishing feature. The
2089 new name refers directly to the setclassresources() function.
2091 PH/65 Added configuration files for NetBSD3.
2093 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2095 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2096 is preferred over IPv4.
2098 PH/68 The bounce_return_message and bounce_return_body options were not being
2099 honoured for bounces generated during the reception of non-SMTP messages.
2100 In particular, this applied to messages rejected by the ACL. This bug has
2101 been fixed. However, if bounce_return_message is true and bounce_return_
2102 body is false, the headers that are returned for a non-SMTP message
2103 include only those that have been read before the error was detected.
2104 (In the case of an ACL rejection, they have all been read.)
2106 PH/69 The HTML version of the specification is now built in a directory called
2107 spec_html instead of spec.html, because the latter looks like a path with
2108 a MIME-type, and this confuses some software.
2110 PH/70 Catch two compiler warnings in sieve.c.
2112 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2113 function verify_get_ident() calls ip_connect() to connect a socket, but
2114 if the "connect()" function timed out, ip_connect() used to close the
2115 socket. However, verify_get_ident() also closes the socket later, and in
2116 between Exim writes to the log, which may get opened at this point. When
2117 the socket was closed in ip_connect(), the log could get the same file
2118 descriptor number as the socket. This naturally causes chaos. The fix is
2119 not to close the socket in ip_connect(); the socket should be closed by
2120 the function that creates it. There was only one place in the code where
2121 this was missing, in the iplookup router, which I don't think anybody now
2122 uses, but I've fixed it anyway.
2124 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2125 well as to direct DNS lookups. Otherwise the handling of names in host
2126 lists is inconsistent and therefore confusing.
2132 PH/01 Two changes to the default runtime configuration:
2134 (1) Move the checks for relay_from_hosts and authenticated clients from
2135 after to before the (commented out) DNS black list checks.
2137 (2) Add control=submission to the relay_from_hosts and authenticated
2138 clients checks, on the grounds that messages accepted by these
2139 statements are most likely to be submissions.
2141 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2143 (1) Generate an error if the third argument for the ${prvs expansion is
2146 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2149 (3) Reset the variables that are obtained from the first argument of
2150 ${prvscheck and used in the second argument before leaving the code,
2151 because their memory is reclaimed, so using them afterwards may do
2154 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2155 one (it's much easier than Tom thought :-).
2157 (5) Because of (4), we can now allow for the use of $prvscheck_result
2158 inside the third argument.
2160 PH/03 For some reason, the default setting of PATH when running a command from
2161 a pipe transport was just "/usr/bin". I have changed it to
2164 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2165 anything to be listed in the output from -bV.
2167 PH/05 When a filter generated an autoreply, the entire To: header line was
2168 quoted in the delivery log line, like this:
2170 => >A.N.Other <ano@some.domain> <original@ddress> ...
2172 This has been changed so that it extracts the operative address. There
2173 may be more than one such address. If so, they are comma-separated, like
2176 => >ano@some.domain,ona@other.domain <original@ddress> ...
2178 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2179 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2180 not being looked up in the rDNS to get a host name, Exim was showing the
2181 IP address twice in Received: lines, even though the IP addresses were
2182 identical. For example:
2184 Received: from [1.2.3.4] (helo=[1.2.3.4])
2186 However, if the real host name was known, it was omitting the HELO data
2187 if it matched the actual IP address. This has been tidied up so that it
2188 doesn't show the same IP address twice.
2190 PH/07 When both +timestamp and +memory debugging was on, the value given by
2191 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2192 called by the debug printing, thereby overwriting the timestamp buffer.
2193 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2195 PH/08 When the original message was included in an autoreply transport, it
2196 always said "this is a copy of the message, including all the headers",
2197 even if body_only or headers_only was set. It now gives an appropriate
2200 PH/09 Applied a patch from the Sieve maintainer which:
2202 o fixes some comments
2203 o adds the (disabled) notify extension core
2204 o adds some debug output for the result of if/elsif tests
2205 o points to the current vacation draft in the documentation
2206 and documents the missing references header update
2210 o fixes a bug in processing the envelope test (when testing
2211 multiple envelope elements, the last element determinted the
2214 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2215 Electronic Mail") by including:
2217 Auto-submitted: auto-generated
2219 in the messages that it generates (bounce messages and others, such as
2220 warnings). In the case of bounce messages for non-SMTP mesages, there was
2221 also a typo: it was using "Auto_submitted" (underscore instead of
2222 hyphen). Since every message generated by Exim is necessarily in response
2223 to another message, thes have all been changed to:
2225 Auto-Submitted: auto-replied
2227 in accordance with these statements in the RFC:
2229 The auto-replied keyword:
2231 - SHOULD be used on messages sent in direct response to another
2232 message by an automatic process,
2234 - MUST NOT be used on manually-generated messages,
2236 - MAY be used on Delivery Status Notifications (DSNs) and Message
2237 Disposition Notifications (MDNs),
2239 - MUST NOT be used on messages generated by automatic or periodic
2240 processes, except for messages which are automatic responses to
2243 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2244 to the default Received: header definition.
2246 PH/12 Added log selector acl_warn_skipped (default on).
2248 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2249 variables because (a) they are in the wrong storage pool and (b) even if
2250 they were copied, it wouldn't work properly because of the caching.
2252 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2253 checking when decoding. Apparently there are clients that generate
2254 overlong encoded strings. Why am I not surprised?
2256 PH/15 If the first argument of "${if match_address" was not empty, but did not
2257 contain an "@" character, Exim crashed. Now it writes a panic log message
2258 and treats the condition as false.
2260 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2262 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2263 extension "envelope-auth". The code is finished and in agreement with
2264 other implementations, but there is no documentation so far and in fact,
2265 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2266 not changing the active code.
2268 Print executed "if" and "elsif" statements when debugging is used. This
2269 helps a great deal to understand what a filter does.
2271 Document more things not specified clearly in RFC3028. I had all this
2272 sorted out, when out of a sudden new issues came to my mind. Oops."
2274 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2277 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2278 canonical form (as documented). However, after a host name lookup from
2279 the IP address, check_host() was doing a simple string comparison with
2280 addresses acquired from the DNS when checking that the found name did
2281 have the original IP as one of its addresses. Since any found IPv6
2282 addresses are likely to be in abbreviated form, the comparison could
2283 fail. Luckily, there already exists a function for doing the comparison
2284 by converting both addresses to binary, so now that is used instead of
2285 the text comparison.
2287 PH/20 There was another similar case to PH/19, when a complete host name was
2288 given in a host list; looking up its IP address could give an abbreviated
2289 form, whereas the current host's name might or might not be abbreviated.
2290 The same fix has been applied.
2296 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2297 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2300 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2301 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2303 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2305 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2306 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2307 Also, for "Auto-Submitted", treat anything other than "no" as
2308 non-personal, in accordance with RFC 3834. (Previously it treated
2309 anything starting "auto-" as non-personal.)
2311 TF/01 The control=submission/name=... option had a problem with syntax
2312 errors if the name included a slash character. The /name= option
2313 now slurps the rest of the string, so it can include any characters
2314 but it must come last in the list of options (after /sender_retain
2317 PH/05 Some modifications to the interface to the fake nameserver for the new
2325 TK/01 Added the "success_on_redirect" address verification option. See
2326 NewStuff for rationale and an example.
2328 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2330 PH/02 Patch to exigrep to allow it to work on syslog lines.
2332 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2333 fread() to skip over the body file's header line, because in Cygwin the
2334 header line is locked and is inaccessible.
2336 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2337 co-exist for some time) to make it clear that it is the Exim ID that is
2338 referenced, not the Message-ID: header line.
2340 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2341 string_format() function, because snprintf() does not exist on all
2344 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2345 ${stat: expansion item.
2347 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2348 protocol synchronization error", to keep the pedants happy.
2350 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2351 well as for IRIX systems, when gcc is being used. See the host.c source
2354 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2356 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2359 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2360 transport and to the smtp transport in LMTP mode.
2362 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2364 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2365 run a filter in a subprocess. This could lead to confusion in subsequent
2366 lookups in the parent process. There should also be a search_tidyup() at
2367 the end of the subprocess.
2369 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2370 only if the host matched helo_try_verify_hosts, which caused the
2371 verification to occur when the EHLO/HELO command was issued. The ACL just
2372 tested the remembered result. Now, if a previous verification attempt has
2373 not happened, "verify = helo" does it there and then.
2375 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2377 TK/03 Fix log output including CR from clamd.
2379 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2380 exist provoked a memory error which could cause a segfault.
2382 PH/15 Installed PCRE 6.2
2384 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2386 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2387 of the problem. Specifically, suggested +O2 rather than +O1 for the
2390 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2392 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2393 record, the port was not correctly logged when the outgoing_port log
2394 selector was set (it logged the transort's default port).
2396 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2397 fallback_hosts, and "hosts" in the smtp transport.
2399 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2400 host errors such as "Connection refused".
2402 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2403 authentication with radiusclient 0.4.9:
2405 - Error returned from rc_read_config was caught wrongly
2406 - Username/password not passed on to radius server due to wrong length.
2408 The presumption is that some radiusclient API changes for 4.51/PH/17
2409 were not taken care of correctly. The code is still untested by me (my
2410 Linux distribution still has 0.3.2 of radiusclient), but it was
2411 contributed by a Radius user.
2413 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2414 expanding the "port" option of the smtp transport.
2416 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2417 while reading a MIME header. Thanks to Tom Hughes for a patch.
2419 PH/24 Include config.h inside local_scan.h so that configuration settings are
2422 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2423 that in an ACL for RCPT (for example), you can examine exactly what was
2426 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2427 commands, but it was not correctly comparing the address with the actual
2428 client host address. Thus, it would show the EHLO address in Received:
2429 header lines when this was not necessary.
2431 PH/27 Added the % operator to ${eval:}.
2433 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2434 it should be ignoring failures (because with -C, for example, it has lost
2435 privilege). It wasn't ignoring creation failures other than "already
2438 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2441 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2442 because an input line was too long, either on its own, or by virtue of
2443 too many continuations, the temporary file was not being removed, and the
2444 return code was incorrect.
2446 PH/31 Missing "BOOL" in function definition in filtertest.c.
2448 PH/32 Applied Sieve patches from the maintainer.
2450 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2452 PH/33 Added "verify = not_blind".
2454 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2455 Local/Makefile (with some defaults set). These are used in built scripts
2456 such as exicyclog, but they have never been used in the exim_install
2457 script (though there are many overriding facilities there). I have
2458 arranged that the exim_install script now takes note of these two
2461 PH/35 Installed configuration files for Dragonfly.
2463 PH/36 When a locally submitted message by a trusted user did not contain a
2464 From: header, and the sender address was obtained from -f or from an SMTP
2465 MAIL command, and the trusted user did not use -F to supply a sender
2466 name, $originator_name was incorrectly used when constructing a From:
2467 header. Furthermore, $originator_name was used for submission mode
2468 messages from external hosts without From: headers in a similar way,
2469 which is clearly wrong.
2471 PH/37 Added control=suppress_local_fixups.
2473 PH/38 When log_selector = +received_sender was set, and the addition of the
2474 sender made the log line's construction buffer exactly full, or one byte
2475 less than full, an overflow happened when the terminating "\n" was
2478 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2479 when the result of a list match is failure because a DNS lookup failed.
2481 PH/40 RM_COMMAND is now used in the building process.
2483 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2484 the "build-* directories that it finds.
2486 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2487 address in a domain literal was a prefix of an interface address.
2489 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2490 when verifying a sender address, unless rewrite_headers is false.
2492 PH/44 Wrote a long comment about why errors_to addresses are verified as
2493 recipients, not senders.
2495 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2496 the ratelimit ACL was added.
2498 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2500 PH/46 Added extra information about PostgreSQL errors to the error string.
2502 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2503 avoiding the need to install special zones in a real server. This is
2504 backwards compatible; if it can't find the fake resolver, it drops back.
2505 Thus, both old and new test suites can be run.
2507 TF/02 Added util/ratelimit.pl
2509 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2510 clock is set back in time.
2512 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2513 Candler <B.Candler@pobox.com>.
2515 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2516 OK for addresses that are the result of redirections.
2518 PH/48 A number of further additions for the benefit of the new test suite,
2519 including a fake gethostbyname() that interfaces to the fake DNS resolver
2522 TF/06 The fix for widen_domains has also been applied to qualify_single and
2523 search_parents which are the other dnslookup options that can cause
2526 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2529 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2531 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2532 service" if it encountered an MX record with an empty target hostname.
2533 The message is now "an MX or SRV record indicated no SMTP service".
2535 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2536 if the DNS of the sending site is misconfigured. This is quite a
2537 common situation. This change restores the behaviour of treating a
2538 helo verification defer as a failure.
2540 PH/51 If self=fail was set on a router, the bounce message did not include the
2541 actual error message.
2547 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2549 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2550 command itself ended in error, the underlying message about the transport
2551 filter timeout was being overwritten with the pipe command error. Now the
2552 underlying error message should be appended to the second error message.
2554 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2556 PH/02 Reduce the amount of output that "make" produces by default. Full output
2557 can still be requested.
2559 PH/03 The warning log line about a condition test deferring for a "warn" verb
2560 was being output only once per connection, rather than after each
2561 occurrence (because it was using the same function as for successful
2562 "warn" verbs). This seems wrong, so I have changed it.
2564 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2565 it should not have, which might have caused a crash in the right
2566 circumstances, but probably never did.
2568 PH/04 Installed a modified version of Tony Finch's patch to make submission
2569 mode fix the return path as well as the Sender: header line, and to
2570 add a /name= option so that you can make the user's friendly name appear
2573 TF/03 Added the control = fakedefer ACL modifier.
2575 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2576 Mark Lowes for thorough testing.
2578 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2580 TK/03 Merged latest SRS patch from Miles Wilton.
2582 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2583 with the definition in sysexits.h (which is #included earlier).
2584 Fortunately, Exim does not actually use EX_OK. The code used to try to
2585 preserve the sysexits.h value, by assumimg that macro definitions were
2586 scanned for macro replacements. I have been disabused of this notion,
2587 so now the code just undefines EX_OK before #including unistd.h.
2589 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2590 in the smtp transport. When a block could not be written in a single
2591 write() function, the timeout was being re-applied to each part-write.
2592 This seems wrong - if the receiver was accepting one byte at a time it
2593 would take for ever. The timeout is now adjusted when this happens. It
2594 doesn't have to be particularly precise.
2596 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2597 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2599 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2600 without a domain if the check to <postmaster@domain> fails.
2602 SC/01 Eximstats: added -xls and the ability to specify output files
2603 (patch written by Frank Heydlauf).
2605 SC/02 Eximstats: use FileHandles for outputing results.
2607 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2609 SC/04 Eximstats: fixed display of large numbers with -nvr option
2611 SC/05 Eximstats: fixed merging of reports with empty tables.
2613 SC/06 Eximstats: added the -include_original_destination flag
2615 SC/07 Eximstats: removed tabs and trailing whitespace.
2617 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2619 TK/06 MBOX spool code: Add real "From " MBOX separator line
2620 so the .eml file is really in mbox format (even though
2621 most programs do not really care). Patch from Alex Miller.
2623 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2624 The latter is generated from $received_to and is only set if the
2625 message has one envelope recipient. SA can use these headers,
2626 obviously out-of-the-box. Patch from Alex Miller.
2628 PH/08 The ${def test on a variable was returning false if the variable's
2629 value was "0", contrary to what the specification has always said!
2630 The result should be true unless the variable is empty.
2632 PH/09 The syntax error of a character other than { following "${if
2633 def:variable_name" (after optional whitespace) was not being diagnosed.
2634 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2635 accidental colon was present, for example, could give incorrect results.
2637 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2638 result is used (not including appendfile, where other changes are about
2641 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2642 This involved changing a lot of size variables from int to off_t. It
2643 should work with maildirs and everything.
2645 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2646 spamd dying while we are connected to it.
2648 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2651 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2652 function for BDB 4.3.
2654 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2656 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2657 the sources for PCRE have been re-arranged and more files are now
2660 PH/15 The code I had for printing potentially long long variables in PH/11
2661 above was not the best (it lost precision). The length of off_t variables
2662 is now inspected at build time, and an appropriate printing format (%ld
2663 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2664 to be "long long int" or "long int". This is needed for the internal
2665 formatting function string_vformat().
2667 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2668 the configuration file to be ":syslog", then the script "guesses" where
2669 the logs files are, rather than using the compiled in default. In our
2670 case the guess is not the same as the compiled default, so the script
2671 suddenly stopped working when I started to use syslog. The patch checks
2672 to see if log_file_path is "". If so, it attempts to read it from exim
2673 with no configuration file to get the compiled in version, before it
2674 falls back to the previous guessing code."
2676 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2677 implementing BATV in an Exim configuration. See NewStuff for the gory
2680 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2681 Makefile that are specific to HP-UX.
2683 PH/18 If the "use_postmaster" option was set for a recipient callout together
2684 with the "random" option, the postmaster address was used as the MAIL
2685 FROM address for the random test, but not for the subsequent recipient
2686 test. It is now used for both.
2688 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2689 patch removes a few documentation additions to RFC 3028, because the
2690 latest draft now contains them. It adds the new en;ascii-case comparator
2691 and a new error check for 8bit text in MIME parts. Comparator and
2692 require names are now matched exactly. I enabled the subaddress
2693 extension, but it is not well tested yet (read: it works for me)."
2695 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2696 rework some of the code of TK/09 above to avoid the hardwired use of
2697 "%lld" and "long long". Replaced the call to snprintf() with a call to
2700 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2701 records point to non-existent hosts", "retry timeout exceeded", and
2702 "retry time not reached for any host after a long failure period".
2704 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2705 experimental DomainKeys support:
2707 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2708 (2) On an error such as an illegally used "control", the wrong name for
2709 the control was given.
2711 These problems did NOT occur unless DomainKeys support was compiled.
2713 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2715 PH/24 Added ${if match_ip condition.
2717 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2718 databases so that it will be absolutely obvious if a crash occurs in the
2719 DB library. This is a regular occurrence (often caused by mis-matched
2722 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2723 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2724 picked up on a user's system that detects such things. There doesn't seem
2725 to be a gcc warning option for this - only an attribute that has to be
2726 put on the function's prototype. It seems that in Fedora Core 4 they have
2727 set this on a number of new functions. No doubt there will be more in due
2730 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2731 specify a transport. However, if an address that was verified by such a
2732 router was the subject of a callout, Exim crashed because it tried to
2733 read the rcpt_include_affixes from the non-existent transport. Now it
2734 just assumes that the setting of that option is false. This bug was
2735 introduced by 4.51/PH/31.
2737 PH/28 Changed -d+all to exclude +memory, because that information is very
2738 rarely of interest, but it makes the output a lot bigger. People tend to
2739 do -d+all out of habit.
2741 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2742 code in os-type was giving problems when libc.so lives in lib64, like on
2745 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2746 aren't the modern standard, and it seems that some systems' include files
2747 don't always have them. Exim was already checking for some of the newer
2748 ones like T_AAAA, and defining it itself. I've added checks for all the
2749 record types that Exim uses.
2751 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2752 not automatically generating a new one, as it is supposed to. This
2753 prevented TLS from working. If the file did exist, but contained invalid
2754 data, a new version was generated, as expected. It was only the case of a
2755 non-existent file that was broken.
2757 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2758 with a change in libdomainkeys > 0.64.
2760 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2761 from DNS. If the selector record carries the flag, it now has
2762 precedence over the domain-wide flag.
2764 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2766 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2767 the use of a port name that isn't defined in /etc/services) occurred, the
2768 message was deferred as in a normal delivery, and thus remained on the
2769 spool, instead of being failed because of the mua_wrapper setting. This
2770 is now fixed, and I tidied up some of the mua_wrapper messages at the
2773 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2774 the messages in a hash of arrays rather than using individual hashes.
2775 This is a bit cleaner and results in dramatic memory savings, albeit
2776 at a slight CPU cost.
2778 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2779 as requested by Marc Sherman.
2781 SC/10 Eximstats: added histograms for user specified patterns as requested
2784 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2786 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2787 fopen() in the content-scanning modules that did not already have it.
2793 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2794 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2796 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2798 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2800 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2801 to test Sieve filters that use "vacation".
2803 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2804 that changes the way the GnuTLS parameters are stored in the cache file.
2805 The new format can be generated externally. For backward compatibility,
2806 if the data in the cache doesn't make sense, Exim assumes it has read an
2807 old-format file, and it generates new data and writes a new file. This
2808 means that you can't go back to an older release without removing the
2811 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2812 work if there are any delivery delays because "one_time" forces the
2813 parent to be marked "delivered", so its unseen clone is never tried
2814 again. For this reason, Exim now forbids the simultaneous setting of
2817 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2818 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2819 entry if you want to know the details. The fix, however, neglected to
2820 consider the case when local delivery batching is involved. The test for
2821 "previously delivered" was not happening when checking to see if an
2822 address could be batched with a previous (undelivered) one; under
2823 certain circumstances this could lead to multiple deliveries to the same
2826 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2827 in its include files, and this causes problems building Exim.
2829 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2830 header_syntax) but Exim was just ignoring anything given after a slash.
2831 In particular, this caused confusion with an attempt to use "verify =
2832 reverse_host_lookup/defer_ok". An error is now given when options are
2833 supplied for verify items that do not have them. (Maybe reverse_host_
2834 lookup should have a defer_ok option, but that's a different point.)
2836 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2837 defined by RFC 821) to 2048, because there were problems with some AUTH
2838 commands, and RFC 1869 says the size should be increased for extended
2839 SMTP commands that take arguments.
2841 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2844 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2845 "unknown" error; now it says that the functionality isn't in the binary.
2847 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2848 an address' error message when a string expansion fails (syntax or
2849 whatever). Otherwise the password may appear in the log. Following change
2850 PH/42 below, there is no longer a chance of it appearing in a bounce
2853 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2855 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2856 of its addresses was ever tried. (Bugzilla bug #2.)
2858 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2859 the result incorrectly in the debug output. (It correctly added a newline
2860 to what was transported.)
2862 TF/01 Added $received_time.
2864 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2865 commented out examples of how to interface to a virus scanner and to
2866 SpamAssassin. Also added commented examples of av_scanner and
2867 spamd_address settings.
2869 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2870 and controls are allowed in which ACLs. There were a couple of minor
2871 errors. Some of the entries in the conditions table (which is a table of
2872 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2873 negation of where the condition IS allowed.
2875 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2877 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2878 header file does not have a version number, so I've had to invent a new
2879 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2880 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2881 radiusclient), but it was contributed by a Radius user.
2883 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2884 files or directories, for OpenSSL.
2886 PH/19 When an Exim process that is running as root has to create an Exim log
2887 file, it does so in a subprocess that runs as exim:exim so as to get the
2888 ownership right at creation (otherwise, other Exim processes might see
2889 the file with the wrong ownership). There was no test for failure of this
2890 fork() call, which would lead to the process getting stuck as it waited
2891 for a non-existent subprocess. Forks do occasionally fail when resources
2892 run out. I reviewed all the other calls to fork(); they all seem to check
2895 PH/20 When checking for unexpected SMTP input at connect time (before writing
2896 the banner), Exim was not dealing correctly with a non-positive return
2897 from the read() function. If the client had disconnected by this time,
2898 the result was a log entry for a synchronization error with an empty
2899 string after "input=" when read() returned zero. If read() returned -1
2900 (an event I could not check), uninitialized data bytes were printed.
2901 There were reports of junk text (parts of files, etc) appearing after
2904 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2906 PH/22 Added support for macro redefinition, and (re)definition in between
2907 driver and ACL definitions.
2909 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2910 forgetting to use the resulting value; it was using the unexpanded value.
2912 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2913 hadn't been configured. The fix is from Juergen Kreileder, who
2914 understands it better than I do:
2916 "Here's what I see happening with three configured cyrus_sasl
2917 authenticators configured (plain, login, cram-md5):
2919 On startup auth_cyrus_sasl_init() gets called for each of these.
2920 This means three calls to sasl_listmech() without a specified mech_list.
2921 => SASL tests which mechs of all available mechs actually work
2922 => three warnings about OTP not working
2923 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2925 With the patch, sasl_listmech() also gets called three times. But now
2926 SASL's mech_list option is set to the server_mech specified in the the
2927 authenticator. Or in other words, the answer from sasl_listmech()
2928 gets limited to just the mech you're testing for (which is different
2930 => the return list contains just 'plain' or 'login', 'cram-md5' or
2931 nothing depending on the value of ob->server_mech.
2933 I've just tested the patch: Authentication still works fine,
2934 unavailable mechs specified in the exim configuration are still
2935 caught, and the auth.log warnings about OTP are gone."
2937 PH/25 When debugging is enabled, the contents of the command line are added
2938 to the debugging output, even when log_selector=+arguments is not
2941 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2942 answer is "GNU", and only if the return is "GNU/something" is the answer
2945 PH/27 $acl_verify_message is now set immediately after the failure of a
2946 verification in an ACL, and so is available in subsequent modifiers. In
2947 particular, the message can be preserved by coding like this:
2949 warn !verify = sender
2950 set acl_m0 = $acl_verify_message
2952 Previously, $acl_verify_message was set only while expanding "message"
2953 and "log_message" when a very denied access.
2955 PH/28 Modified OS/os.c-Linux with
2957 -#ifndef OS_LOAD_AVERAGE
2958 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2960 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2961 nomenclature these days.)
2963 PH/29 Installed patch from the Sieve maintainer that adds the options
2964 sieve_useraddress and sieve_subaddress to the redirect router.
2966 PH/30 In these circumstances:
2967 . Two addresses routed to the same list of hosts;
2968 . First host does not offer TLS;
2969 . First host accepts first address;
2970 . First host gives temporary error to second address;
2971 . Second host offers TLS and a TLS session is established;
2972 . Second host accepts second address.
2973 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2974 and peerdn, if requested) that were in fact used only for the second
2977 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2978 attention to any local part prefix or suffix that was matched by the
2979 router that accepted the address. It now behaves in the same way as it
2980 does for delivery: the affixes are removed from the local part unless
2981 rcpt_include_affixes is set on the transport.
2983 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2984 timeout during the DATA phase of an incoming message.
2986 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2987 applied a patch sanctioned by the Sieve maintainer.
2989 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2990 the uid or gid is negative. A case of a negative gid caused this to be
2991 noticed. The fix allows for either to be negative.
2993 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2994 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2997 PH/36 The change PH/12 above was broken. Fixed it.
2999 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3000 the grounds that routing the same address twice would always produce the
3001 same answer. This might have been true once, but it is certainly no
3002 longer true now. Routing a child address may depend on the previous
3003 routing that produced that child. Some complicated redirection strategies
3004 went wrong when messages had multiple recipients, and made Exim's
3005 behaviour dependent on the order in which the addresses were given.
3007 I have moved the duplicate checking until after the routing is complete.
3008 Exim scans the addresses that are assigned to local and remote
3009 transports, and removes any duplicates. This means that more work will be
3010 done, as duplicates will always all be routed, but duplicates are
3011 presumably rare, so I don't expect this is of any significance.
3013 For deliveries to pipes, files, and autoreplies, the duplicate checking
3014 still happens during the routing process, since they are not going to be
3017 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3018 It corrects a timeout issue with spamd. This is Ian's comment: "The
3019 background is that sometimes spamd either never reads data from a
3020 connection it has accepted, or it never writes response data. The exiscan
3021 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3022 blindly assumes that writes won't block so it may never time out."
3024 PH/39 Allow G after quota size as well as K and M.
3026 PH/40 The value set for $authenticated_id in an authenticator may not contain
3027 binary zeroes or newlines because the value is written to log lines and
3028 to spool files. There was no check on this. Now the value is run through
3029 the string_printing() function so that such characters are converted to
3030 printable escape sequences.
3032 PH/41 $message_linecount is a new variable that contains the total number of
3033 lines in the message. Compare $body_linecount, which is the count for the
3036 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3037 bounce and delay warning messages, except in certain special cases, which
3040 (a) An SMTP error message from a remote host;
3041 (b) A message specified in a :fail: redirection;
3042 (c) A message specified in a "fail" command in a system filter;
3043 (d) A message specified in a FAIL return from the queryprogram router;
3044 (e) A message specified by the cannot_route_message router option.
3046 In these cases only, Exim does include the error details in bounce and
3047 warning messages. There are also a few cases where bland messages such
3048 as "unrouteable address" or "local delivery error" are given.
3050 PH/43 $value is now also set for the "else" part of a ${run expansion.
3052 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3053 being worked on, but at least Exim now implements the latest version to
3056 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3057 process to complete was treated as a delivery failure, a timeout while
3058 writing the message to the pipe was logged, but erroneously treated as a
3059 successful delivery. Such timeouts include transport filter timeouts. For
3060 consistency with the overall process timeout, these timeouts are now
3061 treated as errors, giving rise to delivery failures by default. However,
3062 there is now a new Boolean option for the pipe transport called
3063 timeout_defer, which, if set TRUE, converts the failures into defers for
3064 both kinds of timeout. A transport filter timeout is now identified in
3067 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3068 systems where "make" and "gmake" are different, calling "gmake" at top
3069 level broke things. I've arranged for the value of $(MAKE) to be passed
3070 from the Makefile to this script so that it can call the same version of
3074 A note about Exim versions 4.44 and 4.50
3075 ----------------------------------------
3077 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3078 changes of various kinds. As a consequence, a big documentation update was
3079 needed. This delayed the release for rather longer than seemed good, especially
3080 in the light of a couple of (minor) security issues. Therefore, the changes
3081 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3082 release. So 4.44 and 4.50 are in effect two different branches that both start
3085 I have left the 4.50 change log unchanged; it contains all the changes since
3086 4.43. The change log for 4.44 is below; many of its items are identical to
3087 those for 4.50. This seems to be the most sensible way to preserve the
3088 historical information.
3094 1. Minor wording change to the doc/README.SIEVE file.
3096 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3097 computation of the current number of files was incorrect.
3099 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3100 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3103 4. Give more explanation in the error message when the command for a transport
3104 filter fails to execute.
3106 5. There are several places where Exim runs a non-Exim command in a
3107 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3108 was being done only for the command run by the queryprogram router. It is
3109 now done for all such subprocesses. The other cases are: ${run, transport
3110 filters, and the commands run by the lmtp and pipe transports.
3112 6. Added CONFIGURE_GROUP build-time option.
3114 7. Some older OS have a limit of 256 on the maximum number of file
3115 descriptors. Exim was using setrlimit() to set 1000 as a large value
3116 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3117 systems. I've change it so that if it can't get 1000, it tries for 256.
3119 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3120 was an oversight, and furthermore, ever since the addition of extra
3121 controls (e.g. 4.43/32), the checks on when to allow different forms of
3122 "control" were broken. There should now be diagnostics for all cases when a
3123 control that does not make sense is encountered.
3125 9. Added the /retain_sender option to "control=submission".
3127 10. $recipients is now available in the predata ACL (oversight).
3129 11. Tidy the search cache before the fork to do a delivery from a message
3130 received from the command line. Otherwise the child will trigger a lookup
3131 failure and thereby defer the delivery if it tries to use (for example) a
3132 cached ldap connection that the parent has called unbind on.
3134 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3135 of $address_data from the recipient verification was clobbered by the
3136 sender verification.
3138 13. The value of address_data from a sender verification is now available in
3139 $sender_address_data in subsequent conditions in the ACL statement.
3141 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3143 15. Added a new option "connect=<time>" to callout options, to set a different
3146 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3147 was its contents. (It was OK if the option was not defined at all.)
3149 17. A "Completed" log line is now written for messages that are removed from
3150 the spool by the -Mrm option.
3152 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3153 information about exactly what failed.
3155 19. Added -dd to debug only the daemon process.
3157 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3158 handles timeouts, both on the server side and network timeouts. Renamed the
3159 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3161 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3162 It is now set to "smtps".
3164 22. $host_address is now set to the target address during the checking of
3165 ignore_target_hosts.
3167 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3168 being passed; this would have caused $sender_host_name to have been used if
3169 matching the list had actually called for a host name (not very likely,
3170 since this list is usually IP addresses). A host name is now passed as
3173 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3174 code that specifies a non-restarting handler (typically sigaction() in
3175 modern systems) in an attempt to fix a rare and obscure crash bug.
3177 25. Narrowed the window for a race in the daemon that could cause it to ignore
3178 SIGCHLD signals. This is not a major problem, because they are used only to
3179 wake it up if nothing else does.
3181 26. A malformed maildirsize file could cause Exim to calculate negative values
3182 for the mailbox size or file count. Odd effects could occur as a result.
3183 The maildirsize information is now recalculated if the size or filecount
3186 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3187 support for a long time. Removed HAVE_SYS_VFS_H.
3189 28. Installed the latest version of exipick from John Jetmore.
3191 29. In an address list, if the pattern was not a regular expression, an empty
3192 subject address (from a bounce message) matched only if the pattern was an
3193 empty string. Non-empty patterns were not even tested. This was the wrong
3194 because it is perfectly reasonable to use an empty address as part of a
3195 database query. An empty address is now tested by patterns that are
3196 lookups. However, all the other forms of pattern expect the subject to
3197 contain a local part and a domain, and therefore, for them, an empty
3198 address still always fails if the pattern is not itself empty.
3200 30. Exim went into a mad DNS loop when attempting to do a callout where the
3201 host was specified on an smtp transport, and looking it up yielded more
3202 than one IP address.
3204 31. Re-factored the code for checking spool and log partition space into a
3205 function that finds that data and another that does the check. The former
3206 is then used to implement four new variables: $spool_space, $log_space,
3207 $spool_inodes, and $log_inodes.
3209 32. The RFC2047 encoding function was originally intended for short strings
3210 such as real names; it was not keeping to the 75-character limit for
3211 encoded words that the RFC imposes. It now respects the limit, and
3212 generates multiple encoded words if necessary. To be on the safe side, I
3213 have increased the buffer size for the ${rfc2047: expansion operator from
3216 33. It is now permitted to omit both strings after an "if" condition; if the
3217 condition is true, the result is "true". As before, when the second string
3218 is omitted, a false condition yields an empty string. This makes it less
3219 cumbersome to write custom ACL and router conditions.
3221 34. Failure to deliver a bounce message always caused it to be frozen, even if
3222 there was an errors_to setting on the router. The errors_to setting is now
3225 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3226 canonical form (fully expanded) before being placed in
3227 $sender_host_address.
3229 36. The table in the code that translates DNS record types into text (T_A to
3230 "A" for instance) was missing entries for NS and CNAME. It is just possible
3231 that this could have caused confusion if both these types were looked up
3232 for the same domain, because the text type is used as part of Exim's
3233 per-process caching. But the chance of anyone hitting this buglet seems
3236 37. The dnsdb lookup has been extended in a number of ways.
3238 (1) There is a new type, "zns", which walks up the domain tree until it
3239 finds some nameserver records. It should be used with care.
3241 (2) There is a new type, "mxh", which is like "mx" except that it returns
3242 just the host names, not the priorities.
3244 (3) It is now possible to give a list of domains (or IP addresses) to be
3245 looked up. The behaviour when one of the lookups defers can be
3246 controlled by a keyword.
3248 (4) It is now possible to specify the separator character for use when
3249 multiple records are returned.
3251 38. The dnslists ACL condition has been extended: it is now possible to supply
3252 a list of IP addresses and/or domains to be looked up in a particular DNS
3255 39. Added log_selector=+queue_time_overall.
3257 40. When running the queue in the test harness, wait just a tad after forking a
3258 delivery process, to get repeatability of debugging output.
3260 41. Include certificate and key file names in error message when GnuTLS fails
3261 to set them up, because the GnuTLS error message doesn't include the name
3262 of the failing file when there is a problem reading it.
3264 42. Allow both -bf and -bF in the same test run.
3266 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3268 44. The "Exiscan patch" is now merged into the mainline Exim source.
3270 45. Sometimes the final signoff response after QUIT could fail to get
3271 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3272 < 0 before doing a fflush(). This bug looks as though it goes back to the
3273 introduction of TLS in release 3.20, but "sometimes" must have been rare
3274 because the tests only now provoked it.
3276 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3277 (this can affect the format of dates).
3279 47. exim_tidydb, when checking for the continued existence of a message for
3280 which it has found a message-specific retry record, was not finding
3281 messages that were in split spool directories. Consequently, it was
3282 deleting retry records that should have stayed in existence.
3284 48. Steve fixed some bugs in eximstats.
3286 49. The SPA authentication driver was not abandoning authentication and moving
3287 on to the next authenticator when an expansion was forced to fail,
3288 contradicting the general specification for all authenticators. Instead it
3289 was generating a temporary error. It now behaves as specified.
3291 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3292 (the order specifies the preference for clients). The order is now AES256,
3293 AES128, 3DES, ARCFOUR128.
3295 51. Small patch to Sieve code - explicitly set From: when generating an
3298 52. Exim crashed if a remote delivery caused a very long error message to be
3299 recorded - for instance if somebody sent an entire SpamAssassin report back
3300 as a large number of 550 error lines. This bug was coincidentally fixed by
3301 increasing the size of one of Exim's internal buffers (big_buffer) that
3302 happened as part of the Exiscan merge. However, to be on the safe side, I
3303 have made the code more robust (and fixed the comments that describe what
3306 53. Now that there can be additional text after "Completed" in log lines (if
3307 the queue_time_overall log selector is set), a one-byte patch to exigrep
3308 was needed to allow it to recognize "Completed" as not the last thing in
3311 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3312 patch that reportedly fixes this has been added. I am not expert enough to
3313 create a test for it. This is what the patch creator wrote:
3315 "I found a little strange behaviour of ldap code when working with
3316 Windows 2003 AD Domain, where users was placed in more than one
3317 Organization Units. When I tried to give exim partial DN, the exit code
3318 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3319 But simultaneously result of request was absolutely normal ldap result,
3320 so I produce this patch..."
3322 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3323 REFERENCE, so I have modified the code to exclude the patch when that macro
3326 55. Some experimental protocols are using DNS PTR records for new purposes. The
3327 keys for these records are domain names, not reversed IP addresses. The
3328 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3329 leaves it alone. Component reversal etc. now happens only for IP addresses.
3332 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3334 57. Double the size of the debug message buffer (to 2048) so that more of very
3335 long debug lines gets shown.
3337 58. The exicyclog utility now does better if the number of log files to keep
3338 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3340 59. Two changes related to the smtp_active_hostname option:
3342 (1) $smtp_active_hostname is now available as a variable.
3343 (2) The default for smtp_banner uses $smtp_active_hostname instead
3344 of $primary_hostname.
3346 60. The host_aton() function is supposed to be passed a string that is known
3347 to be a valid IP address. However, in the case of IPv6 addresses, it was
3348 not checking this. This is a hostage to fortune. Exim now panics and dies
3349 if the condition is not met. A case was found where this could be provoked
3350 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3351 components; fortuitously, this particular loophole had already been fixed
3352 by change 4.50/55 above.
3354 If there are any other similar loopholes, the new check in host_aton()
3355 itself should stop them being exploited. The report I received stated that
3356 data on the command line could provoke the exploit when Exim was running as
3357 exim, but did not say which command line option was involved. All I could
3358 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3359 running as the user.
3362 61. There was a buffer overflow vulnerability in the SPA authentication code
3363 (which came originally from the Samba project). I have added a test to the
3364 spa_base64_to_bits() function which I hope fixes it.
3367 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3368 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3370 63. The daemon start-up calls getloadavg() while still root for those OS that
3371 need the first call to be done as root, but it missed one case: when
3372 deliver_queue_load_max is set with deliver_drop_privilege. This is
3373 necessary for the benefit of the queue runner, because there is no re-exec
3374 when deliver_drop_privilege is set.
3376 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3377 This has been fixed.
3379 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3380 was in use, was not putting the data itself into the right store pool;
3381 consequently, it could be overwritten for a subsequent message in the same
3382 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3385 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3387 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3388 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3389 the function were treating the return as a boolean value, which happened to
3390 work because 0=false and not-0=true, but is not correct code.
3392 68. The host_aton() function was not handling scoped IPv6 addresses (those
3393 with, for example, "%eth0" on the end) correctly.
3395 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3396 negated items (that is, ~something) in unsigned ints. Some compilers
3397 apparently mutter when there is no cast.
3399 70. If an address verification called from an ACL failed, and did not produce a
3400 user-specific message (i.e. there was only a "system" message), nothing was
3401 put in $acl_verify_message. In this situation, it now puts the system
3404 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3405 session; change 4.31/43 added the unwanted input to the log line - except
3406 that it did not do this in the start of session case. It now does.
3408 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3409 This is wrong and can cause the other end to generate a synchronization
3410 error if it is another Exim or anything else that does the synchronization
3411 check. A QUIT command is no longer sent after a timeout.
3413 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3414 during host lookups.
3416 74. The defer_ok option of callout verification was not working if it was used
3417 when verifying addresses in header lines, that is, for this case:
3419 verify = header_sender/callout=defer_ok
3421 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3422 those file descriptors could be used for SMTP connections. If anything
3423 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3424 could be sent to the SMTP client, causing chaos. The daemon now opens
3425 stdin, stdout, and stderr to /dev/null when it puts itself into the
3428 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3429 log by default. The user can override this with suitable Perl magic.
3431 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3432 the log message when discard triggers, was not working for the DATA ACL or
3433 for the non-SMTP ACL.
3435 78. Error message wording change in sieve.c.
3437 79. If smtp_accept_max_per_host was set, the number of connections could be
3438 restricted to fewer than expected, because the daemon was trying to set up
3439 a new connection before checking whether the processes handling previous
3440 connections had finished. The check for completed processes is now done
3441 earlier. On busy systems, this bug wouldn't be noticed because something
3442 else would have woken the daemon, and it would have reaped the completed
3445 80. If a message was submitted locally by a user whose login name contained one
3446 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3447 It caused a spool format error. I have fixed the spool reading code. A
3448 related problem was that the "from" clause in the Received: line became
3449 illegal because of the space(s). It is now covered by ${quote_local_part.
3451 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3454 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3456 83. Patch from Sieve maintainer to fix unterminated string problem in
3457 "vacation" handling.
3459 84. Some minor changes to the Linux configuration files to help with other
3460 OS variants using glibc.
3462 85. One more patch for Sieve to update vacation handling to latest spec.
3465 ----------------------------------------------------
3466 See the note above about the 4.44 and 4.50 releases.
3467 ----------------------------------------------------
3473 1. Change 4.43/35 introduced a bug that caused file counts to be
3474 incorrectly computed when quota_filecount was set in an appendfile
3477 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3478 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3481 3. Give more explanation in the error message when the command for a transport
3482 filter fails to execute.
3484 4. There are several places where Exim runs a non-Exim command in a
3485 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3486 was being done only for the command run by the queryprogram router. It is
3487 now done for all such subprocesses. The other cases are: ${run, transport
3488 filters, and the commands run by the lmtp and pipe transports.
3490 5. Some older OS have a limit of 256 on the maximum number of file
3491 descriptors. Exim was using setrlimit() to set 1000 as a large value
3492 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3493 systems. I've change it so that if it can't get 1000, it tries for 256.
3495 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3496 was an oversight, and furthermore, ever since the addition of extra
3497 controls (e.g. 4.43/32), the checks on when to allow different forms of
3498 "control" were broken. There should now be diagnostics for all cases when a
3499 control that does not make sense is encountered.
3501 7. $recipients is now available in the predata ACL (oversight).
3503 8. Tidy the search cache before the fork to do a delivery from a message
3504 received from the command line. Otherwise the child will trigger a lookup
3505 failure and thereby defer the delivery if it tries to use (for example) a
3506 cached ldap connection that the parent has called unbind on.
3508 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3509 of $address_data from the recipient verification was clobbered by the
3510 sender verification.
3512 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3513 was its contents. (It was OK if the option was not defined at all.)
3515 11. A "Completed" log line is now written for messages that are removed from
3516 the spool by the -Mrm option.
3518 12. $host_address is now set to the target address during the checking of
3519 ignore_target_hosts.
3521 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3522 being passed; this would have caused $sender_host_name to have been used if
3523 matching the list had actually called for a host name (not very likely,
3524 since this list is usually IP addresses). A host name is now passed as
3527 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3528 code that specifies a non-restarting handler (typically sigaction() in
3529 modern systems) in an attempt to fix a rare and obscure crash bug.
3531 15. Narrowed the window for a race in the daemon that could cause it to ignore
3532 SIGCHLD signals. This is not a major problem, because they are used only to
3533 wake it up if nothing else does.
3535 16. A malformed maildirsize file could cause Exim to calculate negative values
3536 for the mailbox size or file count. Odd effects could occur as a result.
3537 The maildirsize information is now recalculated if the size or filecount
3540 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3541 support for a long time. Removed HAVE_SYS_VFS_H.
3543 18. Updated exipick to current release from John Jetmore.
3545 19. Allow an empty sender to be matched against a lookup in an address list.
3546 Previously the only cases considered were a regular expression, or an
3549 20. Exim went into a mad DNS lookup loop when doing a callout where the
3550 host was specified on the transport, if the DNS lookup yielded more than
3553 21. The RFC2047 encoding function was originally intended for short strings
3554 such as real names; it was not keeping to the 75-character limit for
3555 encoded words that the RFC imposes. It now respects the limit, and
3556 generates multiple encoded words if necessary. To be on the safe side, I
3557 have increased the buffer size for the ${rfc2047: expansion operator from
3560 22. Failure to deliver a bounce message always caused it to be frozen, even if
3561 there was an errors_to setting on the router. The errors_to setting is now
3564 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3565 canonical form (fully expanded) before being placed in
3566 $sender_host_address.
3568 24. Updated eximstats to version 1.33
3570 25. Include certificate and key file names in error message when GnuTLS fails
3571 to set them up, because the GnuTLS error message doesn't include the name
3572 of the failing file when there is a problem reading it.
3574 26. Expand error message when OpenSSL has problems setting up cert/key files.
3577 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3578 (this can affect the format of dates).
3580 28. exim_tidydb, when checking for the continued existence of a message for
3581 which it has found a message-specific retry record, was not finding
3582 messages that were in split spool directories. Consequently, it was
3583 deleting retry records that should have stayed in existence.
3585 29. eximstats updated to version 1.35
3586 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3587 1.35 - bugfix such that pie charts by volume are generated correctly
3589 30. The SPA authentication driver was not abandoning authentication and moving
3590 on to the next authenticator when an expansion was forced to fail,
3591 contradicting the general specification for all authenticators. Instead it
3592 was generating a temporary error. It now behaves as specified.
3594 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3595 (the order specifies the preference for clients). The order is now AES256,
3596 AES128, 3DES, ARCFOUR128.
3598 31. Small patch to Sieve code - explicitly set From: when generating an
3601 32. Exim crashed if a remote delivery caused a very long error message to be
3602 recorded - for instance if somebody sent an entire SpamAssassin report back
3603 as a large number of 550 error lines. This bug was coincidentally fixed by
3604 increasing the size of one of Exim's internal buffers (big_buffer) that
3605 happened as part of the Exiscan merge. However, to be on the safe side, I
3606 have made the code more robust (and fixed the comments that describe what
3609 33. Some experimental protocols are using DNS PTR records for new purposes. The
3610 keys for these records are domain names, not reversed IP addresses. The
3611 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3612 leaves it alone. Component reversal etc. now happens only for IP addresses.
3615 34. The host_aton() function is supposed to be passed a string that is known
3616 to be a valid IP address. However, in the case of IPv6 addresses, it was
3617 not checking this. This is a hostage to fortune. Exim now panics and dies
3618 if the condition is not met. A case was found where this could be provoked
3619 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3620 components; fortuitously, this particular loophole had already been fixed
3621 by change 4.50/55 or 4.44/33 above.
3623 If there are any other similar loopholes, the new check in host_aton()
3624 itself should stop them being exploited. The report I received stated that
3625 data on the command line could provoke the exploit when Exim was running as
3626 exim, but did not say which command line option was involved. All I could
3627 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3628 running as the user.
3631 35. There was a buffer overflow vulnerability in the SPA authentication code
3632 (which came originally from the Samba project). I have added a test to the
3633 spa_base64_to_bits() function which I hope fixes it.
3636 36. The daemon start-up calls getloadavg() while still root for those OS that
3637 need the first call to be done as root, but it missed one case: when
3638 deliver_queue_load_max is set with deliver_drop_privilege. This is
3639 necessary for the benefit of the queue runner, because there is no re-exec
3640 when deliver_drop_privilege is set.
3642 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3643 was in use, was not putting the data itself into the right store pool;
3644 consequently, it could be overwritten for a subsequent message in the same
3645 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3648 38. Sometimes the final signoff response after QUIT could fail to get
3649 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3650 < 0 before doing a fflush(). This bug looks as though it goes back to the
3651 introduction of TLS in release 3.20, but "sometimes" must have been rare
3652 because the tests only now provoked it.
3658 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3659 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3660 Now it returns an integer. A number of calls were still expecting a T/F
3661 return. Fortuitously, in all cases, the tests worked in OK situations,
3662 which is the norm. However, things would have gone wrong on any write
3663 failures on the smtp file descriptor. This function is used when sending
3664 messages over SMTP and also when doing verify callouts.
3666 2. When Exim is called to do synchronous delivery of a locally submitted
3667 message (the -odf or -odi options), it no longer closes stderr before doing
3670 3. Implemented the mua_wrapper option.
3672 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3674 5. Implemented the functions header_remove(), header_testname(),
3675 header_add_at_position(), and receive_remove_recipient(), and exported them
3678 6. If an ACL "warn" statement specified the addition of headers, Exim already
3679 inserted X-ACL-Warn: at the start if there was no header name. However, it
3680 was not making this test for the second and subsequent header lines if
3681 there were newlines in the string. This meant that an invalid header could
3682 be inserted if Exim was badly configured.
3684 7. Allow an ACL "warn" statement to add header lines at the start or after all
3685 the Received: headers, as well as at the end.
3687 8. Added the rcpt_4xx retry error code.
3689 9. Added postmaster_mailfrom=xxx to callout verification option.
3691 10. Added mailfrom=xxxx to the callout verification option, for verify=
3694 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3695 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3697 12. Inserted some casts to stop certain compilers warning when using pointer
3698 differences as field lengths or precisions in printf-type calls (mostly
3699 affecting debugging statements).
3701 13. Added optional readline() support for -be (dynamically loaded).
3703 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3704 same clock tick as a message's arrival, so that its received time was the
3705 same as the "first fail" time on the retry record, and that message
3706 remained on the queue past the ultimate address timeout, every queue runner
3707 would try a delivery (because it was past the ultimate address timeout) but
3708 after another failure, the ultimate address timeout, which should have then
3709 bounced the address, did not kick in. This was a "< instead of <=" error;
3710 in most cases the first failure would have been in the next clock tick
3711 after the received time, and all would be well.
3713 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3714 being recognized when the domain list was tested by the match_domain
3715 condition in an expansion string.
3717 16. Added the ${str2b64: operator.
3719 17. Exim was always calling setrlimit() to set a large limit for the number of
3720 processes, without checking whether the existing limit was already
3721 adequate. (It did check for the limit on file descriptors.) Furthermore,
3722 errors from getrlimit() and setrlimit() were being ignored. Now they are
3723 logged to the main and panic logs, but Exim does carry on, to try to do its
3724 job under whatever limits there are.
3726 18. Imported PCRE 5.0.
3728 19. Trivial typo in log message " temporarily refused connection" (the leading
3731 20. If the log selector return_path_on_delivery was set and an address was
3732 redirected to /dev/null, the delivery process crashed because it assumed
3733 that a return path would always be set for a "successful" delivery. In this
3734 case, the whole delivery is bypassed as an optimization, and therefore no
3737 21. Internal re-arrangement: the function for sending a challenge and reading
3738 a response while authentication was assuming a zero-terminated challenge
3739 string. It's now changed to take a pointer and a length, to allow for
3740 binary data in such strings.
3742 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3744 23. Exim was not respecting finduser_retries when seeking the login of the
3745 uid under which it was called; it was always trying 10 times. (The default
3746 setting of finduser_retries is zero.) Also, it was sleeping after the final
3747 failure, which is pointless.
3749 24. Implemented tls_on_connect_ports.
3751 25. Implemented acl_smtp_predata.
3753 26. If the domain in control=submission is set empty, Exim assumes that the
3754 authenticated id is a complete email address when it generates From: or
3755 Sender: header lines.
3757 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3758 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3759 chown and chgrp in /bin and hostname in /usr/bin.
3761 28. Exim was keeping the "process log" file open after each use, just as it
3762 does for the main log. This opens the possibility of it remaining open for
3763 long periods when the USR1 signal hits a daemon. Occasional processlog
3764 errors were reported, that could have been caused by this. Anyway, it seems
3765 much more sensible not to leave this file open at all, so that is what now
3768 29. The long-running daemon process does not normally write to the log once it
3769 has entered its main loop, and it closes the log before doing so. This is
3770 so that log files can straightforwardly be renamed and moved. However,
3771 there are a couple of unusual error situations where the daemon does write
3772 log entries, and I had neglected to close the log afterwards.
3774 30. The text of an SMTP error response that was received during a remote
3775 delivery was being truncated at 512 bytes. This is too short for some of
3776 the long messages that one sometimes sees. I've increased the limit to
3779 31. It is now possible to make retry rules that apply only when a message has a
3780 specific sender, in particular, an empty sender.
3782 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3783 it possible to be selective about when SMTP synchronization is enforced.
3785 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3787 32. Implemented hosts_connection_nolog.
3789 33. Added an ACL for QUIT.
3791 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3794 35. Added mailbox_size and mailbox_filecount to appendfile.
3796 36. Added control = no_multiline_responses to ACLs.
3798 37. There was a bug in the logic of the code that waits for the clock to tick
3799 in the case where the clock went backwards by a substantial amount such
3800 that the microsecond fraction of "now" was more than the microsecond
3801 fraction of "then" (but the whole seconds number was less).
3803 38. Added support for the libradius Radius client library this is found on
3804 FreeBSD (previously only the radiusclient library was supported).
3810 1. When certain lookups returned multiple values in the form name=value, the
3811 quoting of the values was not always being done properly. Specifically:
3812 (a) If the value started with a double quote, but contained no whitespace,
3814 (b) If the value contained whitespace other than a space character (i.e.
3815 tabs or newlines or carriage returns) it was not quoted.
3816 This fix has been applied to the mysql and pgsql lookups by writing a
3817 separate quoting function and calling it from the lookup code. The fix
3818 should probably also be applied to nisplus, ibase and oracle lookups, but
3819 since I cannot test any of those, I have not disturbed their existing code.
3821 2. A hit in the callout cache for a specific address caused a log line with no
3822 reason for rejecting RCPT. Now it says "Previous (cached) callout
3823 verification failure".
3825 3. There was an off-by-one bug in the queryprogram router. An over-long
3826 return line was truncated at 256 instead of 255 characters, thereby
3827 overflowing its buffer with the terminating zero. As well as fixing this, I
3828 have increased the buffer size to 1024 (and made a note to document this).
3830 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3831 when Exim is waiting for an SMTP response from a remote server, Exim
3832 restarts its select() call on the socket, thereby resetting its timeout.
3833 This is not a problem when such interrupts are rare. Somebody set up a cron
3834 job to run exiwhat every 2 minutes, which is less than the normal select()
3835 timeout (5 or 10 minutes). This meant that the select() timeout never
3836 kicked in because it was always reset. I have fixed this by comparing the
3837 time when an interrupt arrives with the time at the start of the first call
3838 to select(). If more time than the timeout has elapsed, the interrupt is
3839 treated as a timeout.
3841 5. Some internal re-factoring in preparation for the addition of Sieve
3842 extensions (by MH). In particular, the "personal" test is moved to a
3843 separate function, and given an option for scanning Cc: and Bcc: (which is
3844 not set for Exim filters).
3846 6. When Exim created an email address using the login of the caller as the
3847 local part (e.g. when creating a From: or Sender: header line), it was not
3848 quoting the local part when it contained special characters such as @.
3850 7. Installed new OpenBSD configuration files.
3852 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3853 try to make them clearer.
3855 9. Callout options, other than the timeout value, were being ignored when
3856 verifying sender addresses in header lines. For example, when using
3858 verify = header_sender/callout=no_cache
3860 the cache was (incorrectly) being used.
3862 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3863 only the Cygwin environment.
3865 11. When return_path_on_delivery was set as a log selector, if different remote
3866 addresses in the same message used different return paths and parallel
3867 remote delivery occurred, the wrong values would sometimes be logged.
3868 (Whenever a remote delivery process finished, the return path value from
3869 the most recently started remote delivery process was logged.)
3871 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3872 lines when AUTH and/or TLS are in use. This is the "received protocol"
3873 field. Exim used to use "asmtp" for authenticated SMTP, without any
3874 indication (in the protocol name) for TLS use. Now it follows the RFC and
3875 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3876 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3877 names appear in log lines as well as in Received: header lines.
3879 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3880 extensions, and comparison tests, and to fix some bugs.
3882 14. Changes to the "personal" filter test:
3884 (1) The test was buggy in that it was just doing the equivalent of
3885 "contains" tests on header lines. For example, if a user's address was
3886 anne@some.where, the "personal" test would incorrectly be true for
3888 To: susanne@some.where
3890 This test is now done by extracting each address from the header in turn,
3891 and checking the entire address. Other tests that are part of "personal"
3892 are now done using regular expressions (for example, to check local parts
3893 of addresses in From: header lines).
3895 (2) The list of non-personal local parts in From: addresses has been
3896 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3897 taken from the Sieve specification recommendations.
3899 (3) If the message contains any header line starting with "List-" it is
3900 treated as non-personal.
3902 (4) The test for "circular" in the Subject: header line has been removed
3903 because it now seems ill-conceived.
3905 15. Minor typos in src/EDITME comments corrected.
3907 16. Installed latest exipick from John Jetmore.
3909 17. If headers_add on a router specified a text string that was too long for
3910 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3911 of string_sprintf() is now avoided.
3913 18. $message_body_size was not set (it was always zero) when running the DATA
3914 ACL and the local_scan() function.
3916 19. For the "mail" command in an Exim filter, no default was being set for
3917 the once_repeat time, causing a random time value to be used if "once" was
3918 specified. (If the value happened to be <= 0, no repeat happened.) The
3919 default is now 0s, meaning "never repeat". The "vacation" command was OK
3920 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3921 (I found it when inspecting the code).
3923 20. There is now an overall timeout for performing a callout verification. It
3924 defaults to 4 times the callout timeout, which applies to individual SMTP
3925 commands during the callout. The overall timeout applies when there is more
3926 than one host that can be tried. The timeout is checked before trying the
3927 next host. This prevents very long delays if there are a large number of
3928 hosts and all are timing out (e.g. when the network connections are timing
3929 out). The value of the overall timeout can be changed by specifying an
3930 additional sub-option for "callout", called "maxwait". For example:
3932 verify = sender/callout=5s,maxwait=20s
3934 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3935 to the end before writing, but this should make it even safer).
3937 22. Exim was forgetting that it had advertised PIPELINING for the second and
3938 subsequent messages on an SMTP connection. It was also not resetting its
3939 memory on STARTTLS and an internal HELO.
3941 23. When Exim logs an SMTP synchronization error within a session, it now
3942 records whether PIPELINING has been advertised or not.
3944 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3945 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3946 rather than long int.
3948 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3950 26. Added the never_mail option to autoreply.
3956 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3957 crash if the getsockname() call failed; this can happen if a connection is
3958 closed very soon after it is established. The problem was simply in the
3959 order in which certain operations were done, causing Exim to try to write
3960 to the SMTP stream before it had set up the file descriptor. The bug has
3961 been fixed by making things happen in the correct order.
3967 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3968 before the connection was closed, thus losing the rejection response.
3970 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3971 some early Solaris releases, but causes trouble in current releases where
3972 socklen_t is defined.
3974 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3977 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3980 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3981 that releasing the top of it at the end releases what was used for sub-
3982 expansions (unless the block got too big). However, discard this block if
3983 the first thing is a variable or header, so that we can use its block when
3984 it is dynamic (useful for very large $message_headers, for example).
3986 6. Lookups now cache *every* query, not just the most recent. A new, separate
3987 store pool is used for this. It can be recovered when all lookup caches are
3988 flushed. Lookups now release memory at the end of their result strings.
3989 This has involved some general refactoring of the lookup sources.
3991 7. Some code has been added to the store_xxx() functions to reduce the amount
3992 of flapping under certain conditions.
3994 8. log_incoming_interface used to affect only the <= reception log lines. Now
3995 it causes the local interface and port to be added to several more SMTP log
3996 lines, for example "SMTP connection from", and rejection lines.
3998 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4000 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4002 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4003 could be overwritten at the end of the current message (or the start of a
4004 new message if it was set in a HELO ACL). The value is now preserved for
4005 the duration of the SMTP connection.
4007 12. If a transport had a headers_rewrite setting, and a matching header line
4008 contained an unqualified address, that address was qualified, even if it
4009 did not match any rewriting rules. The underlying bug was that the values
4010 of the flags that permit the existence of unqualified sender and recipient
4011 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4012 non-local messages, and by -bnq for local messages) were not being
4013 preserved with the message after it was received.
4015 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4016 "next input=" as part of the text comprising the host identity instead of
4017 the correct text. The code was using the same buffer for two different
4018 strings. However, depending on which order the printing function evaluated
4019 its arguments, the bug did not always show up. Under Linux, for example, my
4020 test suite worked just fine.
4022 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4023 doesn't work with some older versions of Perl. It has been changed to "my",
4024 which in any case is probably the better facility to use.
4026 15. A really picky compiler found some instances of statements for creating
4027 error messages that either had too many or two few arguments for the format
4030 16. The size of the buffer for calls to the DNS resolver has been increased
4031 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4032 for addresses that have a lot of PTR records. This alleviates a problem; it
4033 does not fully solve it.
4035 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4036 buffer now truncates the list and logs the incident, which is the same
4037 action as happens when Exim is looking up a host name and its aliases.
4038 Previously in this situation something unpredictable would happen;
4039 sometimes it was "internal error: store_reset failed".
4041 18. If a server dropped the connection unexpectedly when an Exim client was
4042 using GnuTLS and trying to read a response, the client delivery process
4043 crashed while trying to generate an error log message.
4045 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4046 string, for example:
4048 warn message = H1: something\nH2: something
4050 the text was added as a single header line from Exim's point of view
4051 though it ended up OK in the delivered message. However, searching for the
4052 second and subsequent header lines using $h_h2: did not work. This has been
4053 fixed. Similarly, if a system filter added multiple headers in this way,
4054 the routers could not see them.
4056 20. Expanded the error message when iplsearch is called with an invalid key to
4057 suggest using net-iplsearch in a host list.
4059 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4060 ACLs are no longer actually imposed (and a message to that effect is
4063 22. If a "gecos" field in a passwd entry contained escaped characters, in
4064 particular, if it contained a \" sequence, Exim got it wrong when building
4065 a From: or a Sender: header from that name. A second bug also caused
4066 incorrect handling when an unquoted " was present following a character
4067 that needed quoting.
4069 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4070 was not being matched caselessly.
4072 24. Arranged for all hyphens in the exim.8 source to be escaped with
4075 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4076 properly. Recipient callout cache records were still being keyed to include
4077 the sender, even when use_sender was set false. This led to far more
4078 callouts that were necessary. The sender is no longer included in the key
4079 when use_sender is false.
4081 26. Added "control = submission" modifier to ACLs.
4083 27. Added the ${base62d: operator to decode base 62 numbers.
4085 28. dnsdb lookups can now access SRV records.
4087 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4088 the configuration file.
4090 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4091 (-v) mode. This makes the output for a verbose queue run more intelligible.
4093 31. Added a use_postmaster feature to recipient callouts.
4095 32. Added the $body_zerocount variable, containing the number of binary zero
4096 bytes in the message body.
4098 33. The time of last modification of the "new" subdirectory is now used as the
4099 "mailbox time last read" when there is a quota error for a maildir
4102 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4104 35. Added +ignore_unknown as a special item in host lists.
4106 36. Code for decoding IPv6 addresses in host lists is now included, even if
4107 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4108 address was recognized as an IP address, but was then not correctly decoded
4109 into binary, causing unexpected and incorrect effects when compared with
4116 1. Very minor rewording of debugging text in manualroute to say "list of
4117 hosts" instead of "hostlist".
4119 2. If verify=header_syntax was set, and a header line with an unqualified
4120 address (no domain) and a large number of spaces between the end of the
4121 name and the colon was received, the reception process suffered a buffer
4122 overflow, and (when I tested it) crashed. This was caused by some obsolete
4123 code that should have been removed. The fix is to remove it!
4125 3. When running in the test harness, delay a bit after writing a bounce
4126 message to get a bit more predictability in the log output.
4128 4. Added a call to search_tidyup() just before forking a reception process. In
4129 theory, someone could use a lookup in the expansion of smtp_accept_max_
4130 per_host which, without the tidyup, could leave open a database connection.
4132 5. Added the variables $recipient_data and $sender_data which get set from a
4133 lookup success in an ACL "recipients" or "senders" condition, or a router
4134 "senders" option, similar to $domain_data and $local_part_data.
4136 6. Moved the writing of debug_print from before to after the "senders" test
4139 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4140 problems for message scanning, either using a data ACL, or using
4141 local_scan() because the Received: header was not generated till after they
4142 were called (in order to set the time as the time of reception completion).
4143 I have revised the way this works. The header is now generated after the
4144 body is received, but before the ACL or local_scan() are called. After they
4145 are run, the timestamp in the header is updated.
4151 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4152 before starting a queue runner without re-exec. This happened only when
4153 deliver_drop_privilege was set or when the Exim user was set to root. The
4154 effect of the bug was that timeouts during subsequent deliveries caused
4155 crashes instead of being properly handled. The handler is now left at its
4156 default (and expected) setting.
4158 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4159 message, again when deliver_drop_privilege is set or Exim is run as root.
4160 The bug described in (1) was not present in this case, but the tidying up
4161 of the other signals was missing. I have made the two cases consistent.
4163 3. The ignore_target_hosts setting on a manualroute router was being ignored
4164 for hosts that were looked up using the /MX notation.
4166 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4169 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4170 operated on the sender address. After changing the $sender_address to <>
4171 for the sender address verify, Exim was re-instated it as the original
4172 (before rewriting) address, but remembering that it had rewritten it, so it
4173 wasn't rewriting it again. This bug also had the effect of breaking the
4174 sender address verification caching when the sender address was rewritten.
4176 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4177 This has been changed so that if the ip literal address matches
4178 ignore_target_hosts, the router declines.
4180 7. Added expansion conditions match_domain, match_address, and match_local_
4181 part (NOT match_host).
4183 8. The placeholder for the Received: header didn't have a length field set.
4185 9. Added code to Exim itself and to exim_lock to test for a specific race
4186 condition that could lead to file corruption when using MBX delivery. The
4187 issue is with the lockfile that is created in /tmp. If this file is removed
4188 after a process has opened it but before that process has acquired a lock,
4189 there is the potential for a second process to recreate the file and also
4190 acquire a lock. This could lead to two Exim processes writing to the file
4191 at the same time. The added code performs the same test as UW imapd; it
4192 checks after acquiring the lock that its file descriptor still refers to
4193 the same named file.
4195 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4196 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4199 11. Added the smtp_active_hostname option. If used, this will typically be made
4200 to depend on the incoming interface address. Because $interface_address is
4201 not set up until the daemon has forked a reception process, error responses
4202 that can happen earlier (such as "too many connections") no longer contain
4205 12. If an expansion in a condition on a "warn" statement fails because a lookup
4206 defers, the "warn" statement is abandoned, and the next ACL statement is
4207 processed. Previously this caused the whole ACL to be aborted.
4209 13. Added the iplsearch lookup type.
4211 14. Added ident_timeout as a log selector.
4213 15. Added tls_certificate_verified as a log selector.
4215 16. Added a global option tls_require_ciphers (compare the smtp transport
4216 option of the same name). This controls incoming TLS connections.
4218 17. I finally figured out how to make tls_require_ciphers do a similar thing
4219 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4220 before starting the TLS session.
4222 18. Tabs are now shown as \t in -bP output.
4224 19. If the log selector return_path_on_delivery was set, Exim crashed when
4225 bouncing a message because it had too many Received: header lines.
4227 20. If two routers both had headers_remove settings, and the first one included
4228 a superfluous trailing colon, the final name in the first list and the
4229 first name in the second list were incorrectly joined into one item (with a
4230 colon in the middle).
4236 1. Added -C and -D options to the exinext utility, mainly to make it easier
4237 to include in the automated testing, but these could be helpful when
4238 multiple configurations are in use.
4240 2. The exinext utility was not formatting the output nicely when there was
4241 an alternate port involved in the retry record key, nor when there was a
4242 message id as well (for retries that were specific to a specific message
4243 and a specific host). It was also confused by IPv6 addresses, because of
4244 the additional colons they contain. I have fixed the IPv4 problem, and
4245 patched it up to do a reasonable job for IPv6.
4247 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4248 delivery, the log line now contains "pipelined" if PIPELINING was used.
4250 4. An SMTP transport process used to panic and die if the bind() call to set
4251 an explicit outgoing interface failed. This has been changed; it is now
4252 treated in the same way as a connect() failure.
4254 5. A reference to $sender_host_name in the part of a conditional expansion
4255 that was being skipped was still causing a DNS lookup. This no longer
4258 6. The def: expansion condition was not recognizing references to header lines
4259 that used bh_ and bheader_.
4261 7. Added the _cache feature to named lists.
4263 8. The code for checking quota_filecount in the appendfile transport was
4264 allowing one more file than it should have been.
4266 9. For compatibility with Sendmail, the command line option
4274 and sets the incoming protocol and host name (for trusted callers). The
4275 host name and its colon can be omitted when only the protocol is to be set.
4276 Note the Exim already has two private options, -pd and -ps, that refer to
4277 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4278 "s", but I don't think that's a major issue.
4280 10. A number of refactoring changes to the code, none of which should affect
4283 (a) The number of logging options was getting close to filling up the
4284 32-bit word that was used as a bit map. I have split them into two classes:
4285 those that are passed in the argument to log_write(), and those that are
4286 only ever tested independently outside of that function. These are now in
4287 separate 32-bit words, so there is plenty of room for expansion again.
4288 There is no change in the user interface or the logging behaviour.
4290 (b) When building, for example, log lines, the code previously used a
4291 macro that called string_cat() twice, in order to add two strings. This is
4292 not really sufficiently general. Furthermore, there was one instance where
4293 it was actually wrong because one of the argument was used twice, and in
4294 one call a function was used. (As it happened, calling the function twice
4295 did not affect the overall behaviour.) The macro has been replaced by a
4296 function that can join an arbitrary number of extra strings onto a growing
4299 (c) The code for expansion conditions now uses a table and a binary chop
4300 instead of a serial search (which was left over from when there were very
4301 few conditions). Also, it now recognizes conditions like "pam" even when
4302 the relevant support is not compiled in: a suitably worded error message is
4303 given if an attempt is made to use such a condition.
4305 11. Added ${time_interval:xxxxx}.
4307 12. A bug was causing one of the ddress fields not to be passed back correctly
4308 from remote delivery subprocesses. The field in question was not being
4309 subsequently used, so this caused to problems in practice.
4311 13. Added new log selectors queue_time and deliver_time.
4313 14. Might have fixed a bug in maildirsizefile handling that threw up
4314 "unexpected character" debug warnings, and recalculated the data
4315 unnecessarily. In any case, I expanded the warning message to give more
4318 15. Added the message "Restricted characters in address" to the statements in
4319 the default ACL that block characters like @ and % in local parts.
4321 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4322 Three changes have been made:
4324 (a) There was a serious bug; a negative response to MAIL caused the whole
4325 recipient domain to be cached as invalid, thereby blocking all messages
4326 to all local parts at the same domain, from all senders. This bug has
4327 been fixed. The domain is no longer cached after a negative response to
4328 MAIL if the sender used is not empty.
4330 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4333 (c) A new callout option, "use_sender" has been added for people who want
4334 the modified behaviour.
4340 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4343 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4344 indeed breaks things for older releases.
4346 3. Added additional logging to the case where there is a problem reading data
4347 from a filter that is running in a subprocess using a pipe, in order to
4348 try to track down a specific problem.
4350 4. Testing facility fudge: when running in the test harness and attempting
4351 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4352 getting "No route to host". Convert this to a timeout.
4354 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4357 6. Some OS don't have socklen_t but use size_t instead. This affects the
4358 fifth argument of getsockopt() amongst other things. This is now
4359 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4360 can be set for individual OS. I have set it for SunOS5, OSF1, and
4361 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4362 some earlier ones do not.
4364 7. Change 4.30/15 was not doing the test caselessly.
4366 8. The standard form for an IPv6 address literal was being rejected by address
4367 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4368 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4369 this, as well as the form without the "IPv6" on the front (but only when
4370 address literals are enabled, of course).
4372 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4374 10. Exim crashed if a message with an empty sender address specified by -f
4375 encountered a router with an errors_to setting. This could be provoked only
4376 by a command such as
4380 where an empty string was supplied; "<>" did not hit this bug.
4382 11. Installed PCRE release 4.5.
4384 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4385 remained set. It is now erased.
4387 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4388 times from message ids (which are base 36 rather than the normal 62).
4390 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4391 were being counted as actual protocol errors, and logged if the log
4392 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4393 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4394 and DATA following a set of rejected RCPTs do not count as protocol errors.
4395 In other words, Exim assumes they were pipelined, though this may not
4396 actually be the case. Of course, in all cases the client gets an
4397 appropriate error code.
4399 15. If a lookup fails in an ACL condition, a message about the failure may
4400 be available; it is used if testing the ACL cannot continue, because most
4401 such messages specify what the cause of the deferral is. However, some
4402 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4403 that caused an old message to be retained and used if a later statement
4404 caused a defer, replacing the real cause of the deferral.
4406 16. If an IP address had so many PTR records that the DNS lookup buffer
4407 was not large enough to hold them, Exim could crash while trying to process
4408 the truncated data. It now detects and logs this case.
4410 17. Further to 4.21/58, another change has been made: if (and only if) the
4411 first line of a message (the first header line) ends with CRLF, a bare LF
4412 in a subsequent header line has a space inserted after it, so as not to
4413 terminate the header.
4415 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4416 unnecessarily, used atoi() instead of strtol(), and didn't check the
4417 termination when getting file sizes from file names by regex.
4419 19. Completely re-implemented the support for maildirsize files, in the light
4420 of a number of problems with the previous contributed implementation
4421 (4.30/29). In particular:
4423 . If the quota is zero, the maildirsize file is maintained, but no quota is
4426 . If the maildir directory does not exist, it is created before any attempt
4427 to write a maildirsize file.
4429 . The quota value in the file is just a cache; if the quota is changed in
4430 the transport, the new value overrides.
4432 . A regular expression is available for excluding directories from the
4435 20. The autoreply transport checks the characters in options that define the
4436 message's headers; it allows continued headers, but it was checking with
4437 isspace() after an embedded newline instead of explicitly looking for a
4440 21. If all the "regular" hosts to which an address was routed had passed their
4441 expiry times, and had not reached their retry times, the address was
4442 bounced, even if fallback hosts were defined. Now Exim should go on to try
4445 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4446 equivalent code in the SMTP transport. Some hosts send humungous responses
4447 to HELO/EHLO, more than 1024 it seems.
4449 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4450 gives compiler warnings in some environments. I've now done it "properly",
4453 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4454 (because of problems with the built-in one) was declared to return uschar *
4455 instead of char *, causing compiler failure.
4457 25. Fixed a file descriptor leak when processing alias/forward files.
4459 26. Fixed a minor format string issue in dbfn.c.
4461 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4463 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4464 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4467 29. When a sender address is verified, it is cached, to save repeating the test
4468 when there is more than one recipient in a message. However, when the
4469 verification involves a callout, it is possible for different callout
4470 options to be set for different recipients. It is too complicated to keep
4471 track of this in the cache, so now Exim always runs a verification when a
4472 callout is required, relying on the callout cache for the optimization.
4473 The overhead is duplication of the address routing, but this should not be
4476 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4477 to be verified with callout=postmaster, and the main callout worked but the
4478 postmaster check failed, the verification correctly failed. However, if a
4479 subsequent RCPT command asked for sender verification *without* the
4480 postmaster check, incorrect caching caused this verification also to fail,
4483 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4484 it was not caching the DNS options (qualify_single, search_parents) that
4485 were used when the lookup failed. A subsequent lookup with different
4486 options therefore always gave the same answer, though there were cases
4487 where it should not have. (Example: a "domains = !$mx_any" option on a
4488 dnslookup router: the "domains" option is always processed without any
4489 widening, but the router might have qualify_single set.) Now Exim uses the
4490 cached value only when the same options are set.
4492 32. Added John Jetmore's "exipick" utility to the distribution.
4494 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4495 than a timeout (e.g. a certificate is required, and is not provided), an
4496 Exim server now closes the connection immediately. Previously it waited for
4497 the client to close - but if the client is SSL, it seems that they each
4498 wait for each other, leading to a delay before one of them times out.
4500 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4501 maintained 0.8.x compatibility because I don't think many are using it, and
4502 it is clearly obsolete.
4504 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4507 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4508 client certificate was expired. A simple patch fixes this, though I don't
4509 understand the full logic of why the verify callback is called multiple
4512 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4513 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4514 which causes problems with some clients (such as the Certicom SSL Plus
4515 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4516 disables the coutermeasure allowing Eudora to connect."
4518 38. Exim was not checking that a write() to a log file succeeded. This could
4519 lead to Bad Things if a log got too big, in particular if it hit a file
4520 size limit. Exim now panics and dies if it cannot write to a log file, just
4521 as it does if it cannot open a log file.
4523 39. Modified OS/Makefile-Linux so that it now contains
4525 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4527 The two -D definitions ensure that Exim is compiled with large file
4528 support, which makes it possible to handle log files that are bigger than
4531 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4532 instance) a domain was checked against a named list that involved a lookup,
4533 causing $domain_data to be set, then another domain was checked against the
4534 same list, then the first domain was re-checked, the value of $domain_data
4535 after the final check could be wrong. In particular, if the second check
4536 failed, it could be set empty. This bug probably also applied to
4539 41. The strip_trailing_dot option was not being applied to the address given
4540 with the -f command-line option.
4542 42. The code for reading a message's header from the spool was incrementing
4543 $received_count, but never initializing it. This meant that the value was
4544 incorrect (doubled) while delivering a message in the same process in which
4545 it was received. In the most common configuration of Exim, this never
4546 happens - a fresh exec is done - but it can happen when
4547 deliver_drop_privilege is set.
4549 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4550 it now includes up to 150 characters of the unexpected data in the log
4553 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4554 and building data strings. The size of both of these buffers was 10 000
4555 bytes - far larger than anybody would *ever* want, thought I. Needless to
4556 say, somebody hit the limit. I have increased the maximum line length to
4557 20 000 and the maximum data length of concatenated lines to 100 000. I have
4558 also fixed two bugs, because there was no checking on these buffers. Tsk,
4559 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4560 buffer is too small.
4562 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4563 lsearch lookups. Now it does.
4565 46. When parsing a route_list item in a manualroute router, a fixed-length
4566 buffer was used for the list of hosts. I made this 1024 bytes long,
4567 thinking that nobody would ever have a list of hosts that long. Wrong.
4568 Somebody had a whole pile of complicated expansion conditions, and the
4569 string was silently truncated, leading to an expansion error. It turns out
4570 that it is easier to change to an unlimited length (owing to other changes
4571 that have happened since this code was originally written) than to build
4572 structure for giving a limitation error. The length of the item that
4573 expands into the list of hosts is now unlimited.
4575 47. The lsearch lookup could not handle data where the length of text line was
4576 more than 4095 characters. Such lines were truncated, leading to shortened
4577 data being returned. It should now handle lines of any length.
4579 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4580 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4583 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4584 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4585 no longer happens in comment lines. A list of replacements is now placed
4586 at the head of all of the source files, except those whose only change is
4587 to replace PERL_COMMAND in the very first #! line.
4589 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4590 messages on the queue, with a bottom-up merge sort, using code contributed
4591 by Michael Haardt. This should make operations like -bp somewhat faster on
4592 large queues. It won't affect queue runners, except when queue_run_in_order
4595 51. Installed eximstats 1.31 in the distribution.
4597 52. Added support for SRV lookups to the dnslookup router.
4599 53. If an ACL referred to $message_body or $message_body_end, the value was not
4600 reset for any messages that followed in the same SMTP session.
4602 54. The store-handling optimization for building very long strings was not
4603 differentiating between the different store pools. I don't think this
4604 actually made any difference in practice, but I've tidied it.
4606 55. While running the routers to verify a sender address, $sender_address
4607 was still set to the sender address. This is wrong, because when routing to
4608 send a bounce to the sender, it would be empty. Therefore, I have changed
4609 it so that, while verifying a sender address, $sender_address is set to <>.
4610 (There is no change to what happens when verifying a recipient address.)
4612 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4613 target A or AAAA records (if not already returned) without resetting the
4614 qualify_single or search_parents options of the DNS resolver. These are
4615 inappropriate in this case because the targets of MX and SRV records must
4616 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4617 target that, when qualified, matched something in the local domain. These
4618 two options are now turned off when doing these lookups.
4620 57. It seems that at least some releases of Reiserfs (which does not have the
4621 concept of a fixed number of inodes) returns zero and not -1 for the
4622 number of available inodes. This interacted badly with check_spool_inodes,
4623 which assumed that -1 was the "no such thing" setting. What I have done is
4624 to check that the total number of inodes is greater than zero before doing
4625 the test of how many are available.
4627 58. When a "warn" ACL statement has a log_message modifier, the message is
4628 remembered, and not repeated. This is to avoid a lot of repetition when a
4629 message has many recipients that cause the same warning to be written.
4630 Howewer, Exim was preserving the list of already written lines for an
4631 entire SMTP session, which doesn't seem right. The memory is now reset if a
4632 new message is started.
4634 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4635 debugging output unless log_rewrite was also set.
4637 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4638 of (int)(handle) when we know that handle contains (void *)(-1).
4640 61. The Exim daemon panic-logs an error return when it closes the incoming
4641 connection. However "connection reset by peer" seems to be common, and
4642 isn't really an error worthy of noting specially, so that particular error
4645 62. When Exim is trying to find all the local interfaces, it used to panic and
4646 die if the ioctl to get the interface flags failed. However, it seems that
4647 on at least one OS (Solaris 9) it is possible to have an interface that is
4648 included in the list of interfaces, but for which you get a failure error
4649 for this call. This happens when the interface is not "plumbed" into a
4650 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4651 failure of the "get flags" call assumes that the interface is down.
4653 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4654 makes life easier for people who are doing arithmetic on fields extracted
4655 from dates, where you often get leading zeros that should not be
4656 interpreted as octal.
4658 64. Added qualify_domain to the redirect router, to override the global
4661 65. If a pathologically long header line contained very many addresses (the
4662 report of this problem mentioned 10 000) and each of them was rewritten,
4663 Exim could use up a very large amount of memory. (It kept on making new
4664 copies of the header line as it rewrote, and never released the old ones.)
4665 At the expense of a bit more processing, the header rewriting function has
4666 been changed so that it no longer eats memory in this way.
4668 66. The generation of the Received: header has been moved from the time that a
4669 message starts to be received, to the time that it finishes. The timestamp
4670 in the Received: header should now be very close to that of the <= log
4671 line. There are two side-effects of this change:
4673 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4674 logged header lines no longer include the local Received: line, because
4675 it has not yet been created. The same applies to a copy of the message
4676 that is returned to a non-SMTP sender when a message is rejected.
4678 (b) When a filter file is tested using -bf, no additional Received: header
4679 is added to the test message. After some thought, I decided that this
4682 This change does not affect the value of $received_for. It is still set
4683 after address rewriting, but before local_scan() is called.
4685 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4687 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4688 gave an unhelpful panic error message, and a defer error. I have managed to
4689 change this behaviour so that it now rejects any supplied certificate,
4690 which seems right, as the list of acceptable certificates is empty.
4692 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4693 gave an unhelpful defer error. I have not managed to make this reject any
4694 supplied certificates, but the error message it gives is "no certificate
4695 supplied", which is not helpful.
4697 70. exigrep's output now also includes lines that are not associated with any
4698 message, but which match the given pattern. Implemented by a patch from
4699 Martin Sluka, which also tidied up the Perl a bit.
4701 71. Recipient callout verification, like sender verification, was using <> in
4702 the MAIL FROM command. This isn't really the right thing, since the actual
4703 sender may affect whether the remote host accepts the recipient or not. I
4704 have changed it to use the actual sender in the callout; this means that
4705 the cache record is now keyed on a recipient/sender pair, not just the
4706 recipient address. There doesn't seem to be a real danger of callout loops,
4707 since a callout by the remote host to check the sender would use <>.
4708 [SEE ABOVE: changed after hitting problems.]
4710 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4711 temporary errors. However, in the case of such a code being given after
4712 the end of a data transmission (i.e. after ".") Exim was failing to write
4713 a retry record for the message. (Yes, there was some broken host that was
4714 actually sending 8xx at this point.)
4716 73. An unknown lookup type in a host list could cause Exim to panic-die when
4717 the list was checked. (An example that provoked this was putting <; in the
4718 middle of a list instead of at the start.) If this happened during a DATA
4719 ACL check, a -D file could be left lying around. This kind of configuration
4720 error no longer causes Exim to die; instead it causes a defer errror. The
4721 incident is still logged to the main and panic logs.
4723 74. Buglet left over from Exim 3 conversion. The message "too many messages
4724 in one connection" was written to the rejectlog but not the mainlog, except
4725 when address rewriting (yes!) was being logged.
4727 75. Added write_rejectlog option.
4729 76. When a system filter was run not as root (that is, when system_filter_user
4730 was set), the values of the $n variables were not being returned to the
4731 main process; thus, they were not subsequently available in the $sn
4734 77. Added +return_path_on_delivery log selector.
4736 78. A connection timeout was being treated differently from recipients deferred
4737 when testing hosts_max_try with a message that was older than the host's
4738 retry timeout. (The host should not be counted, thus allowing all hosts to
4739 be tried at least once before bouncing.) This may have been the cause of an
4740 occasionally reported bug whereby a message would remain on the queue
4741 longer than the retry timeout, but would be bounced if a delivery was
4742 forced. I say "may" because I never totally pinned down the problem;
4743 setting up timeout/retry tests is difficult. See also the next item.
4745 79. The ultimate address timeout was not being applied to errors that involved
4746 a combination of host plus message (for example, a timeout on a MAIL
4747 command). When an address resolved to a number of possible hosts, and they
4748 were not all tried for each delivery (e.g. because of hosts_max_try), a
4749 message could remain on the queue longer than the retry timeout.
4751 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4754 81. Fixed an obscure SMTP outgoing bug which required at least the following
4755 conditions: (a) there was another message waiting for the same server;
4756 (b) the server returned 5xx to all RCPT commands in the first message so
4757 that the message was not completed; (c) the server dropped the connection
4758 or gave a negative response to the RSET that Exim sends to abort the
4759 transaction. The observed case was a dropped connection after DATA that had
4760 been sent in pipelining mode. That is, the server had advertised PIPELINING
4761 but was not implementing it correctly. The effect of the bug was incorrect
4762 behaviour, such as trying another host, and this could lead to a crash.
4768 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4769 and daemon.c were passed as pointers to ints; they should have been
4770 pointers to socklen_t variables (which are typically unsigned ints).
4772 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4775 3. Fixed a really odd bug that affected only the testing scheme; patching a
4776 certain fixed string in the binary changed the value of another string that
4777 happened to be identical to the end of the original first string.
4779 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4780 name", it returns that address as the IP address. On some operating
4781 systems (e.g. Solaris), it also passes back the IP address string as the
4782 "host name". However, on others (e.g. Linux), it passes back an empty
4783 string. Exim wasn't checking for this, and was changing the host name to an
4784 empty string, assuming it had been canonicized.
4786 5. Although rare, it is permitted to have more than one PTR record for a given
4787 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4788 all the names associated with an address, because they do in Solaris.
4789 However, it seems that they do not in Linux for data that comes from the
4790 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4791 I found this out when I moved to a new Linux workstation and tried to run
4792 the Exim test suite.
4794 To get round this problem I have changed the code so that it now does its
4795 own call to the DNS to look up PTR records when searching for a host name.
4796 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4797 addresses that are only in /etc/hosts are still found.
4799 This behaviour is, however, controlled by an option called host_lookup_
4800 order, which defaults to "bydns:byaddr". If people want to use the other
4801 order, or indeed, just use one or the other means of lookup, they can
4802 specify it in this variable.
4804 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4805 some operating systems, this comes back from gethostbyaddr() as an empty
4806 string, and this is what Exim used to test for. However, it seems that in
4807 other systems, "." is yielded. Exim now tests for this case too.
4809 7. The values of check_spool_space and check_log_space are now held internally
4810 as a number of kilobytes instead of an absolute number of bytes. If a
4811 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4812 kilobyte. This means that much larger values can be stored.
4814 8. Exim monitor: an attempt to get the action menu when not actually pointing
4815 at a message produces an empty menu entitled "No message selected". This
4816 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4817 no entries in it ("Shell widget menu has zero width and/or height"). So I
4818 have added a single, blank menu entry in this case.
4820 9. Added ${quote_local_part.
4822 10. MIME decoding is now applied to the contents of Subject: header lines when
4825 11. Now that a reference to $sender_host_address automatically causes a reverse
4826 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4827 host lookup before query-style lookups in lists that might use this
4828 variable. This has therefore been abolished, and the "net-" prefix is no
4829 longer necessary for query-style lookups.
4831 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4832 have been a typo for LFLAGS, so it has been changed.
4834 13. The install script calls Exim with "-C /dev/null" in order to find the
4835 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4836 to be output. Howeve, since Exim outputs its version number before the
4837 error, it didn't break the script. It just looked ugly. I fixed this by
4838 always allowing "-C /dev/null" if the caller is root.
4840 14. Ignore overlarge ACL variable number when reading spool file - insurance
4841 against a later release with more variables having written the file.
4843 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4844 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4845 this, as well as the form without the "IPv6" on the front.
4847 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4848 OS/Makefile-Darwin file.
4850 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4851 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4853 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4854 variable. However, some OS use ints for inodes. I've added cast to long int
4855 to get rid of the compiler warning.
4857 19. I had forgotten to lock out "/../" in configuration file names when
4858 ALT_CONFIG_PREFIX was set.
4860 20. Routers used for verification do not need to specify transports. However,
4861 if such a router generated a host list, and callout was configured, Exim
4862 crashed, because it could not find a port number from the (non-existent)
4863 transport. It now assumes port 25 in this circumstance.
4865 21. Added the -t option to exigrep.
4867 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4868 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4869 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4870 them is set, LOOKUP_LSEARCH is forced.
4872 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4874 24. Added sender and host information to the "rejected by local_scan()" log
4875 line; previously there was no indication of these.
4877 25. Added .include_if_exists.
4879 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4880 Linux. It turns out that not all file systems support this. Apparently some
4881 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4882 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4883 not supported on the file descriptor, is now ignored when Exim is trying to
4884 sync a directory. This applies only to Linux.
4886 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4888 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4889 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4892 29. Added support for maildirsize files from supplied patch (modified a bit).
4894 30. The use of :fail: followed by an empty string could lead Exim to respond to
4895 sender verification failures with (e.g.):
4897 550 Verification failed for <xxx>
4898 550 Sender verify failed
4900 where the first response line was missing the '-' that indicates it is not
4901 the final line of the response.
4903 31. The loop for finding the name of the user that called Exim had a hardwired
4904 limit of 10; it now uses the value of finduser_retries, which is used for
4905 all other user lookups.
4907 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4910 33. Exim was neglecting to zero errno before one call of strtol() when
4911 expanding a string and expecting an integer value. On some systems this
4912 resulted in spurious "integer overflow" errors. Also, it was casting the
4913 result into an int without checking.
4915 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4916 did not work. The code looks as if it has *never* worked, though it appears
4917 to have been documented since at least releast 1.62. I have made it work.
4919 35. The "timeout_DNS" error in retry rules, also documented since at least
4920 1.62, also never worked. As it isn't clear exactly what this means, and
4921 clearly it isn't a major issue, I have abolished the feature by treating it
4922 as "timeout", and writing a warning to the main and panic logs.
4924 36. The display of retry rules for -brt wasn't always showing the error code
4927 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4928 timeout_connect_A, timeout_connect_MX.
4930 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4931 to the empty sender.
4933 39. The daemon was not analysing the content of -oX till after it had closed
4934 stderr and disconnected from the controlling terminal. This meant that any
4935 syntax errors were only noted on the panic log, and the return code from
4936 the command was 0. By re-arranging the code a little, I've made the
4937 decoding happen first, so such errors now appear on stderr, and the return
4938 code is 1. However, the actual setting up of the sockets still happens in
4939 the disconnected process, so errors there are still only recorded on the
4942 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4943 connections (as happens on some IP stacks) was logged at start up time as
4944 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4945 it from "IPv6 and IPv4", which means that two separate sockets are being
4948 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4949 says whether AF_INET or AF_INET6 was passed as an argument.
4951 42. Exiwhat output was messed up when time zones were included in log
4954 43. Exiwhat now gives more information about the daemon's listening ports,
4955 and whether -tls-on-connect was used.
4957 44. The "port" option of the smtp transport is now expanded.
4959 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4960 silently ignored. Now an error message is written to the main and panic
4963 46. There's a new ACL modifier called "logwrite" which writes to a log file
4964 as soon as it is encountered.
4966 47. Added $local_user_uid and $local_user_gid at routing time.
4968 48. Exim crashed when trying to verify a sender address that was being
4971 49. Exim was recognizing only a space character after ".include". It now also
4972 recognizes a tab character.
4974 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4975 extracting the relevant information from the specification. The man page no
4976 longer contains scrambled data for the -d option, and I've added a section
4977 at the front about calling Exim under different names.
4979 51. Added "extra_headers" argument to the "mail" command in filter files.
4981 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4984 53. Installed eximstats 1.29.
4986 54. Added transport_filter_timeout as a generic transport option.
4988 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4989 Cc: header lines. This was required by RFC 822, but it not required by RFC
4992 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4993 incoming messages that did not have them. Now it does so only if the
4994 message originates locally, that is, if there is no associated remote host
4995 address. When Resent- header lines are present, this applies to the Resent-
4996 lines rather than the non-Resent- lines.
4998 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4999 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5001 58. Messages for configuration errors now include the name of the main
5002 configuration file - useful now that there may be more than one file in a
5003 list (.included file names were always shown).
5005 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5006 for those rare installations that do not start the daemon as root or run it
5007 setuid root. I've cut out the call to initgroups() if the daemon is not
5010 60. The Exim user and group can now be bound into the binary as text strings
5011 that are looked up at the start of Exim's processing.
5013 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5015 62. Added $mailstore_basename variable.
5017 63. Installed patch to sieve.c from Michael Haardt.
5019 64. When Exim failed to open the panic log after failing to open the main log,
5020 the original message it was trying to log was written to stderr and debug
5021 output, but if they were not available (the usual case in production), it
5022 was lost. Now it is written to syslog before the two lines that record the
5023 failures to open the logs.
5025 65. Users' Exim filters run in subprocesses under the user's uid. It is
5026 possible for a "deliver" command or an alias in a "personal" command to
5027 provoke an address rewrite. If logging of address rewriting is configured,
5028 this fails because the process is not running as root or exim. There may be
5029 a better way of dealing with this, but for the moment (because 4.30 needs
5030 to be released), I have disabled address rewrite logging when running a
5031 filter in a non-root, non-exim process.
5037 1. The buildconfig auxiliary program wasn't quoting the value set for
5038 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5039 not defined. This bug was masked in 4.22 by the effect that was fixed in
5042 2. Some messages that were rejected after a message id was allocated were
5043 shown as "incomplete" by exigrep. It no longer does this for messages that
5044 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5046 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5047 have allow_domain_literals set, the ID did not get logged in the <= line.
5048 Domain literals are now always recognized in Message-ID: header lines.
5050 4. The first argument for a ${extract expansion item is the key name or field
5051 number. Leading and trailing spaces in this item were not being ignored,
5052 causing some misleading effects.
5054 5. When deliver_drop_privilege was set, single queue runner processes started
5055 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5056 same command in the process it spins off) were not dropping privilege.
5058 6. When the daemon running as "exim" started a queue runner, it always
5059 re-executed Exim in the spun-off process. This is a waste of effort when
5060 deliver_drop_privilege is set. The new process now just calls the
5061 queue-runner function directly.
5067 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5070 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5071 ignored. Though the use of -oP was forcing the writing of a pid file, it
5072 was always written to the default place.
5074 3. If the message "no IP address found for host xxxx" is generated during
5075 incoming verification, it is now followed by identification of the incoming
5076 connection (so you can more easily find what provoked it).
5078 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5080 5. Added some features to "harden" Exim a bit more against certain attacks:
5082 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5083 be put in Local/Makefile. This is like the never_users runtime option,
5084 but it cannot be overridden. The default setting is "root".
5086 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5087 prefix string with which any file named in a -C command line option
5090 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5091 is retained for -C and -D only if the caller of Exim is root. Without
5092 it, the exim user may also use -C and -D and retain privilege.
5094 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5095 command line option is disabled.
5097 6. Macro names set by the -D option must start with an upper case letter, just
5098 like macro names defined in the configuration file.
5100 7. Added "dereference=" facility to LDAP.
5102 8. Two instances of the typo "uknown" in the source files are fixed.
5104 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5105 the Configure-Makefile script screwed up while processing it.
5107 10. Incorporated PCRE 4.4.
5109 11. The SMTP synchronization check was not operating right at the start of an
5110 SMTP session. For example, it could not catch a HELO sent before the client
5111 waited for the greeting. There is now a check for outstanding input at the
5112 point when the greeting is written. Because of the duplex, asynchronous
5113 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5114 way, but not yet received, when the check is performed.
5116 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5117 on TCP/IP sockets, because this apparently causes some broken clients to
5120 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5121 unchanged) from the Cygwin maintainer.
5123 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5124 was supported instead of testing for mbx. Effectively a typo.
5126 15. The spa authenticator server code was not checking that the input it
5127 received was valid base64.
5129 16. The debug output line for the "set" modifier in ACLs was not showing the
5130 name of the variable that was being set.
5132 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5134 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5135 lookup on its own. Something else was needed to trigger the lookup. For
5136 example, a match in host_lookup or the need for a host name in a host list.
5137 Now, if $sender_host_name is referenced and the host name has not yet been
5138 looked up, a lookup is performed. If the lookup fails, the variable remains
5139 empty, and $host_lookup_failed is set to "1".
5141 19. Added "eqi" as a case-independent comparison operator.
5143 20. The saslauthd authentication condition could segfault if neither service
5144 nor realm was specified.
5146 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5147 error message that was logged was misleading, and incoming SMTP
5148 connections were dropped. The message is now more accurate, and temporary
5149 errors are given to SMTP connections.
5151 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5152 (except RSET) with a 421 error, until QUIT is received. However, it was
5153 failing to send a response to QUIT.
5155 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5156 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5157 ACL is now run after the helo_try_verify_hosts code.
5159 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5160 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5161 case-independent, so other case variants are also recognized). Apparently
5162 some systems use these upper case variants.
5164 25. If more than two messages were waiting for the same host, and a transport
5165 filter was specified for the transport, Exim sent two messages over the
5166 same TCP/IP connection, and then failed with "socket operation on non-
5167 socket" when it tried to send the third.
5169 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5171 27. The extern definition of crypt16() in expand.c was not being excluded when
5172 the OS had its own crypt16() function.
5174 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5175 as a preferred synonym for return_size_limit, both as an option and as an
5178 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5180 30. Changed the default configuration ACL to relax the local part checking rule
5181 for addresses that are not in any local domains. For these addresses,
5182 slashes and pipe symbols are allowed within local parts, but the sequence
5183 /../ is explicitly forbidden.
5185 31. SPA server authentication was not clearing the challenge buffer before
5188 32. log_message in a "warn" ACL statement was writing to the reject log as
5189 well as to the main log, which contradicts the documentation and doesn't
5190 seem right (because no rejection is happening). So I have stopped it.
5192 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5193 However, I am unable to do any testing of this.
5195 34. Fixed an infelicity in the appendfile transport. When checking directories
5196 for a mailbox, to see if any needed to be created, it was accidentally
5197 using path names with one or more superfluous leading slashes; tracing
5198 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5200 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5201 contents are added to the log line that is written for every discarded
5202 recipient. (Previously a log_message setting was ignored.)
5204 36. The ${quote: operator now quotes the string if it is empty.
5206 37. The install script runs exim in order to find its version number. If for
5207 some reason other than non-existence or emptiness, which it checks, it
5208 could not run './exim', it was installing it with an empty version number,
5209 i.e. as "exim-". This error state is now caught, and the installation is
5212 38. An argument was missing from the function that creates an error message
5213 when Exim fails to connect to the socket for saslauthd authentication.
5214 This could cause Exim to crash, or give a corrupted message.
5216 39. Added isip, isip4, and isip6 to ${if conditions.
5218 40. The ACL variables $acl_xx are now saved with the message, and can be
5219 accessed later in routers, transports, and filters.
5221 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5222 strings in the file are not string-expanded.
5224 42. If a MAIL command specified a SIZE value that was too large to fit into an
5225 int variable, the check against message_size_limit failed. Such values are
5226 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5227 day this will have to be increased, but I don't think I want to be around
5228 when emails are that large.
5235 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5236 iconv() is not standard in FreeBSD.
5238 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5239 IPv6 enabled. The observed symptom was a segmentation fault on return from
5240 the function os_common_find_running_interfaces() in src/os.c.
5242 3. In the check_special_case() function in daemon.c I had used "errno" as an
5243 argument name, which causes warnings on some systems. This was basically a
5244 typo, since it was named "eno" in the comments!
5246 4. The code that waits for the clock to tick (at a resolution of some fraction
5247 of a second) so as to ensure message-id uniqueness was always waiting for
5248 at least one whole tick, when it could have waited for less. [This is
5249 almost certainly not relevant at current processor speeds, where it is
5250 unlikely to ever wait at all. But we try to future-proof.]
5252 5. The function that sleeps for a time interval that includes fractions of a
5253 second contained a race. It did not block SIGALRM between setting the
5254 timer, and suspending (a couple of lines later). If the interval was short
5255 and the sigsuspend() was delayed until after it had expired, the suspension
5256 never ended. On busy systems this could lead to processes getting stuck for
5259 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5260 process, before it forks any delivery processes. The open lookup caching
5261 mechanism meant that the open file or database connection was passed into
5262 the delivery process. The problem was that delivery processes always tidy
5263 up cached lookup data. This could cause a problem for the next delivery
5264 process started by the queue runner, because the external queue runner
5265 process does not know about the closure. So the next delivery process
5266 still has data in the lookup cache. In the case of a file lookup, there was
5267 no problem because closing a file descriptor in a subprocess doesn't affect
5268 the parent. However, if the lookup was caching a connection to a database,
5269 the connection was closed, and the second delivery process was likely to
5270 see errors such as "PGSQL: query failed: server closed the connection
5271 unexpectedly". The problem has been fixed by closing all cached lookups
5272 in a queue runner before running a delivery process.
5274 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5275 seem to have the "const" qualifier which it has on other OS. I've
5278 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5279 *of the same type* (client or server) with the same public name that an
5280 error should be diagnosed.
5282 9. When Exim looked up a host name for an IP address, but failed to find the
5283 original IP address when looking up the host name (a safety check), it
5284 output the message "<ip address> does not match any IP for NULL", which was
5285 confusing, to say the least. The bug was that the host name should have
5286 appeared instead of "NULL".
5288 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5289 user that is built into the binary, and the -C or -D options is used, root
5290 privilege is dropped before the configuration file is read. In addition,
5291 logging is switched to stderr instead of the normal log files. If the
5292 configuration then re-defines the Exim user, the unprivileged environment
5293 is probably not what is expected, so Exim logs a panic warning message (but
5296 However, if deliver_drop_privilege is set, the unprivileged state may well
5297 be exactly what is intended, so the warning has been cut out in that case,
5298 and Exim is allowed to try to write to its normal log files.
5304 1. smtp_return_error_details was not giving details for temporary sender
5305 or receiver verification errors.
5307 2. Diagnose a configuration error if two authenticators have the same public
5310 3. Exim used not to create the message log file for a message until the first
5311 delivery attempt. This could be confusing when incoming messages were held
5312 for policy or load reasons. The message log file is now created at the time
5313 the message is received, and an initial "Received" line is written to it.
5315 4. The automatically generated man page for command line options had a minor
5316 bug that caused no ill effects; however, a more serious problem was that
5317 the procedure for building the man page automatically didn't always
5318 operate. Consequently, release 4.20 contains an out-of-date version. This
5319 shouldn't happen again.
5321 5. When building Exim with embedded Perl support, the script that builds the
5322 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5323 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5325 6. The freeze_tell option was not being used for messages that were frozen on
5326 arrival, either by an ACL or by local_scan().
5328 7. Added the smtp_incomplete_transaction log selector.
5330 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5331 was accepting AUTH without a new EHLO.
5333 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5334 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5337 10. Logging of TCP/IP connections (when configured) now happens in the main
5338 daemon process instead of the child process, so that the TCP/IP connection
5339 count is more accurate (but it can never be perfect).
5341 11. The use of "drop" in a nested ACL was not being handled correctly in the
5342 outer ACL. Now, if condition failure induced by the nested "drop" causes
5343 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5344 or "require"), the connection is dropped.
5346 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5347 that yield "discard" can now be used with an "accept" or a "discard" verb,
5348 but an error is generated for any others (because I can't see a useful way
5349 to define what should happen).
5351 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5352 are now processed in the same way as lines in the Exim configuration file.
5353 In particular, continuation lines are supported.
5355 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5357 15. Added -ti meaning -t -i.
5359 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5360 domains, and warn by logging if others are found.
5362 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5363 returned from GIFCONF when Exim is trying to find the list of interfaces on
5364 a host. The code in os.c has been modified to copy each ifreq to an aligned
5365 structure in all cases.
5367 Also, in some cases, the returned ifreq's were being copied to a 'struct
5368 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5369 means the last couple of bytes of an IPv6 address could be chopped if the
5370 ifreq contained only a normal sockaddr (14 bytes storage).
5372 18. Named domain lists were not supported in the hosts_treat_as_local option.
5373 An entry such as +xxxx was not recognized, and was treated as a literal
5376 19. Ensure that header lines added by a DATA ACL are included in the reject log
5377 if the ACL subsequently rejects the message.
5379 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5380 MD5 (which is deprecated).
5382 21. When testing a filter file using -bf, Exim was writing a message when it
5383 took the sender from a "From " line in the message, but it was not doing so
5384 when it took $return_path from a Return-Path: header line. It now does.
5386 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5387 with a valid header line field name (a series of printing characters
5388 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5390 23. Changed "disc" in the source to "disk" to conform to the documentation and
5391 the book and for uniformity.
5393 24. Ignore Sendmail's -Ooption=value command line item.
5395 25. When execve() failed while trying to run a command in a pipe transport,
5396 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5397 could be confused with a return value of 69 from the command itself. This
5398 has been changed to 127, the value the shell returns if it is asked to run
5399 a non-existent command. The wording for the related log line suggests a
5400 non-existent command as the problem.
5402 26. If received_header_text expands to an empty string, do not add a Received:
5403 header line to the message. (Well, it adds a token one on the spool, but
5404 marks it "old" so that it doesn't get used or transmitted.)
5406 27. Installed eximstats 1.28 (addition of -nt option).
5408 28. There was no check for failure on the call to getsockname() in the daemon
5409 code. This can fail if there is a shortage of resources on the system, with
5410 ENOMEM, for example. A temporary error is now given on failure.
5412 29. Contrary to the C standard, it seems that in some environments, the
5413 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5414 program. Exim now does this explicitly; it affects the formatting of
5415 timestamps using strftime().
5417 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5418 complaints. I've now initialized all the variables, to avoid this.
5420 32. Header lines added by a system filter were not being "seen" during
5421 transport-time rewrites.
5423 33. The info_callback() function passed to OpenSSL is set up with type void
5424 (*)(SSL *, int, int), as described somewhere. However, when calling the
5425 function (actually a macro) that sets it up, the type void(*)() is
5426 expected. I've put in a cast to prevent warnings from picky compilers.
5428 34. If a DNS black list lookup found a CNAME record, but there were no A
5429 records associated with the domain it pointed at, Exim crashed.
5431 35. If a DNS black list lookup returned more than one A record, Exim ignored
5432 all but the first. It now scans all returned addresses if a particular IP
5433 value is being sought. In this situation, the contents of the
5434 $dnslist_value variable are a list of all the addresses, separated by a
5437 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5438 to accept a host name and all its aliases if the forward lookup for any of
5439 them yielded the IP address of the incoming connection. Now it accepts only
5440 those names whose forward lookup yields the correct IP address. Any other
5441 names are discarded. This closes a loophole whereby a rogue DNS
5442 administrator could create reverse DNS records to break through a
5443 wildcarded host restriction in an ACL.
5445 37. If a user filter or a system filter that ran in a subprocess used any of
5446 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5447 the wrong values were passed to the pipe command ($thisaddress had the
5448 value of $0, $0 had the value of $1, etc). This bug was introduced by
5449 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5451 38. Improved the line breaking for long SMTP error messages from ACLs.
5452 Previously, if there was no break point between 40 and 75 characters, Exim
5453 left the rest of the message alone. Two changes have been made: (a) I've
5454 reduced the minimum length to 35 characters; (b) if it can't find a break
5455 point between 35 and 75 characters, it looks ahead and uses the first one
5456 that it finds. This may give the occasional overlong line, but at least the
5457 remaining text gets split now.
5459 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5460 file descriptors might be low, and that setting 1000 would always raise it.
5461 It turns out that in some environments, the limit is already over 1000 and
5462 that lowering it causes trouble. So now Exim takes care not to decrease it.
5464 40. When delivering a message, the value of $return_path is set to $sender_
5465 address at the start of routing (routers may change the value). By an
5466 oversight, this default was not being set up when an address was tested by
5467 -bt or -bv, which affected the outcome if any router or filter referred to
5470 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5471 log only when "message" or "log_message" are set. However, if one of the
5472 conditions was an address verification, or a call to a nested ACL, the
5473 messages generated by the underlying test were being passed through. This
5474 no longer happens. The underlying message is available in $acl_verify_
5475 message for both "message" and "log_message" expansions, so it can be
5476 passed through if needed.
5478 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5479 new expansion $bh_ to give the encoded byte string without charset
5480 translation. Translation happens only if iconv() is available; HAVE_ICONV
5481 indicates this at build time. HEADERS_CHARSET gives the charset to
5482 translate to; headers_charset can change it in the configuration, and
5483 "headers charset" can change it in an individual filter file.
5485 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5486 that creates RFC 2047 encoded "words" labels them as that charset instead
5487 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5488 expansion operator; (ii) when Exim creates a From: line for a local
5489 message; (iii) when a header line is rewritten to include a "phrase" part.
5491 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5492 buggy, causing it to skip the first lines of messages whose message ID
5493 ended in 'D'. This would not have bitten before Exim release 4.14, because
5494 message IDs were unlikely to end in 'D' before then. The effect was to have
5495 incorrect size information for certain domains.
5497 45. #include "config.h" was missing at the start of the crypt16.c module. This
5498 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5501 46. If there was a timeout during a "random" callout check, Exim treated it as
5502 a failure of the random address, and carried on sending RSET and the real
5503 address. If the delay was just some slowness somewhere, the response to the
5504 original RCPT would be taken as a response to RSET and so on, causing
5505 mayhem of various kinds.
5507 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5508 when I implemented it. It didn't allow for the fact that some option values
5509 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5510 the right test for positive values.
5512 48. Domain names in DNS records are case-independent. Exim always looks them up
5513 in lower case. Some resolvers return domain names in exactly the case they
5514 appear in the zone file, that is, they may contain uppercase letters. Not
5515 all resolvers do this - some return always lower case. Exim was treating a
5516 change of case by a resolver as a change of domain, similar to a widening
5517 of a domain abbreviation. This triggered its re-routing code and so it was
5518 trying to route what was effectively the same domain again. This normally
5519 caused routing to fail (because the router wouldn't handle the domain
5520 twice). Now Exim checks for this case specially, and just changes the
5521 casing of the domain that it ultimately uses when it transmits the message
5524 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5527 50. If a filter generated a file delivery with a non-absolute name (possible if
5528 no home directory exists for the router), the forbid_file option was not
5531 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5532 the existing equality matching.
5534 52. Exim was using ints instead of ino_t variables in some places where it was
5535 dealing with inode numbers.
5537 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5538 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5539 called TMPDIR, and if it finds it is different, it changes its value.
5541 54. The smtp_printf() function is now made available to local_scan() so
5542 additional output lines can be written before returning. There is also an
5543 smtp_fflush() function to enable the detection of a dropped connection.
5544 The variables smtp_input and smtp_batched_input are exported to
5547 55. Changed the default runtime configuration: the message "Unknown user"
5548 has been removed from the ACL, and instead placed on the localuser router,
5549 using the cannot_route_message feature. This means that any verification
5550 failures that generate their own messages won't get overridden. Similarly,
5551 the "Unrouteable address" message that was in the ACL for unverifiable
5552 relay addresses has also been removed.
5554 56. Added hosts_avoid_esmtp to the smtp transport.
5556 57. The exicyclog script was not checking for the esoteric option
5557 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5558 will work only if exicyclog is run under the appropriate euid.
5560 58. Following a discussion on the list, the rules by which Exim recognises line
5561 endings on incoming messages have been changed. The -dropcr and drop_cr
5562 options are now no-ops, retained only for backwards compatibility. The
5563 following line terminators are recognized: LF CRLF CR. However, special
5564 processing applies to CR:
5566 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5567 nor a local message in the state where . is a terminator.
5569 (ii) If a bare CR is encountered in a header line, an extra space is added
5570 after the line terminator so as not to end the header. The reasoning
5571 behind this is that bare CRs in header lines are most likely either
5572 to be mistakes, or people trying to play silly games.
5574 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5575 was being incorrectly given as 18 bytes larger than it should have been.
5576 This is a VOB (very old bug).
5578 60. This may never have affected anything current, but just in case it has:
5579 When the local host is found other than at the start of a list of hosts,
5580 the local host, those with the same MX, and any that follow, are discarded.
5581 When the list in question was part of a longer list of hosts, the following
5582 hosts (not currently being processed) were also being discarded. This no
5583 longer happens. I'm not sure if this situation could ever has previously
5586 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5589 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5590 header. This is something that is recommended in a new Internet Draft, and
5591 is something that is documented as being done by Sendmail. There are two
5592 possible values. For messages generated by the autoreply transport, Exim
5595 Auto-Submitted: auto-replied
5597 whereas for all other generated messages (e.g. bounces) it adds
5599 Auto-Submitted: auto-generated
5601 63. The "personal" condition in filters now includes a test for the
5602 Auto-Submitted: header. If it contains the string "auto-" the message it
5603 not considered personal.
5605 64. Added rcpt_include_affixes as a generic transport option.
5607 65. Added queue_only_override (default true).
5609 66. Added the syslog_duplication option.
5611 67. If what should have been the first header line of a message consisted of
5612 a space followed by a colon, Exim was mis-interpreting it as a header line.
5613 It isn't of course - it is syntactically invalid and should therefore be
5614 treated as the start of the message body. The misbehaviour could have
5615 caused a number of strange effects, including loss of data in subsequent
5616 header lines, and spool format errors.
5618 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5619 client host had authenticated. This control can now be exercised by an ACL
5620 for more flexibility.
5622 69. By default, callouts do not happen when testing with -bh. There is now a
5623 variant, -bhc, which does actually run the callout code, including
5624 consulting and updating the callout cache.
5626 70. Added support for saslauthd authentication, courtesy of Alexander
5629 71. If statvfs() failed on the spool or log directories while checking their
5630 size for availability, Exim confusingly gave the error "space shortage".
5631 Furthermore, in debugging mode it crashed with a floating point exception.
5632 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5633 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5634 serious problem, Exim now writes to the main and panic logs when this
5635 happens, with details of the failure. It then refuses to accept the
5636 incoming message, giving the message "spool directory problem" or "log
5637 directory problem" with a 421 code for SMTP messages.
5639 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5640 0, 1, and 2 exist, because some OS complain for execs without them (see
5641 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5642 descriptors. However, the code omitted to check that the open succeeded,
5643 causing mysterious errors if for some reason the permissions on /dev/null
5644 got screwed. Now Exim writes a message to the main and panic logs, and
5645 bombs out if it can't open /dev/null.
5647 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5648 interact so that it is all more flexible. It is supposed to remain
5649 backwards compatible. Also added extra_local_interfaces.
5651 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5652 to bomb out with an assertion failure - to the client this appears as a
5653 connection drop. This problem occurs in the part of the code that was taken
5654 from the Samba project. Fortunately, the assertion is in a very simple
5655 function, so I have fixed this by reproducing the function inline in the
5656 one place where it is called, and arranging for authentication to fail
5657 instead of killing the process with assert().
5659 75. The SPA client code was not working when the server requested OEM rather
5660 than Unicode encoding.
5662 76. Added code to make require_files with a specific uid setting more usable in
5663 the case where statting the file as root fails - usually a non-root-mounted
5664 NFS file system. When this happens and the failure is EACCES, Exim now
5665 forks a subprocess and does the per-uid checking as the relevant uid.
5667 77. Added process_log_path.
5669 78. If log_file_path was not explicitly set, a setting of check_log_space or
5670 check_log_inodes was ignored.
5672 79. If a space check for the spool or log partitions fails, the incident is now
5673 logged. Of course, in the latter case the data may get lost...
5675 80. Added the %p formatting code to string_format() so that it can be used to
5676 print addresses in debug_print(). Adjusted all the address printing in the
5677 debugging in store.c to use %p rather than %d.
5679 81. There was a concern that a line of code in smtp_in.c could overflow a
5680 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5681 initially expressed, the concern was not well-founded, because trailing
5682 spaces are removed early. However, if the trailing spaces were followed by
5683 a NULL, they did not get removed, so the overflow was possible. Two fixes
5686 (a) I re-wrote the offending code in a cleaner fashion.
5687 (b) If an incoming SMTP command contains a NULL character, it is rejected
5690 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5691 runs initgroups(), so that if the Exim user is in any additional groups,
5692 they will be used during message reception.
5698 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob