1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
9 PP/02 Make -n do something, by making it not do something.
10 When combined with -bP, the name of an option is not output.
12 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
15 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
16 $sender_host_name and config options to manage this, and basic check
19 PP/05 DSCP support for outbound connections and control modifier for inbound.
21 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
22 (Only plugin which currently uses this is kerberos4, which nobody should
23 be using, but we should make it available and other future plugins might
24 conceivably use it, even though it would break NAT; stuff *should* be
25 using channel bindings instead).
27 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
28 name; added for Sendmail compatibility; requires admin caller.
29 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
30 just ignore it); requires trusted caller.
31 Also parse but ignore: -Ac -Am -X<logfile>
34 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
36 TL/02 Add +smtp_confirmation as a default logging option.
38 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
39 Patch by Magnus Holmgren from 2007-02-20.
41 TL/04 Bugzilla 1281 - Spec typo.
43 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
45 JH/02 Support "G" suffix to numbers in ${if comparisons.
47 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
49 NM/01 Bugzilla 1197 - Spec typo
50 Bugzilla 1196 - Spec examples corrections
52 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
54 PP/09 Add gnutls_enable_pkcs11 option.
56 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
57 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
59 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
60 "acl {{name}{arg}...}", and optional args on acl condition
63 JH/05 Permit multiple router/transport headers_add/remove lines.
68 PP/01 Handle short writes when writing local log-files.
69 In practice, only affects FreeBSD (8 onwards).
70 Bugzilla 1053, with thanks to Dmitry Isaikin.
72 NM/01 Bugzilla 949 - Documentation tweak
74 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
77 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
79 PP/02 Implemented gsasl authenticator.
81 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
83 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
84 `pkg-config foo` for cflags/libs.
86 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
87 with rest of GSASL and with heimdal_gssapi.
89 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
90 `pkg-config foo` for cflags/libs for the TLS implementation.
92 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
93 properties get this fed in as external SSF. A number of robustness
94 and debugging improvements to the cyrus_sasl authenticator.
96 PP/08 cyrus_sasl server now expands the server_realm option.
98 PP/09 Bugzilla 1214 - Log authentication information in reject log.
99 Patch by Jeremy Harris.
101 PP/10 Added dbmjz lookup type.
103 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
105 PP/12 MAIL args handles TAB as well as SP, for better interop with
106 non-compliant senders.
107 Analysis and variant patch by Todd Lyons.
109 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
110 Bug report from Lars Müller <lars@samba.org> (via SUSE),
111 Patch from Dirk Mueller <dmueller@suse.com>
113 PP/13 tls_peerdn now print-escaped for spool files.
114 Observed some $tls_peerdn in wild which contained \n, which resulted
115 in spool file corruption.
117 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
118 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
119 or write after TLS renegotiation, which otherwise led to messages
122 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
123 as a tracking header (ie: a signed header comes before the signature).
124 Patch from Wolfgang Breyha.
126 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
127 comma-sep list; embedded commas doubled.
129 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
131 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
133 Report and patch from Dmitry Banschikov.
135 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
136 Removed SSL_clear() after SSL_new() which led to protocol negotiation
137 failures. We appear to now support TLS1.1+ with Exim.
139 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
140 lets Exim select keys and certificates based upon TLS SNI from client.
141 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
142 before an outbound SMTP session. New log_selector, +tls_sni.
144 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
145 NULL dereference. Report and patch from Alun Jones.
147 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
148 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
149 Not seeing resolver debug output on NetBSD, but suspect this is a
150 resolver implementation change.
152 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
153 Left warnings. Added "eximon gdb" invocation mode.
155 PP/21 Defaulting "accept_8bitmime" to true, not false.
157 PP/22 Added -bw for inetd wait mode support.
159 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
160 locate the relevant includes and libraries. Made this the default.
162 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
163 Bugzilla 1246, report and most of solution from Tomasz Kusy.
165 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
166 This may cause build issues on older platforms.
168 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
169 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
170 gnutls_require_mac & gnutls_require_protocols (no longer supported).
171 Added SNI support via GnuTLS too.
172 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
174 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
176 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
177 Applied second patch from Janne, implementing suggestion to default
178 multiple-strings-in-record handling to match SPF spec.
180 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
182 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
183 read-only, out of scope).
184 Patch from Wolfgang Breyha, report from Stuart Northfield.
186 PP/29 Fix three issues highlighted by clang analyser static analysis.
187 Only crash-plausible issue would require the Cambridge-specific
188 iplookup router and a misconfiguration.
189 Report from Marcin Mirosław.
191 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
193 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
194 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
195 As part of this, removing so much warning spew let me fix some minor
196 real issues in debug logging.
198 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
199 assignment on my part. Fixed.
201 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
202 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
203 Janne Snabb (who went above and beyond: thank you).
205 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
206 string otherwise requires a connection and a bunch more work and it's
207 relatively easy to get wrong. Should also expose TLS library linkage
210 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
211 64-bit ${eval} (JH/03).
213 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
214 GNU libc to support some of the 64-bit stuff, should not lead to
215 conflicts. Defined before os.h is pulled in, so if a given platform
216 needs to override this, it can.
218 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
219 protection layer was required, which is not implemented.
220 Bugzilla 1254, patch from Wolfgang Breyha.
222 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
223 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
224 tls_dhparam take prime identifiers. Also unbreak combination of
225 OpenSSL+DH_params+TLSSNI.
227 PP/39 Disable SSLv2 by default in OpenSSL support.
233 PP/01 Solaris build fix for Oracle's LDAP libraries.
234 Bugzilla 1109, patch from Stephen Usher.
236 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
238 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
241 TF/02 Fix a couple more cases where we did not log the error message
242 when unlink() failed. See also change 4.74-TF/03.
244 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
245 lock up or crash if it happened to be inside a call to libc when it
246 got a SIGUSR1 from exiwhat.
248 The SIGUSR1 handler appends the current process status to the process
249 log which is later printed by exiwhat. It used to use the general
250 purpose logging code to do this, but several functions it calls are
251 not safe for signals.
253 The new output code in the SIGUSR1 handler is specific to the process
254 log, and simple enough that it's easy to inspect for signal safety.
255 Removing some special cases also simplifies the general logging code.
256 Removing the spurious timestamps from the process log simplifies
259 TF/04 Improved ratelimit ACL condition.
261 The /noupdate option has been deprecated in favour of /readonly which
262 has clearer semantics. The /leaky, /strict, and /readonly update modes
263 are mutually exclusive. The update mode is no longer included in the
264 database key; it just determines when the database is updated. (This
265 means that when you upgrde Exim will forget old rate measurements.)
267 Exim now checks that the per_* options are used with an update mode that
268 makes sense for the current ACL. For example, when Exim is processing a
269 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
270 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
271 must specify per_mail/readonly. If you omit the update mode it defaults to
272 /leaky where that makes sense (as before) or /readonly where required.
274 The /noupdate option is now undocumented but still supported for
275 backwards compatibility. It is equivalent to /readonly except that in
276 ACLs where /readonly is required you may specify /leaky/noupdate or
277 /strict/noupdate which are treated the same as /readonly.
279 A useful new feature is the /count= option. This is a generalization
280 of the per_byte option, so that you can measure the throughput of other
281 aggregate values. For example, the per_byte option is now equivalent
282 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
284 The per_rcpt option has been generalized using the /count= mechanism
285 (though it's more complicated than the per_byte equivalence). When it is
286 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
287 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
288 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
289 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
290 Note that using per_rcpt with a non-readonly update mode in more than
291 one ACL will cause the recipients to be double-counted. (The per_mail
292 and per_byte options don't have this problem.)
294 The handling of very low rates has changed slightly. If the computed rate
295 is less than the event's count (usually one) then this event is the first
296 after a long gap. In this case the rate is set to the same as this event's
297 count, so that the first message of a spam run is counted properly.
299 The major new feature is a mechanism for counting the rate of unique
300 events. The new per_addr option counts the number of different
301 recipients that someone has sent messages to in the last time period. It
302 behaves like per_rcpt if all the recipient addresses are different, but
303 duplicate recipient addresses do not increase the measured rate. Like
304 the /count= option this is a general mechanism, so the per_addr option
305 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
306 example, measure the rate that a client uses different sender addresses
307 with the options per_mail/unique=$sender_address. There are further
308 details in the main documentation.
310 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
312 TF/06 Removed a few PCRE remnants.
314 TF/07 Automatically extract Exim's version number from tags in the git
315 repository when doing development or release builds.
317 PP/02 Raise smtp_cmd_buffer_size to 16kB.
318 Bugzilla 879. Patch from Paul Fisher.
320 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
321 Heavily based on revision 40f9a89a from Simon Arlott's tree.
324 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
326 PP/05 Variable $av_failed, true if the AV scanner deferred.
327 Bugzilla 1078. Patch from John Horne.
329 PP/06 Stop make process more reliably on build failure.
330 Bugzilla 1087. Patch from Heiko Schlittermann.
332 PP/07 Make maildir_use_size_file an _expandable_ boolean.
333 Bugzilla 1089. Patch from Heiko Schlittermann.
335 PP/08 Handle ${run} returning more data than OS pipe buffer size.
336 Bugzilla 1131. Patch from Holger Weiß.
338 PP/09 Handle IPv6 addresses with SPF.
339 Bugzilla 860. Patch from Wolfgang Breyha.
341 PP/10 GnuTLS: support TLS 1.2 & 1.1.
343 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
346 PP/11 match_* no longer expand right-hand-side by default.
347 New compile-time build option, EXPAND_LISTMATCH_RHS.
348 New expansion conditions, "inlist", "inlisti".
350 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
352 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
354 PP/14 fix log_write() format string regression from TF/03.
355 Bugzilla 1152. Patch from Dmitry Isaikin.
361 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
363 PP/02 Harmonised TLS library version reporting; only show if debugging.
364 Layout now matches that introduced for other libraries in 4.74 PP/03.
366 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
368 PP/04 New "dns_use_edns0" global option.
370 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
373 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
374 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
376 TK/01 Updated PolarSSL code to 0.14.2.
377 Bugzilla 1097. Patch from Andreas Metzler.
379 PP/07 Catch divide-by-zero in ${eval:...}.
382 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
385 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
386 format-string attack -- SECURITY: remote arbitrary code execution.
388 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
389 time unintentionally subject to list matching rules, letting the header
390 cause arbitrary Exim lookups (of items which can occur in lists, *not*
391 arbitrary string expansion). This allowed for information disclosure.
393 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
394 INT_MIN/-1 -- value coerced to INT_MAX.
400 NM/01 Workround for PCRE version dependency in version reporting
403 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
404 This fixes portability to compilers other than gcc, notably
405 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
407 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
408 makefiles for portability to HP-UX and POSIX correctness.
410 PP/01 Permit LOOKUP_foo enabling on the make command-line.
411 Also via indented variable definition in the Makefile.
412 (Debugging by Oliver Heesakkers).
414 PP/02 Restore caching of spamd results with expanded spamd_address.
415 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
417 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
418 Improves build reliability. Fix from: Frank Elsner
420 NM/02 Fix wide character breakage in the rfc2047 coding
421 Fixes bug 1064. Patch from Andrey N. Oktyabrski
423 NM/03 Allow underscore in dnslist lookups
424 Fixes bug 1026. Patch from Graeme Fowler
426 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
427 Code patches from Adam Ciarcinski of NetBSD.
429 NM/04 Fixed exiqgrep to cope with mailq missing size issue
432 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
433 is logged, to avoid truncation. Patch from John Horne.
435 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
436 Patch from Jakob Hirsch.
438 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
439 SQL string expansion failure details.
440 Patch from Andrey Oktyabrski.
442 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
443 Patch from Simon Arlott.
445 PP/09 New lookups functionality failed to compile on old gcc which rejects
446 extern declarations in function scope.
447 Patch from Oliver Fleischmann
449 PP/10 Use sig_atomic_t for flags set from signal handlers.
450 Check getgroups() return and improve debugging.
451 Fixed developed for diagnosis in bug 927 (which turned out to be
454 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
455 Patch from Mark Zealey.
457 PP/12 Bugzilla 1056: Improved spamd server selection.
458 Patch from Mark Zealey.
460 PP/13 Bugzilla 1086: Deal with maildir quota file races.
461 Based on patch from Heiko Schlittermann.
463 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
464 Patch from Uwe Doering, sign-off by Michael Haardt.
466 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
467 variable declaration deep within a block. Bug and patch from
470 PP/15 lookups-Makefile IRIX compatibilty coercion.
472 PP/16 Make DISABLE_DKIM build knob functional.
474 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
475 Patch by Simon Arlott
477 TF/03 Fix valgrind.h portability to C89 compilers that do not support
478 variable argument macros. Our copy now differs from upstream.
484 TF/01 Failure to get a lock on a hints database can have serious
485 consequences so log it to the panic log.
487 TF/02 Log LMTP confirmation messages in the same way as SMTP,
488 controlled using the smtp_confirmation log selector.
490 TF/03 Include the error message when we fail to unlink a spool file.
492 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
493 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
494 for maintaining out-of-tree patches for some time.
496 PP/01 Bugzilla 139: Documentation and portability issues.
497 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
498 Handle per-OS dynamic-module compilation flags.
500 PP/02 Let /dev/null have normal permissions.
501 The 4.73 fixes were a little too stringent and complained about the
502 permissions on /dev/null. Exempt it from some checks.
503 Reported by Andreas M. Kirchwitz.
505 PP/03 Report version information for many libraries, including
506 Exim version information for dynamically loaded libraries. Created
507 version.h, now support a version extension string for distributors
508 who patch heavily. Dynamic module ABI change.
510 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
511 privilege escalation vulnerability whereby the Exim run-time user
512 can cause root to append content of the attacker's choosing to
515 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
518 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
519 If dropping privileges for untrusted macros, we disabled normal logging
520 on the basis that it would fail; for the Exim run-time user, this is not
521 the case, and it resulted in successful deliveries going unlogged.
522 Fixed. Reported by Andreas Metzler.
528 PP/01 Date: & Message-Id: revert to normally being appended to a message,
529 only prepend for the Resent-* case. Fixes regression introduced in
530 Exim 4.70 by NM/22 for Bugzilla 607.
532 PP/02 Include check_rfc2047_length in configure.default because we're seeing
533 increasing numbers of administrators be bitten by this.
535 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
537 PP/03 Bugzilla 994: added openssl_options main configuration option.
539 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
541 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
543 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
545 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
546 without a peer certificate, leading to a segfault because of an
547 assumption that peers always have certificates. Be a little more
548 paranoid. Problem reported by Martin Tscholak.
550 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
551 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
552 NB: ClamAV planning to remove STREAM in "middle of 2010".
553 CL also introduces -bmalware, various -d+acl logging additions and
554 more caution in buffer sizes.
556 PP/09 Implemented reverse_ip expansion operator.
558 PP/10 Bugzilla 937: provide a "debug" ACL control.
560 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
562 PP/12 Bugzilla 973: Implement --version.
564 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
566 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
568 PP/15 Bugzilla 816: support multiple condition rules on Routers.
570 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
571 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
572 ignore trailing whitespace.
574 JJ/02 prevent non-panic DKIM error from being sent to paniclog
576 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
579 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
580 Notification from Dr Andrew Aitchison.
582 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
583 ExtendedDetectionInfo response format.
584 Notification from John Horne.
586 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
589 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
590 XSL and documented dependency on system catalogs, with examples of how
593 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
596 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
597 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
598 configuration file which is writeable by the Exim user or group.
600 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
601 of configuration files to cover files specified with the -C option if
602 they are going to be used with root privileges, not just the default
605 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
606 option (effectively making it always true).
608 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
609 files to be used while preserving root privileges.
611 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
612 that rogue child processes cannot use them.
614 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
615 run-time user, instead of root.
617 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
618 Exim run-time user without dropping privileges.
620 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
621 result string, instead of calling string_vformat() twice with the same
624 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
625 for other users. Others should always drop root privileges if they use
626 -C on the command line, even for a whitelisted configure file.
628 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
630 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
636 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
637 $data_path, and $header_path variables; fixed documentation bugs and
640 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
641 exipick to access non-standard spools, including the "frozen" queue
644 NM/01 Bugzilla 965: Support mysql stored procedures.
645 Patch from Alain Williams
647 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
649 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
650 Patch from Andreas Metzler
652 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
653 Patch from Kirill Miazine
655 NM/05 Bugzilla 671: Added umask to procmail example.
657 JJ/03 installed exipick 20100323.0, fixing doc bug
659 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
660 directory. Notification and patch from Dan Rosenberg.
662 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
664 TK/02 Improve log output when DKIM signing operation fails.
666 MH/01 Treat the transport option dkim_domain as a colon separated
667 list, not as a single string, and sign the message with each element,
668 omitting multiple occurences of the same signer.
670 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
671 Bugzilla 985, 986. Patch by Simon Arlott
673 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
674 Patch by Simon Arlott
676 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
677 MBX locking. Notification from Dan Rosenberg.
683 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
685 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
687 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
689 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
691 NM/04 Bugzilla 915: Flag broken perl installation during build.
697 TK/01 Added patch by Johannes Berg that expands the main option
698 "spamd_address" if it starts with a dollar sign.
700 TK/02 Write list of recipients to X-Envelope-Sender header when building
701 the mbox-format spool file for content scanning (suggested by Jakob
704 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
705 (http://www.dcc-servers.net/) support via dccifd. Activated by
706 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
708 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
709 by Mark Daniel Reidel <mr@df.eu>.
711 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
712 When building exim an external PCRE library is now needed -
713 PCRE is a system library on the majority of modern systems.
714 See entry on PCRE_LIBS in EDITME file.
716 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
717 conversation. Added nologin parameter to request.
718 Patch contributed by Kirill Miazine.
720 TF/01 Do not log submission mode rewrites if they do not change the address.
722 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
724 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
725 log files in place. Contributed by Roberto Lima.
727 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
729 TF/03 Bugzilla 615: When checking the local_parts router precondition
730 after a local_part_suffix or local_part_prefix option, Exim now
731 does not use the address's named list lookup cache, since this
732 contains cached lookups for the whole local part.
734 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
735 Robert Millan. Documentation is in experimental-spec.txt.
737 TF/04 Bugzilla 668: Fix parallel build (make -j).
739 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
741 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
742 Patch provided by Jan Srzednicki.
744 TF/05 Leading white space used to be stripped from $spam_report which
745 wrecked the formatting. Now it is preserved.
747 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
748 that they are available at delivery time.
750 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
752 TF/08 TLS error reporting now respects the incoming_interface and
753 incoming_port log selectors.
755 TF/09 Produce a more useful error message if an SMTP transport's hosts
756 setting expands to an empty string.
758 NM/06 Bugzilla 744: EXPN did not work under TLS.
759 Patch provided by Phil Pennock.
761 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
762 Patch provided by Richard Godbee.
764 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
765 acl_smtp_notquit, added index entry.
767 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
768 Patch provided by Eugene Bujak.
770 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
771 accept(). Patch provided by Maxim Dounin.
773 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
774 Patch provided by Phil Pennock.
776 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
778 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
779 Patch provided by Brad "anomie" Jorsch.
781 NM/14 Bugzilla 622: Added timeout setting to transport filter.
782 Patch provided by Dean Brooks.
784 TK/05 Add native DKIM support (does not depend on external libraries).
786 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
787 Patch provided by Graeme Fowler.
789 NM/16 Bugzilla 851: Documentation example syntax fix.
791 NM/17 Changed NOTICE file to remove references to embedded PCRE.
793 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
796 NM/19 Bugzilla 745: TLS version reporting.
797 Patch provided by Phil Pennock.
799 NM/20 Bugzilla 167: bool: condition support.
800 Patch provided by Phil Pennock.
802 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
803 clients. Patch provided by Phil Pennock.
805 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
806 Patch provided by Brad "anomie" Jorsch.
808 NM/23 Bugzilla 687: Fix misparses in eximstats.
809 Patch provided by Heiko Schlittermann.
811 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
812 Patch provided by Heiko Schlittermann.
814 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
815 plus update to original patch.
817 NM/26 Bugzilla 799: Documentation correction for ratelimit.
819 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
820 Patch provided by David Brownlee.
822 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
824 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
826 NM/30 Bugzilla 888: TLS documentation bugfixes.
828 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
830 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
831 Unlike the original bugzilla I have changed all shell scripts in src tree.
833 NM/33 Bugzilla 898: Transport filter timeout fix.
834 Patch by Todd Rinaldo.
836 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
837 Patch by Serge Demonchaux.
839 NM/35 Bugzilla 39: Base64 decode bug fixes.
840 Patch by Jakob Hirsch.
842 NM/36 Bugzilla 909: Correct connect() call in dcc code.
844 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
846 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
848 NM/39 Bugzilla 911: Fixed MakeLinks build script.
854 TK/01 Add preliminary DKIM support. Currently requires a forked version of
855 ALT-N's libdkim that I have put here:
856 http://duncanthrax.net/exim-experimental/
858 Note to Michael Haardt: I had to rename some vars in sieve.c. They
859 were called 'true' and it seems that C99 defines that as a reserved
860 keyword to be used with 'bool' variable types. That means you could
861 not include C99-style headers which use bools without triggering
862 build errors in sieve.c.
864 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
865 as mailq or other aliases. Changed the --help handling significantly
866 to do whats expected. exim_usage() emits usage/help information.
868 SC/01 Added the -bylocaldomain option to eximstats.
870 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
872 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
874 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
880 PH/01 Another patch from the Sieve maintainer.
882 PH/02 When an IPv6 address is converted to a string for single-key lookup
883 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
884 dots are used instead of colons so that keys in lsearch files need not
885 contain colons. This was done some time before quoting was made available
886 in lsearch files. However, iplsearch files do require colons in IPv6 keys
887 (notated using the quote facility) so as to distinguish them from IPv4
888 keys. This meant that lookups for IP addresses in host lists did not work
889 for iplsearch lookups.
891 This has been fixed by arranging for IPv6 addresses to be expressed with
892 colons if the lookup type is iplsearch. This is not incompatible, because
893 previously such lookups could never work.
895 The situation is now rather anomolous, since one *can* have colons in
896 ordinary lsearch keys. However, making the change in all cases is
897 incompatible and would probably break a number of configurations.
899 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
902 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
903 conversion specification without a maximum field width, thereby enabling
904 a rogue spamd server to cause a buffer overflow. While nobody in their
905 right mind would setup Exim to query an untrusted spamd server, an
906 attacker that gains access to a server running spamd could potentially
907 exploit this vulnerability to run arbitrary code as the Exim user.
909 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
910 $primary_hostname instead of what libspf2 thinks the hosts name is.
912 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
913 a directory entry by the name of the lookup key. Previously, if a
914 symlink pointed to a non-existing file or a file in a directory that
915 Exim lacked permissions to read, a lookup for a key matching that
916 symlink would fail. Now it is enough that a matching directory entry
917 exists, symlink or not. (Bugzilla 503.)
919 PH/03 The body_linecount and body_zerocount variables are now exported in the
922 PH/04 Added the $dnslist_matched variable.
924 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
925 This means they are set thereafter only if the connection becomes
928 PH/06 Added the client_condition to authenticators so that some can be skipped
929 by clients under certain conditions.
931 PH/07 The error message for a badly-placed control=no_multiline_responses left
932 "_responses" off the end of the name.
934 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
936 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
937 (without spaces) instead of just copying the configuration text.
939 PH/10 Added the /noupdate option to the ratelimit ACL condition.
941 PH/11 Added $max_received_linelength.
943 PH/12 Added +ignore_defer and +include_defer to host lists.
945 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
946 way in which PCRE > 7.0 is built.
948 PH/14 Implemented queue_only_load_latch.
950 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
951 MAIL command. The effect was to mangle the value on 64-bit systems.
953 PH/16 Another patch from the Sieve maintainer.
955 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
957 PH/18 If a system quota error occurred while trying to create the file for
958 a maildir delivery, the message "Mailbox is full" was not appended to the
959 bounce if the delivery eventually timed out. Change 4.67/27 below applied
960 only to a quota excession during the actual writing of the file.
962 PH/19 It seems that peer DN values may contain newlines (and other non-printing
963 characters?) which causes problems in log lines. The DN values are now
964 passed through string_printing() before being added to log lines.
966 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
967 and InterBase are left for another time.)
969 PH/21 Added message_body_newlines option.
971 PH/22 Guard against possible overflow in moan_check_errorcopy().
973 PH/23 POSIX allows open() to be a macro; guard against that.
975 PH/24 If the recipient of an error message contained an @ in the local part
976 (suitably quoted, of course), incorrect values were put in $domain and
977 $local_part during the evaluation of errors_copy.
983 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
984 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
987 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
988 issue a MAIL command.
990 PH/02 In an ACL statement such as
992 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
994 if a client was not listed at all, or was listed with a value other than
995 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
996 the condition was not true (as it should be), so access was not denied.
997 The bug was that the ! inversion was incorrectly passed on to the second
998 item. This has been fixed.
1000 PH/03 Added additional dnslists conditions == and =& which are different from
1001 = and & when the dns lookup returns more than one IP address.
1003 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1004 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1006 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1007 FSYNC, which compiles an option called disable_fsync that allows for
1008 bypassing fsync(). The documentation is heavily laced with warnings.
1010 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1012 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1013 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1014 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1015 including adding "make clean"; (3) Added -fPIC when compiling the test
1016 dynamically loaded module, to get rid of a warning.
1018 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1019 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1020 = 0s. The bug is otherwise harmless.
1022 PH/07 There was a bug in the dovecot authenticator such that the value of
1023 $auth1 could be overwritten, and so not correctly preserved, after a
1024 successful authentication. This usually meant that the value preserved by
1025 the server_setid option was incorrect.
1027 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1029 PH/09 Installed PCRE release 7.0.
1031 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1032 run for batched SMTP input. It is now run at the start of every message
1033 in the batch. While fixing this I discovered that the process information
1034 (output by running exiwhat) was not always getting set for -bs and -bS
1035 input. This is fixed, and it now also says "batched" for BSMTP.
1037 PH/11 Added control=no_pipelining.
1039 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1040 patch, slightly modified), and move the expansion of helo_data till after
1041 the connection is made in the smtp transport (so it can use these
1044 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1046 PH/14 Added log_selector = +pid.
1048 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1050 PH/16 Add ${if forany and ${if forall.
1052 PH/17 Added dsn_from option to vary the From: line in DSNs.
1054 PH/18 Flush SMTP output before performing a callout, unless control =
1055 no_callout_flush is set.
1057 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1058 was true (the default) a successful delivery failed to delete the retry
1059 item, thus causing premature timeout of the address. The bug is now
1062 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1064 PH/21 Long custom messages for fakedefer and fakereject are now split up
1065 into multiline reponses in the same way that messages for "deny" and
1066 other ACL rejections are.
1068 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1069 with slight modification.
1071 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1072 draft, changing the syntax and factoring some duplicate code".
1074 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1075 for deliveries of the second and subsequent messages over the same SMTP
1078 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1079 ${reduce, with only minor "tidies".
1081 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1083 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1084 expansion side effects.
1086 PH/27 When a message times out after an over-quota error from an Exim-imposed
1087 quota, the bounce message says "mailbox is full". This message was not
1088 being given when it was a system quota that was exceeded. It now should
1091 MH/03 Made $recipients available in local_scan(). local_scan() already has
1092 better access to the recipient list through recipients_list[], but
1093 $recipients can be useful in postmaster-provided expansion strings.
1095 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1096 in the case of a MAIL command with additional options following the
1097 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1098 were accidentally chopped off.
1100 PH/29 SMTP synchronization checks are implemented when a command is read -
1101 there is a check that no more input is waiting when there shouldn't be
1102 any. However, for some commands, a delay in an ACL can mean that it is
1103 some time before the response is written. In this time, more input might
1104 arrive, invalidly. So now there are extra checks after an ACL has run for
1105 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1106 pipelining has not been advertised.
1108 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1110 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1111 returned in the SMTP response when smtp_return_error_details was false.
1112 This has been fixed.
1114 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1115 instead of the C I/O that was originally supplied, because problems were
1116 reported on Solaris.
1118 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1119 Exim which did not show up earlier: it was assuming that a call to
1120 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1121 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1122 was a macro that became an assignment, so it seemed to work. This has
1123 changed to a proper function call with a void return, hence the compile
1124 error. Exim's code has been fixed.
1126 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1129 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1131 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1133 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1134 because it turns out to be too big or there is a timeout) there is no
1135 "Completed" line in the log. When some messages of this type were
1136 selected by exigrep, they were listed as "not completed". Others were
1137 picked up by some special patterns. I have improved the selection
1138 criteria to be more general.
1140 PH/38 The host_find_failed option in the manualroute router can now be set
1141 to "ignore", to completely ignore a host whose IP address cannot be
1142 found. If all hosts are ignored, the behaviour is controlled by the new
1143 host_all_ignored option.
1145 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1146 homing or because of multiple MX records with /mx) generated more than
1147 one IP address, and the following item turned out to be the local host,
1148 all the secondary addresses of the first item were incorrectly removed
1149 from the list, along with the local host and any following hosts (which
1150 is what is supposed to happen).
1152 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1153 whoever called Exim into the -H file. In the case of the daemon it was
1154 behaving confusingly. When first started, it used values for whoever
1155 started the daemon, but after a SIGHUP it used the Exim user (because it
1156 calls itself on a restart). I have changed the code so that it now always
1159 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1160 message are rejected with the same error (e.g. no authentication or bad
1161 sender address), and a DATA command is nevertheless sent (as can happen
1162 with PIPELINING or a stupid MUA), the error message that was given to the
1163 RCPT commands is included in the rejection of the DATA command. This is
1164 intended to be helpful for MUAs that show only the final error to their
1167 PH/42 Another patch from the Sieve maintainer.
1169 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1170 Eximstats - Fixed some broken HTML links and added missing column headers
1172 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1173 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1175 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1177 PH/43 Yet another patch from the Sieve maintainer.
1179 PH/44 I found a way to check for a TCP/IP connection going away before sending
1180 the response to the final '.' that terminates a message, but only in the
1181 case where the client has not sent further data following the '.'
1182 (unfortunately, this is allowed). However, in many cases there won't be
1183 any further data because there won't be any more messages to send. A call
1184 to select() can be used: if it shows that the input is "ready", there is
1185 either input waiting, or the socket has been closed. An attempt to read
1186 the next input character can distinguish the two cases. Previously, Exim
1187 would have sent an OK response which the client would never have see.
1188 This could lead to message repetition. This fix should cure that, at
1189 least in a lot of common cases.
1191 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1192 advertised in response to EHLO.
1198 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1199 fixed by 4.65/MH/01 (is this a record?) are fixed:
1201 (i) An empty string was always treated as zero by the numeric comparison
1202 operators. This behaviour has been restored.
1204 (ii) It is documented that the numeric comparison operators always treat
1205 their arguments as decimal numbers. This was broken in that numbers
1206 starting with 0 were being interpreted as octal.
1208 While fixing these problems I realized that there was another issue that
1209 hadn't been noticed. Values of message_size_limit (both the global option
1210 and the transport option) were treated as octal if they started with 0.
1211 The documentation was vague. These values are now always treated as
1212 decimal, and I will make that clear in the documentation.
1218 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1219 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1222 MH/01 Don't check that the operands of numeric comparison operators are
1223 integers when their expansion is in "skipping" mode (fixes bug
1224 introduced by 4.64-PH/07).
1226 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1227 child addresses, Exim now panics and dies. Previously, because the count
1228 is held in a short int, deliveries were likely to be lost. As such a
1229 large number of recipients for a single message is ridiculous
1230 (performance will be very, very poor), I have chosen to impose a limit
1231 rather than extend the field.
1237 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1238 leftover -K file (the existence of which was triggered by #402).
1239 While we were at it, introduced process PID as part of the -K
1240 filename. This should rule out race conditions when creating
1243 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1244 processing considerably. Previous code took too long for large mails,
1245 triggering a timeout which in turn triggers #401.
1247 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1248 in the DK code in transports.c. sendfile() is not really portable,
1249 hence the _LINUX specificness.
1251 TF/01 In the add_headers option to the mail command in an Exim filter,
1252 there was a bug that Exim would claim a syntax error in any
1253 header after the first one which had an odd number of characters
1256 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1257 callout verification, Exim cached a "reject" for the entire domain. This
1258 is correct for most verifications, but it is not correct for a recipient
1259 verification with use_sender or use_postmaster set, because in that case
1260 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1261 case of MAIL FROM:<> rejection from other early rejections (e.g.
1262 rejection of HELO). When verifying a recipient using a non-null MAIL
1263 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1264 Whatever the result of the callout, the value of the domain cache is
1265 left unchanged (for any other kind of callout, getting as far as trying
1266 RCPT means that the domain itself is ok).
1268 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1271 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1272 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1273 session, the daemon ignores ECONNECT errors and logs others; it now
1274 ignores EPIPE as well.
1276 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1277 (quoted-printable decoding).
1279 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1280 later the small subsequent patch to fix an introduced bug.
1282 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1284 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1286 PH/08 An error is now given if message_size_limit is specified negative.
1288 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1289 to be given (somewhat) arbitrary names.
1291 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1294 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1295 miscellaneous code fixes
1297 PH/10 Added the log_reject_target ACL modifier to specify where to log
1300 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1301 hostname. This is wrong, because it relates to the incoming message (and
1302 probably the interface on which it is arriving) and not to the outgoing
1303 callout (which could be using a different interface). This has been
1304 changed to use the value of the helo_data option from the smtp transport
1305 instead - this is what is used when a message is actually being sent. If
1306 there is no remote transport (possible with a router that sets up host
1307 addresses), $smtp_active_hostname is used.
1309 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1310 tweaks were necessary in order to get it to work (see also 21 below):
1311 (a) The code assumed that strncpy() returns a negative number on buffer
1312 overflow, which isn't the case. Replaced with Exim's string_format()
1314 (b) There were several signed/unsigned issues. I just did the minimum
1315 hacking in of casts. There is scope for a larger refactoring.
1316 (c) The code used strcasecmp() which is not a standard C function.
1317 Replaced with Exim's strcmpic() function.
1318 (d) The code set only $1; it now sets $auth1 as well.
1319 (e) A simple test gave the error "authentication client didn't specify
1320 service in request". It would seem that Dovecot has changed its
1321 interface. Fortunately there's a specification; I followed it and
1322 changed what the client sends and it appears to be working now.
1324 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1327 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1328 address A is aliased to B and C, where B exists and C does not. Without
1329 -v the output is "A verified" because verification stops after a
1330 successful redirection if more than one address is generated. However,
1331 with -v the child addresses are also verified. Exim was outputting "A
1332 failed to verify" and then showing the successful verification for C,
1333 with its parentage. It now outputs "B failed to verify", showing B's
1334 parentage before showing the successful verification of C.
1336 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1337 look up a TXT record in a specific list after matching in a combined
1340 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1341 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1342 they consult the DNS. I had assumed they would set it the way they
1343 wanted; and indeed my experiments on Linux seem to show that in some
1344 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1345 To be on the safe side, however, I have now made the interface to
1346 host_find_byname() similar to host_find_bydns(), with an argument
1347 containing the DNS resolver options. The host_find_byname() function now
1348 sets these options at its start, just as host_find_bydns() does. The smtp
1349 transport options dns_qualify_single and dns_search_parents are passed to
1350 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1351 of host_find_byname() use the default settings of RES_DEFNAMES
1352 (qualify_single) but not RES_DNSRCH (search_parents).
1354 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1355 spool_read_header() do less string testing, by means of a preliminary
1356 switch on the second character of optional "-foo" lines. (This is
1357 overdue, caused by the large number of possibilities that now exist.
1358 Originally there were few.) While I was there, I also converted the
1359 str(n)cmp tests so they don't re-test the leading "-" and the first
1360 character, in the hope this might squeeze out yet more improvement.
1362 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1363 flag allowing group syntax was set by the header_syntax check but not
1364 turned off, possible causing trouble later; (2) The flag was not being
1365 set at all for the header_verify test, causing "group"-style headers to
1366 be rejected. I have now set it in this case, and also caused header_
1367 verify to ignore an empty address taken from a group. While doing this, I
1368 came across some other cases where the code for allowing group syntax
1369 while scanning a header line wasn't quite right (mostly, not resetting
1370 the flag correctly in the right place). These bugs could have caused
1371 trouble for malformed header lines. I hope it is now all correct.
1373 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1374 with the "reply" argument non-NULL. The code, however (which originally
1375 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1376 but it didn't always do it. This confused somebody who was copying the
1377 code for some other use. I have removed all the tests.
1379 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1380 feature that was used to support insecure browsers during the U.S. crypto
1381 embargo. It requires special client support, and Exim is probably the
1382 only MTA that supported it -- and would never use it because real RSA is
1383 always available. This code has been removed, because it had the bad
1384 effect of slowing Exim down by computing (never used) parameters for the
1385 RSA_EXPORT functionality.
1387 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1388 authenticator to fail if there's a tab character in the incoming data
1389 (there should never be unless someone is messing about, as it's supposed
1390 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1391 if the connection is using TLS or if the remote IP is the same as the
1392 local IP, and the "valid-client-cert option" if a client certificate has
1395 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1396 authenticators. This can be used for authorization after authentication
1397 succeeds. (In the case of plaintext, it servers for both authentication
1400 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1401 if any retry times were supplied.
1403 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1404 connection, where there is no client IP address to check. In this
1405 situation, the verify now always succeeds.
1407 PH/25 Applied John Jetmore's -Mset patch.
1409 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1411 PH/27 In a string expansion for a processed (not raw) header when multiple
1412 headers of the same name were present, leading whitespace was being
1413 removed from all of them, but trailing whitespace was being removed only
1414 from the last one. Now trailing whitespace is removed from each header
1415 before concatenation. Completely empty headers in a concatenation (as
1416 before) are ignored.
1418 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1419 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1421 PH/29 [Removed. This was a change that I later backed out, and forgot to
1422 correct the ChangeLog entry (that I had efficiently created) before
1423 committing the later change.]
1425 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1426 address errors (4xx response to RCPT) over the same connection as other
1427 messages routed to the same hosts. Such deliveries are always "forced",
1428 so retry times are not inspected. This resulted in far too many retries
1429 for the affected addresses. The effect occurred only when there were more
1430 hosts than the hosts_max_try setting in the smtp transport when it had
1431 the 4xx errors. Those hosts that it had tried were not added to the list
1432 of hosts for which the message was waiting, so if all were tried, there
1433 was no problem. Two fixes have been applied:
1435 (i) If there are any address or message errors in an SMTP delivery, none
1436 of the hosts (tried or untried) are now added to the list of hosts
1437 for which the message is waiting, so the message should not be a
1438 candidate for sending over the same connection that was used for a
1439 successful delivery of some other message. This seems entirely
1440 reasonable: after all the message is NOT "waiting for some host".
1441 This is so "obvious" that I'm not sure why it wasn't done
1442 previously. Hope I haven't missed anything, but it can't do any
1443 harm, as the worst effect is to miss an optimization.
1445 (ii) If, despite (i), such a delivery is accidentally attempted, the
1446 routing retry time is respected, so at least it doesn't keep
1447 hammering the server.
1449 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1450 in ${readsocket because some servers need this prod.
1452 PH/32 Added some extra debug output when updating a wait-xxx database.
1454 PH/33 The hint "could be header name not terminated by colon", which has been
1455 given for certain expansion errors for a long time, was not being given
1456 for the ${if def:h_colon_omitted{... case.
1458 PH/34 The spec says: "With one important exception, whenever a domain list is
1459 being scanned, $domain contains the subject domain." There was at least
1460 one case where this was not true.
1462 PH/35 The error "getsockname() failed: connection reset by peer" was being
1463 written to the panic log as well as the main log, but it isn't really
1464 panic-worthy as it just means the connection died rather early on. I have
1465 removed the panic log writing for the ECONNRESET error when getsockname()
1468 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1469 runs only) independently of the message's sender address. This meant
1470 that, if the 4xx error was in fact related to the sender, a different
1471 message to the same recipient with a different sender could confuse
1472 things. In particualar, this can happen when sending to a greylisting
1473 server, but other circumstances could also provoke similar problems.
1474 I have changed the default so that the retry time for these errors is now
1475 based a combination of the sender and recipient addresses. This change
1476 can be overridden by setting address_retry_include_sender=false in the
1479 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1480 remote server are returned as part of bounce messages. This was not
1481 happening for LMTP over a pipe (the lmtp transport), but now it is the
1482 same for both kinds of LMTP.
1484 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1485 in header lines that were in fact CNAMEs. This is no longer the case.
1487 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1488 and queue runs started by the daemon processed all messages. This has
1489 been fixed so that -R and -S can now usefully be given with -q<time>.
1491 PH/40 Import PCRE release 6.7 (fixes some bugs).
1493 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1495 PH/42 Give an error if -q is specified more than once.
1497 PH/43 Renamed the variables $interface_address and $interface_port as
1498 $received_ip_address and $received_port, to make it clear that these
1499 values apply to message reception, and not to the outgoing interface when
1500 a message is delivered. (The old names remain recognized, of course.)
1502 PH/44 There was no timeout on the connect() call when using a Unix domain
1503 socket in the ${readsocket expansion. There now is.
1505 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1506 be meaningful with "accept".
1508 SC/01 Eximstats V1.43
1509 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1511 SC/02 Eximstats V1.44
1512 Use a glob alias rather than an array ref in the generated
1513 parser. This improves both readability and performance.
1515 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1516 Collect SpamAssassin and rejection statistics.
1517 Don't display local sender or destination tables unless
1518 there is data to show.
1519 Added average volumes into the top table text output.
1521 SC/04 Eximstats V1.46
1522 Collect data on the number of addresses (recipients)
1523 as well as the number of messages.
1525 SC/05 Eximstats V1.47
1526 Added 'Message too big' to the list of mail rejection
1527 reasons (thanks to Marco Gaiarin).
1529 SC/06 Eximstats V1.48
1530 Mainlog lines which have GMT offsets and are too short to
1531 have a flag are now skipped.
1533 SC/07 Eximstats V1.49 (Alain Williams)
1534 Added the -emptyok flag.
1536 SC/08 Eximstats V1.50
1537 Fixes for obtaining the IP address from reject messages.
1539 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1540 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1541 whitesspace changes from 4.64-PH/27
1543 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1546 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1547 are found, allow negative numbers in numeric criteria)
1549 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1551 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1552 to match changes made in 4.64-PH/43
1554 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1556 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1557 "domain" is actually an IP address, and force a failure. This locks out
1558 those revolvers/nameservers that support "A-for-A" lookups, in
1559 contravention of the specifications.
1561 PH/48 When a host name was looked up from an IP address, and the subsequent
1562 forward lookup of the name timed out, the host name was left in
1563 $sender_host_name, contrary to the specification.
1565 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1566 restricted to single-key lookups, Exim was not diagnosing an error if
1567 * or *@ was used with a query-style lookup.
1569 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1571 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1572 long ago, but noone interested enough thought of it. Let's just say that
1573 the "1.1" means that there are some new functions that weren't there at
1574 some point in the past.
1576 PH/51 Error processing for expansion failure of helo_data from an smtp
1577 transport during callout processing was broken.
1579 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1580 tested/used via the -bh/-bhc/-bs options.
1582 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1583 bug, fixed in subsequent PCRE releases).
1585 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1586 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1588 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1594 SC/01 Use a glob alias rather than an array ref in eximstats generated
1595 parser. This improves both readability and performance.
1597 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1598 Don't display local sender or destination tables in eximstats unless
1599 there is data to show.
1600 Added average volumes into the eximstats top table text output.
1602 SC/03 Collect data on the number of addresses (recipients) as well
1603 as the number of messages in eximstats.
1605 TF/01 Correct an error in the documentation for the redirect router. Exim
1606 does (usually) call initgroups() when daemonizing.
1608 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1609 with consistent privilege compared to when running as a daemon.
1611 TF/03 Note in the spec that $authenticated_id is not set for local
1612 submissions from trusted users.
1614 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1615 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1617 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1618 by adding some example configuration directives to the default
1619 configuration file. A little bit of work is required to uncomment the
1620 directives and define how usernames and passwords are checked, but
1621 there is now a framework to start from.
1623 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1624 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1625 without this. I don't know how relevant this is to other LDAP libraries.
1627 PH/02 Add the verb name to the "unknown ACL verb" error.
1629 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1631 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1633 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1634 directory not expanded when it should be if an expanded home directory
1635 was set for the address (which is overridden by the transport).
1637 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1640 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1641 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1642 because it is too late at that time, and has no effect.
1644 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1645 security issue with \' (bugzilla #107). I could not use the
1646 PQescapeStringConn() function, because it needs a PGconn value as one of
1649 PH/08 When testing addresses using -bt, indicate those final addresses that
1650 are duplicates that would not cause an additional delivery. At least one
1651 person was confused, thinking that -bt output corresponded to deliveries.
1652 (Suppressing duplicates isn't a good idea as you lose the information
1653 about possibly different redirections that led to the duplicates.)
1655 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1656 systems where poll() doesn't work, in particular OS X.
1658 PH/10 Added more information to debugging output for retry time not reached.
1660 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1661 operations in malware.c.
1663 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1666 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1667 syslog_duplication set false, log lines that would normally be written
1668 both the the main log and to the reject log were not written to syslog at
1671 PH/14 In the default configuration, change the use of "message" in ACL warn
1672 statements to "add_header".
1674 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1675 not followed by a command (e.g. "seen endif").
1677 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1678 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1681 PH/17 Added extra conditions to the default value of delay_warning_condition
1685 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1686 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1687 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1690 The Auto-Submitted: and various List- headers are standardised, whereas I
1691 don't think Precedence: ever was.
1693 PH/18 Refactored debugging code in route_finduser() to show more information,
1694 in particular, the error code if getpwnam() issues one.
1696 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1697 This is apparently needed in addition to the PH/07 change above to avoid
1698 any possible encoding problems.
1700 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1701 but not after initializing Perl.
1703 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1704 output them only if debugging. By default they are written stderr,
1705 apparently, which is not desirable.
1707 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1710 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1713 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1715 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1716 authenticated or an ident call has been made. Suppress the default
1717 values for $authenticated_id and $authenticated_sender (but permit -oMai
1718 and -oMas) when testing with -bh.
1720 PH/24 Re-jigged the order of the tests in the default configuration so that the
1721 tests for valid domains and recipients precede the DNS black list and CSA
1722 tests, on the grounds that those ones are more expensive.
1724 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1725 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1726 This bug exists in every version of Exim that I still have, right back to
1729 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1730 However, an attempt to turn on case-sensitivity in a regex key by
1731 including (?-i) didn't work because the subject string was already
1732 lowercased, and the effects were non-intuitive. It turns out that a
1733 one-line patch can be used to allow (?-i) to work as expected.
1739 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1740 other effects) broke the use of negated acl sub-conditions.
1742 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1745 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1746 "Deny" causes Exim to reject the incoming connection with a 554 error.
1747 Unfortunately, if there is a major crisis, such as a disk failure,
1748 tcp-wrappers gives "deny", whereas what one would like would be some
1749 kind of temporary error. A kludge has been added to help with this.
1750 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1751 554 error is used if errno is still zero or contains ENOENT (which occurs
1752 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1755 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1757 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1758 errors. Otherwise a message that provokes a temporary error (when other
1759 messages do not) can cause a whole host to time out.
1761 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1762 addresses were routed directly to files or pipes from a redirect router.
1763 File deliveries just didn't batch; pipe deliveries might have suffered
1766 PH/06 A failure to get a lock for a hints database would erroneously always say
1767 "Failed to get write lock", even when it was really a read lock.
1769 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1770 of 0600. This has been changed to use the value of the lockfile_mode
1771 option (which defaults to 0600).
1773 PH/08 Applied small patch from the Sieve maintainer.
1775 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1776 folder from quota calculations, a direct delivery into this folder messed
1777 up the contents of the maildirsize file. This was because the regex was
1778 used only to exclude .Trash (or whatever) when the size of the mailbox
1779 was calculated. There was no check that a delivery was happening into an
1780 excluded directory. This bug has been fixed by ignoring all quota
1781 processing for deliveries into excluded directories.
1783 PH/10 Added the maildirfolder_create_regex option to appendfile.
1789 PH/01 The code for finding all the local interface addresses on a FreeBSD
1790 system running IPv6 was broken. This may well have applied to all BSD
1791 systems, as well as to others that have similar system calls. The broken
1792 code found IPv4 interfaces correctly, but gave incorrect values for the
1793 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1794 that it would not match correctly against @[] and not recognize the IPv6
1797 PH/02 The ipliteral router was not recognizing addresses of the form user@
1798 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1800 PH/03 Added disable_ipv6.
1802 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1803 decoded form, because it is most often used to construct To: headers
1804 lines in autoreplies, and the decoded form may well be syntactically
1805 invalid. However, $reply_address has leading white space removed, and all
1806 newlines turned into spaces so that the autoreply transport does not
1809 PH/05 If group was specified without a user on a router, and no group or user
1810 was specified on a transport, the group from the router was ignored.
1812 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1813 for visible compile-time settings that can be used to change these
1814 numbers, for those that want even more. Backwards compatibility with old
1815 spool files has been maintained. However, going back to a previous Exim
1816 release will lost any variables that are in spool files.
1818 PH/07 Two small changes when running in the test harness: increase delay when
1819 passing a TCP/IP connection to a new process, in case the original
1820 process has to generate a bounce, and remove special handling of
1821 127.0.0.2 (sic), which is no longer necessary.
1823 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1824 be the same on different OS.
1826 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1829 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1830 whether --show-vars was specified or not
1832 JJ/02 exipick: Added support for new ACL variable spool format introduced
1835 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1836 syntactically invalid From: or Reply-to: line, and a filter used this to
1837 generate an autoreply, and therefore failed to obtain an address for the
1838 autoreply, Exim could try to deliver to a non-existent relative file
1839 name, causing unrelated and misleading errors. What now happens is that
1840 it logs this as a hard delivery error, but does not attempt to create a
1843 PH/11 The exinext utility has a -C option for testing purposes, but although
1844 the given file was scanned by exinext itself; it wasn't being passed on
1845 when Exim was called.
1847 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1848 an end-of-file indication when reading a command response.
1850 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1851 compiled. In many other places in Exim, IPv6 addresses are always
1852 recognized, so I have changed this. It also means that IPv4 domain
1853 literals of the form [IPV4:n.n.n.n] are now always recognized.
1855 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1856 used if the router is not running as root, for example, when verifying at
1857 ACL time, or when using -bh. The debugging output from this situation was
1858 non-existent - all you got was a failure to exec. I have made two
1861 (a) Failures to set uid/gid, the current directory, or a process leader
1862 in a subprocess such as that created by queryprogram now generate
1863 suitable debugging ouput when -d is set.
1865 (b) The queryprogram router detects when it is not running as root,
1866 outputs suitable debugging information if -d is set, and then runs
1867 the subprocess without attempting to change uid/gid.
1869 PH/15 Minor change to Makefile for building test_host (undocumented testing
1872 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1873 additional section of a DNS packet that returns MX or SRV records.
1874 Instead, it always explicitly searches for A/AAAA records. This avoids
1875 major problems that occur when a DNS server includes only records of one
1876 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1877 fixed another bug: if SRV records were looked up and the corresponding
1878 address records were *not* found in the additional section, the port
1879 values from the SRV records were lost.
1881 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1882 using the correct key (the original address) when searching the retry
1883 rules in order to find which one to use for generating the retry hint.
1885 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1886 adding the default one. Similarly, if it contains a Reply-To: header, the
1887 errors_reply_to option, if set, is not used.
1889 PH/19 When calculating a retry time, Exim used to measure the "time since
1890 failure" by looking at the "first failed" field in the retry record. Now
1891 it does not use this if it is later than than the arrival time of the
1892 message. Instead it uses the arrival time. This makes for better
1893 behaviour in cases where some deliveries succeed, thus re-setting the
1894 "first failed" field. An example is a quota failure for a huge message
1895 when small messages continue to be delivered. Without this change, the
1896 "time since failure" will always be short, possible causing more frequent
1897 delivery attempts for the huge message than are intended.
1898 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1900 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1901 $1, $2, $3) because the numerical variables can be reset during some
1902 expansion items (e.g. "match"), thereby losing the authentication data.
1904 PH/21 Make -bV show the size of off_t variables so that the test suite can
1905 decide whether to run tests for quotas > 2G.
1907 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1908 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1909 filecount value is greater than 2G or if a quota value is greater than 2G
1910 on a system where the size of off_t is not greater than 4, a panic error
1913 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1914 never match. The debug and -bh output now contains an explicit error
1915 message indicating a malformed IPv4 address or mask.
1917 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1918 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1919 PH/23 above applies.
1921 PH/25 Do not write to syslog when running in the test harness. The only
1922 occasion when this arises is a failure to open the main or panic logs
1923 (for which there is an explicit test).
1925 PH/26 Added the /no_tell option to "control=freeze".
1927 PH/27 If a host name lookup failed very early in a connection, for example, if
1928 the IP address matched host_lookup and the reverse lookup yielded a name
1929 that did not have a forward lookup, an error message of the form "no IP
1930 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1931 could be logged. Now it outputs the IP address instead of "NULL".
1933 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1934 allows the sender and the authenticated sender to be set when
1935 submitting a message from within Exim. Since child_open_exim() is
1936 documented for local_scan(), the new function should be too.
1938 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1939 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1940 results in an empty string is now treated as unset.
1942 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1944 PH/31 Added sender_verify_fail logging option.
1946 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1947 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1948 tidied the source and removed it altogether.
1950 PH/33 When a queue run was abandoned because the load average was too high, a
1951 log line was always written; now it is written only if the queue_run log
1952 selector is set. In addition, the log line for abandonment now contains
1953 information about the queue run such as the pid. This is always present
1954 in "start" and "stop" lines but was omitted from the "abandon" line.
1956 PH/34 Omit spaces between a header name and the colon in the error message that
1957 is given when verify = headers_syntax fails (if there are lots of them,
1958 the message gets confusing).
1960 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1961 names, as there are now some PTR records that contain slashes. This check
1962 is only to protect against broken name servers that fall over on strange
1963 characters, so the fact that it applies to all lookups doesn't matter.
1965 PH/36 Now that the new test suite is complete, we can remove some of the
1966 special code in Exim that was needed for the old test suite. For example,
1967 sorting DNS records because real resolvers return them in an arbitrary
1968 order. The new test suite's fake resolver always returns records in the
1971 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1972 bounces) except when queue_only is set, to avoid logging races between
1973 the different processes.
1975 PH/38 Panic-die if .include specifies a non-absolute path.
1977 PH/39 A tweak to the "H" retry rule from its user.
1979 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1980 a label. They prevented compilation on older perls.
1982 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1983 a warning to be raised on newish perls.
1985 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1986 on queue. Changes to match documented behaviour of showing count of
1987 messages matching specified criteria.
1989 PH/40 Changed the default ident timeout from 30s to 5s.
1991 PH/41 Added support for the use of login_cap features, on those BSD systems
1992 that have them, for controlling the resources used by pipe deliveries.
1994 PH/42 The content-scanning code uses fopen() to create files in which to put
1995 message data. Previously it was not paying any attention to the mode of
1996 the files. Exim runs with umask(0) because the rest of the code creates
1997 files with open(), and sets the required mode explicitly. Thus, these
1998 files were ending up world-writeable. This was not a big issue, because,
1999 being within the spool directory, they were not world-accessible. I have
2000 created a function called modefopen, which takes an additional mode
2001 argument. It sets umask(777), creates the file, chmods it to the required
2002 mode, then resets the umask. All the relevant calls to fopen() in the
2003 content scanning code have been changed to use this function.
2005 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2006 to 24 hours. This avoids potential overflow problems when processing G
2007 and H retry rules. I suspect nobody ever tinkers with this value.
2009 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2011 PH/45 When the plaintext authenticator is running as a client, the server's
2012 challenges are checked to ensure they are valid base64 strings. By
2013 default, the authentication attempt is cancelled if an invalid string is
2014 received. Setting client_ignore_invalid_base64 true ignores these errors.
2015 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2016 they are received. Thus, the responses can be made to depend on the
2017 challenges. If an invalid string is ignored, an empty string is placed in
2020 PH/46 Messages that are created by the autoreply transport now contains a
2021 References: header, in accordance with RFCs 2822 and 3834.
2023 PH/47 Added authenticated_sender_force to the smtp transport.
2025 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2027 PH/49 Installed latest patch from the Sieve maintainer.
2029 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2030 was also set, the appendfile transport was unnecessarily scanning a
2031 directory of message files (e.g. for maildir delivery) to find the count
2032 of files (along with the size), even though it did not need this
2033 information. It now does the scan only if it needs to find either the
2034 size of the count of files.
2036 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2038 PH/52 Two bugs concerned with error handling when the smtp transport is
2041 (i) Exim was not creating retry information for temporary errors given
2042 for individual recipients after the DATA command when the smtp transport
2043 was used in LMTP mode. This meant that they could be retried too
2044 frequently, and not timed out correctly.
2046 (ii) Exim was setting the flag that allows error details to be returned
2047 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2048 recipients that were returned after the DATA command.
2050 PH/53 This is related to PH/52, but is more general: for any failing address,
2051 when detailed error information was permitted to be returned to the
2052 sender, but the error was temporary, then after the final timeout, only
2053 "retry timeout exceeded" was returned. Now it returns the full error as
2054 well as "retry timeout exceeded".
2056 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2057 do this, and (what is worse) MTAs that accept it.
2059 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2060 will now be deprecated.
2062 PH/56 New os.c-cygwin from the Cygwin maintainer.
2064 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2065 formats (previously only available in exim formats via -bpr, -bpru,
2066 and -bpra. Now also available in native and exiqgrep formats)
2068 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2069 with very large, slow to parse queues
2071 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2073 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2075 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2076 responses to authentication challenges, though it was showing the
2077 challenges; (ii) I've removed the CR characters from the debug output for
2080 PH/58 Allow for the insertion of a newline as well as a space when a string
2081 is turned into more than one encoded-word during RFC 2047 encoding. The
2082 Sieve code now uses this.
2084 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2085 data_4xx, lost_connection, tls_required.
2087 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2088 message was being sent as an SMTP response.
2090 PH/61 Add -l and -k options to exicyclog.
2092 PH/62 When verifying, if an address was redirected to one new address, so that
2093 verification continued, and the new address failed or deferred after
2094 having set something in $address_data, the value of $address_data was not
2095 passed back to the ACL. This was different to the case when no
2096 redirection occurred. The value is now passed back in both cases.
2098 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2099 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2100 use login_cap.h, so on its own it isn't the distinguishing feature. The
2101 new name refers directly to the setclassresources() function.
2103 PH/65 Added configuration files for NetBSD3.
2105 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2107 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2108 is preferred over IPv4.
2110 PH/68 The bounce_return_message and bounce_return_body options were not being
2111 honoured for bounces generated during the reception of non-SMTP messages.
2112 In particular, this applied to messages rejected by the ACL. This bug has
2113 been fixed. However, if bounce_return_message is true and bounce_return_
2114 body is false, the headers that are returned for a non-SMTP message
2115 include only those that have been read before the error was detected.
2116 (In the case of an ACL rejection, they have all been read.)
2118 PH/69 The HTML version of the specification is now built in a directory called
2119 spec_html instead of spec.html, because the latter looks like a path with
2120 a MIME-type, and this confuses some software.
2122 PH/70 Catch two compiler warnings in sieve.c.
2124 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2125 function verify_get_ident() calls ip_connect() to connect a socket, but
2126 if the "connect()" function timed out, ip_connect() used to close the
2127 socket. However, verify_get_ident() also closes the socket later, and in
2128 between Exim writes to the log, which may get opened at this point. When
2129 the socket was closed in ip_connect(), the log could get the same file
2130 descriptor number as the socket. This naturally causes chaos. The fix is
2131 not to close the socket in ip_connect(); the socket should be closed by
2132 the function that creates it. There was only one place in the code where
2133 this was missing, in the iplookup router, which I don't think anybody now
2134 uses, but I've fixed it anyway.
2136 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2137 well as to direct DNS lookups. Otherwise the handling of names in host
2138 lists is inconsistent and therefore confusing.
2144 PH/01 Two changes to the default runtime configuration:
2146 (1) Move the checks for relay_from_hosts and authenticated clients from
2147 after to before the (commented out) DNS black list checks.
2149 (2) Add control=submission to the relay_from_hosts and authenticated
2150 clients checks, on the grounds that messages accepted by these
2151 statements are most likely to be submissions.
2153 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2155 (1) Generate an error if the third argument for the ${prvs expansion is
2158 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2161 (3) Reset the variables that are obtained from the first argument of
2162 ${prvscheck and used in the second argument before leaving the code,
2163 because their memory is reclaimed, so using them afterwards may do
2166 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2167 one (it's much easier than Tom thought :-).
2169 (5) Because of (4), we can now allow for the use of $prvscheck_result
2170 inside the third argument.
2172 PH/03 For some reason, the default setting of PATH when running a command from
2173 a pipe transport was just "/usr/bin". I have changed it to
2176 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2177 anything to be listed in the output from -bV.
2179 PH/05 When a filter generated an autoreply, the entire To: header line was
2180 quoted in the delivery log line, like this:
2182 => >A.N.Other <ano@some.domain> <original@ddress> ...
2184 This has been changed so that it extracts the operative address. There
2185 may be more than one such address. If so, they are comma-separated, like
2188 => >ano@some.domain,ona@other.domain <original@ddress> ...
2190 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2191 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2192 not being looked up in the rDNS to get a host name, Exim was showing the
2193 IP address twice in Received: lines, even though the IP addresses were
2194 identical. For example:
2196 Received: from [1.2.3.4] (helo=[1.2.3.4])
2198 However, if the real host name was known, it was omitting the HELO data
2199 if it matched the actual IP address. This has been tidied up so that it
2200 doesn't show the same IP address twice.
2202 PH/07 When both +timestamp and +memory debugging was on, the value given by
2203 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2204 called by the debug printing, thereby overwriting the timestamp buffer.
2205 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2207 PH/08 When the original message was included in an autoreply transport, it
2208 always said "this is a copy of the message, including all the headers",
2209 even if body_only or headers_only was set. It now gives an appropriate
2212 PH/09 Applied a patch from the Sieve maintainer which:
2214 o fixes some comments
2215 o adds the (disabled) notify extension core
2216 o adds some debug output for the result of if/elsif tests
2217 o points to the current vacation draft in the documentation
2218 and documents the missing references header update
2222 o fixes a bug in processing the envelope test (when testing
2223 multiple envelope elements, the last element determinted the
2226 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2227 Electronic Mail") by including:
2229 Auto-submitted: auto-generated
2231 in the messages that it generates (bounce messages and others, such as
2232 warnings). In the case of bounce messages for non-SMTP mesages, there was
2233 also a typo: it was using "Auto_submitted" (underscore instead of
2234 hyphen). Since every message generated by Exim is necessarily in response
2235 to another message, thes have all been changed to:
2237 Auto-Submitted: auto-replied
2239 in accordance with these statements in the RFC:
2241 The auto-replied keyword:
2243 - SHOULD be used on messages sent in direct response to another
2244 message by an automatic process,
2246 - MUST NOT be used on manually-generated messages,
2248 - MAY be used on Delivery Status Notifications (DSNs) and Message
2249 Disposition Notifications (MDNs),
2251 - MUST NOT be used on messages generated by automatic or periodic
2252 processes, except for messages which are automatic responses to
2255 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2256 to the default Received: header definition.
2258 PH/12 Added log selector acl_warn_skipped (default on).
2260 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2261 variables because (a) they are in the wrong storage pool and (b) even if
2262 they were copied, it wouldn't work properly because of the caching.
2264 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2265 checking when decoding. Apparently there are clients that generate
2266 overlong encoded strings. Why am I not surprised?
2268 PH/15 If the first argument of "${if match_address" was not empty, but did not
2269 contain an "@" character, Exim crashed. Now it writes a panic log message
2270 and treats the condition as false.
2272 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2274 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2275 extension "envelope-auth". The code is finished and in agreement with
2276 other implementations, but there is no documentation so far and in fact,
2277 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2278 not changing the active code.
2280 Print executed "if" and "elsif" statements when debugging is used. This
2281 helps a great deal to understand what a filter does.
2283 Document more things not specified clearly in RFC3028. I had all this
2284 sorted out, when out of a sudden new issues came to my mind. Oops."
2286 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2289 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2290 canonical form (as documented). However, after a host name lookup from
2291 the IP address, check_host() was doing a simple string comparison with
2292 addresses acquired from the DNS when checking that the found name did
2293 have the original IP as one of its addresses. Since any found IPv6
2294 addresses are likely to be in abbreviated form, the comparison could
2295 fail. Luckily, there already exists a function for doing the comparison
2296 by converting both addresses to binary, so now that is used instead of
2297 the text comparison.
2299 PH/20 There was another similar case to PH/19, when a complete host name was
2300 given in a host list; looking up its IP address could give an abbreviated
2301 form, whereas the current host's name might or might not be abbreviated.
2302 The same fix has been applied.
2308 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2309 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2312 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2313 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2315 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2317 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2318 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2319 Also, for "Auto-Submitted", treat anything other than "no" as
2320 non-personal, in accordance with RFC 3834. (Previously it treated
2321 anything starting "auto-" as non-personal.)
2323 TF/01 The control=submission/name=... option had a problem with syntax
2324 errors if the name included a slash character. The /name= option
2325 now slurps the rest of the string, so it can include any characters
2326 but it must come last in the list of options (after /sender_retain
2329 PH/05 Some modifications to the interface to the fake nameserver for the new
2337 TK/01 Added the "success_on_redirect" address verification option. See
2338 NewStuff for rationale and an example.
2340 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2342 PH/02 Patch to exigrep to allow it to work on syslog lines.
2344 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2345 fread() to skip over the body file's header line, because in Cygwin the
2346 header line is locked and is inaccessible.
2348 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2349 co-exist for some time) to make it clear that it is the Exim ID that is
2350 referenced, not the Message-ID: header line.
2352 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2353 string_format() function, because snprintf() does not exist on all
2356 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2357 ${stat: expansion item.
2359 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2360 protocol synchronization error", to keep the pedants happy.
2362 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2363 well as for IRIX systems, when gcc is being used. See the host.c source
2366 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2368 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2371 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2372 transport and to the smtp transport in LMTP mode.
2374 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2376 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2377 run a filter in a subprocess. This could lead to confusion in subsequent
2378 lookups in the parent process. There should also be a search_tidyup() at
2379 the end of the subprocess.
2381 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2382 only if the host matched helo_try_verify_hosts, which caused the
2383 verification to occur when the EHLO/HELO command was issued. The ACL just
2384 tested the remembered result. Now, if a previous verification attempt has
2385 not happened, "verify = helo" does it there and then.
2387 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2389 TK/03 Fix log output including CR from clamd.
2391 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2392 exist provoked a memory error which could cause a segfault.
2394 PH/15 Installed PCRE 6.2
2396 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2398 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2399 of the problem. Specifically, suggested +O2 rather than +O1 for the
2402 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2404 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2405 record, the port was not correctly logged when the outgoing_port log
2406 selector was set (it logged the transort's default port).
2408 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2409 fallback_hosts, and "hosts" in the smtp transport.
2411 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2412 host errors such as "Connection refused".
2414 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2415 authentication with radiusclient 0.4.9:
2417 - Error returned from rc_read_config was caught wrongly
2418 - Username/password not passed on to radius server due to wrong length.
2420 The presumption is that some radiusclient API changes for 4.51/PH/17
2421 were not taken care of correctly. The code is still untested by me (my
2422 Linux distribution still has 0.3.2 of radiusclient), but it was
2423 contributed by a Radius user.
2425 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2426 expanding the "port" option of the smtp transport.
2428 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2429 while reading a MIME header. Thanks to Tom Hughes for a patch.
2431 PH/24 Include config.h inside local_scan.h so that configuration settings are
2434 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2435 that in an ACL for RCPT (for example), you can examine exactly what was
2438 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2439 commands, but it was not correctly comparing the address with the actual
2440 client host address. Thus, it would show the EHLO address in Received:
2441 header lines when this was not necessary.
2443 PH/27 Added the % operator to ${eval:}.
2445 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2446 it should be ignoring failures (because with -C, for example, it has lost
2447 privilege). It wasn't ignoring creation failures other than "already
2450 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2453 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2454 because an input line was too long, either on its own, or by virtue of
2455 too many continuations, the temporary file was not being removed, and the
2456 return code was incorrect.
2458 PH/31 Missing "BOOL" in function definition in filtertest.c.
2460 PH/32 Applied Sieve patches from the maintainer.
2462 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2464 PH/33 Added "verify = not_blind".
2466 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2467 Local/Makefile (with some defaults set). These are used in built scripts
2468 such as exicyclog, but they have never been used in the exim_install
2469 script (though there are many overriding facilities there). I have
2470 arranged that the exim_install script now takes note of these two
2473 PH/35 Installed configuration files for Dragonfly.
2475 PH/36 When a locally submitted message by a trusted user did not contain a
2476 From: header, and the sender address was obtained from -f or from an SMTP
2477 MAIL command, and the trusted user did not use -F to supply a sender
2478 name, $originator_name was incorrectly used when constructing a From:
2479 header. Furthermore, $originator_name was used for submission mode
2480 messages from external hosts without From: headers in a similar way,
2481 which is clearly wrong.
2483 PH/37 Added control=suppress_local_fixups.
2485 PH/38 When log_selector = +received_sender was set, and the addition of the
2486 sender made the log line's construction buffer exactly full, or one byte
2487 less than full, an overflow happened when the terminating "\n" was
2490 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2491 when the result of a list match is failure because a DNS lookup failed.
2493 PH/40 RM_COMMAND is now used in the building process.
2495 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2496 the "build-* directories that it finds.
2498 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2499 address in a domain literal was a prefix of an interface address.
2501 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2502 when verifying a sender address, unless rewrite_headers is false.
2504 PH/44 Wrote a long comment about why errors_to addresses are verified as
2505 recipients, not senders.
2507 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2508 the ratelimit ACL was added.
2510 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2512 PH/46 Added extra information about PostgreSQL errors to the error string.
2514 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2515 avoiding the need to install special zones in a real server. This is
2516 backwards compatible; if it can't find the fake resolver, it drops back.
2517 Thus, both old and new test suites can be run.
2519 TF/02 Added util/ratelimit.pl
2521 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2522 clock is set back in time.
2524 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2525 Candler <B.Candler@pobox.com>.
2527 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2528 OK for addresses that are the result of redirections.
2530 PH/48 A number of further additions for the benefit of the new test suite,
2531 including a fake gethostbyname() that interfaces to the fake DNS resolver
2534 TF/06 The fix for widen_domains has also been applied to qualify_single and
2535 search_parents which are the other dnslookup options that can cause
2538 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2541 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2543 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2544 service" if it encountered an MX record with an empty target hostname.
2545 The message is now "an MX or SRV record indicated no SMTP service".
2547 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2548 if the DNS of the sending site is misconfigured. This is quite a
2549 common situation. This change restores the behaviour of treating a
2550 helo verification defer as a failure.
2552 PH/51 If self=fail was set on a router, the bounce message did not include the
2553 actual error message.
2559 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2561 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2562 command itself ended in error, the underlying message about the transport
2563 filter timeout was being overwritten with the pipe command error. Now the
2564 underlying error message should be appended to the second error message.
2566 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2568 PH/02 Reduce the amount of output that "make" produces by default. Full output
2569 can still be requested.
2571 PH/03 The warning log line about a condition test deferring for a "warn" verb
2572 was being output only once per connection, rather than after each
2573 occurrence (because it was using the same function as for successful
2574 "warn" verbs). This seems wrong, so I have changed it.
2576 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2577 it should not have, which might have caused a crash in the right
2578 circumstances, but probably never did.
2580 PH/04 Installed a modified version of Tony Finch's patch to make submission
2581 mode fix the return path as well as the Sender: header line, and to
2582 add a /name= option so that you can make the user's friendly name appear
2585 TF/03 Added the control = fakedefer ACL modifier.
2587 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2588 Mark Lowes for thorough testing.
2590 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2592 TK/03 Merged latest SRS patch from Miles Wilton.
2594 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2595 with the definition in sysexits.h (which is #included earlier).
2596 Fortunately, Exim does not actually use EX_OK. The code used to try to
2597 preserve the sysexits.h value, by assumimg that macro definitions were
2598 scanned for macro replacements. I have been disabused of this notion,
2599 so now the code just undefines EX_OK before #including unistd.h.
2601 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2602 in the smtp transport. When a block could not be written in a single
2603 write() function, the timeout was being re-applied to each part-write.
2604 This seems wrong - if the receiver was accepting one byte at a time it
2605 would take for ever. The timeout is now adjusted when this happens. It
2606 doesn't have to be particularly precise.
2608 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2609 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2611 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2612 without a domain if the check to <postmaster@domain> fails.
2614 SC/01 Eximstats: added -xls and the ability to specify output files
2615 (patch written by Frank Heydlauf).
2617 SC/02 Eximstats: use FileHandles for outputing results.
2619 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2621 SC/04 Eximstats: fixed display of large numbers with -nvr option
2623 SC/05 Eximstats: fixed merging of reports with empty tables.
2625 SC/06 Eximstats: added the -include_original_destination flag
2627 SC/07 Eximstats: removed tabs and trailing whitespace.
2629 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2631 TK/06 MBOX spool code: Add real "From " MBOX separator line
2632 so the .eml file is really in mbox format (even though
2633 most programs do not really care). Patch from Alex Miller.
2635 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2636 The latter is generated from $received_to and is only set if the
2637 message has one envelope recipient. SA can use these headers,
2638 obviously out-of-the-box. Patch from Alex Miller.
2640 PH/08 The ${def test on a variable was returning false if the variable's
2641 value was "0", contrary to what the specification has always said!
2642 The result should be true unless the variable is empty.
2644 PH/09 The syntax error of a character other than { following "${if
2645 def:variable_name" (after optional whitespace) was not being diagnosed.
2646 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2647 accidental colon was present, for example, could give incorrect results.
2649 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2650 result is used (not including appendfile, where other changes are about
2653 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2654 This involved changing a lot of size variables from int to off_t. It
2655 should work with maildirs and everything.
2657 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2658 spamd dying while we are connected to it.
2660 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2663 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2664 function for BDB 4.3.
2666 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2668 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2669 the sources for PCRE have been re-arranged and more files are now
2672 PH/15 The code I had for printing potentially long long variables in PH/11
2673 above was not the best (it lost precision). The length of off_t variables
2674 is now inspected at build time, and an appropriate printing format (%ld
2675 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2676 to be "long long int" or "long int". This is needed for the internal
2677 formatting function string_vformat().
2679 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2680 the configuration file to be ":syslog", then the script "guesses" where
2681 the logs files are, rather than using the compiled in default. In our
2682 case the guess is not the same as the compiled default, so the script
2683 suddenly stopped working when I started to use syslog. The patch checks
2684 to see if log_file_path is "". If so, it attempts to read it from exim
2685 with no configuration file to get the compiled in version, before it
2686 falls back to the previous guessing code."
2688 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2689 implementing BATV in an Exim configuration. See NewStuff for the gory
2692 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2693 Makefile that are specific to HP-UX.
2695 PH/18 If the "use_postmaster" option was set for a recipient callout together
2696 with the "random" option, the postmaster address was used as the MAIL
2697 FROM address for the random test, but not for the subsequent recipient
2698 test. It is now used for both.
2700 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2701 patch removes a few documentation additions to RFC 3028, because the
2702 latest draft now contains them. It adds the new en;ascii-case comparator
2703 and a new error check for 8bit text in MIME parts. Comparator and
2704 require names are now matched exactly. I enabled the subaddress
2705 extension, but it is not well tested yet (read: it works for me)."
2707 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2708 rework some of the code of TK/09 above to avoid the hardwired use of
2709 "%lld" and "long long". Replaced the call to snprintf() with a call to
2712 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2713 records point to non-existent hosts", "retry timeout exceeded", and
2714 "retry time not reached for any host after a long failure period".
2716 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2717 experimental DomainKeys support:
2719 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2720 (2) On an error such as an illegally used "control", the wrong name for
2721 the control was given.
2723 These problems did NOT occur unless DomainKeys support was compiled.
2725 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2727 PH/24 Added ${if match_ip condition.
2729 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2730 databases so that it will be absolutely obvious if a crash occurs in the
2731 DB library. This is a regular occurrence (often caused by mis-matched
2734 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2735 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2736 picked up on a user's system that detects such things. There doesn't seem
2737 to be a gcc warning option for this - only an attribute that has to be
2738 put on the function's prototype. It seems that in Fedora Core 4 they have
2739 set this on a number of new functions. No doubt there will be more in due
2742 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2743 specify a transport. However, if an address that was verified by such a
2744 router was the subject of a callout, Exim crashed because it tried to
2745 read the rcpt_include_affixes from the non-existent transport. Now it
2746 just assumes that the setting of that option is false. This bug was
2747 introduced by 4.51/PH/31.
2749 PH/28 Changed -d+all to exclude +memory, because that information is very
2750 rarely of interest, but it makes the output a lot bigger. People tend to
2751 do -d+all out of habit.
2753 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2754 code in os-type was giving problems when libc.so lives in lib64, like on
2757 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2758 aren't the modern standard, and it seems that some systems' include files
2759 don't always have them. Exim was already checking for some of the newer
2760 ones like T_AAAA, and defining it itself. I've added checks for all the
2761 record types that Exim uses.
2763 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2764 not automatically generating a new one, as it is supposed to. This
2765 prevented TLS from working. If the file did exist, but contained invalid
2766 data, a new version was generated, as expected. It was only the case of a
2767 non-existent file that was broken.
2769 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2770 with a change in libdomainkeys > 0.64.
2772 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2773 from DNS. If the selector record carries the flag, it now has
2774 precedence over the domain-wide flag.
2776 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2778 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2779 the use of a port name that isn't defined in /etc/services) occurred, the
2780 message was deferred as in a normal delivery, and thus remained on the
2781 spool, instead of being failed because of the mua_wrapper setting. This
2782 is now fixed, and I tidied up some of the mua_wrapper messages at the
2785 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2786 the messages in a hash of arrays rather than using individual hashes.
2787 This is a bit cleaner and results in dramatic memory savings, albeit
2788 at a slight CPU cost.
2790 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2791 as requested by Marc Sherman.
2793 SC/10 Eximstats: added histograms for user specified patterns as requested
2796 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2798 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2799 fopen() in the content-scanning modules that did not already have it.
2805 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2806 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2808 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2810 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2812 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2813 to test Sieve filters that use "vacation".
2815 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2816 that changes the way the GnuTLS parameters are stored in the cache file.
2817 The new format can be generated externally. For backward compatibility,
2818 if the data in the cache doesn't make sense, Exim assumes it has read an
2819 old-format file, and it generates new data and writes a new file. This
2820 means that you can't go back to an older release without removing the
2823 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2824 work if there are any delivery delays because "one_time" forces the
2825 parent to be marked "delivered", so its unseen clone is never tried
2826 again. For this reason, Exim now forbids the simultaneous setting of
2829 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2830 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2831 entry if you want to know the details. The fix, however, neglected to
2832 consider the case when local delivery batching is involved. The test for
2833 "previously delivered" was not happening when checking to see if an
2834 address could be batched with a previous (undelivered) one; under
2835 certain circumstances this could lead to multiple deliveries to the same
2838 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2839 in its include files, and this causes problems building Exim.
2841 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2842 header_syntax) but Exim was just ignoring anything given after a slash.
2843 In particular, this caused confusion with an attempt to use "verify =
2844 reverse_host_lookup/defer_ok". An error is now given when options are
2845 supplied for verify items that do not have them. (Maybe reverse_host_
2846 lookup should have a defer_ok option, but that's a different point.)
2848 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2849 defined by RFC 821) to 2048, because there were problems with some AUTH
2850 commands, and RFC 1869 says the size should be increased for extended
2851 SMTP commands that take arguments.
2853 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2856 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2857 "unknown" error; now it says that the functionality isn't in the binary.
2859 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2860 an address' error message when a string expansion fails (syntax or
2861 whatever). Otherwise the password may appear in the log. Following change
2862 PH/42 below, there is no longer a chance of it appearing in a bounce
2865 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2867 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2868 of its addresses was ever tried. (Bugzilla bug #2.)
2870 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2871 the result incorrectly in the debug output. (It correctly added a newline
2872 to what was transported.)
2874 TF/01 Added $received_time.
2876 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2877 commented out examples of how to interface to a virus scanner and to
2878 SpamAssassin. Also added commented examples of av_scanner and
2879 spamd_address settings.
2881 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2882 and controls are allowed in which ACLs. There were a couple of minor
2883 errors. Some of the entries in the conditions table (which is a table of
2884 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2885 negation of where the condition IS allowed.
2887 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2889 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2890 header file does not have a version number, so I've had to invent a new
2891 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2892 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2893 radiusclient), but it was contributed by a Radius user.
2895 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2896 files or directories, for OpenSSL.
2898 PH/19 When an Exim process that is running as root has to create an Exim log
2899 file, it does so in a subprocess that runs as exim:exim so as to get the
2900 ownership right at creation (otherwise, other Exim processes might see
2901 the file with the wrong ownership). There was no test for failure of this
2902 fork() call, which would lead to the process getting stuck as it waited
2903 for a non-existent subprocess. Forks do occasionally fail when resources
2904 run out. I reviewed all the other calls to fork(); they all seem to check
2907 PH/20 When checking for unexpected SMTP input at connect time (before writing
2908 the banner), Exim was not dealing correctly with a non-positive return
2909 from the read() function. If the client had disconnected by this time,
2910 the result was a log entry for a synchronization error with an empty
2911 string after "input=" when read() returned zero. If read() returned -1
2912 (an event I could not check), uninitialized data bytes were printed.
2913 There were reports of junk text (parts of files, etc) appearing after
2916 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2918 PH/22 Added support for macro redefinition, and (re)definition in between
2919 driver and ACL definitions.
2921 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2922 forgetting to use the resulting value; it was using the unexpanded value.
2924 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2925 hadn't been configured. The fix is from Juergen Kreileder, who
2926 understands it better than I do:
2928 "Here's what I see happening with three configured cyrus_sasl
2929 authenticators configured (plain, login, cram-md5):
2931 On startup auth_cyrus_sasl_init() gets called for each of these.
2932 This means three calls to sasl_listmech() without a specified mech_list.
2933 => SASL tests which mechs of all available mechs actually work
2934 => three warnings about OTP not working
2935 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2937 With the patch, sasl_listmech() also gets called three times. But now
2938 SASL's mech_list option is set to the server_mech specified in the the
2939 authenticator. Or in other words, the answer from sasl_listmech()
2940 gets limited to just the mech you're testing for (which is different
2942 => the return list contains just 'plain' or 'login', 'cram-md5' or
2943 nothing depending on the value of ob->server_mech.
2945 I've just tested the patch: Authentication still works fine,
2946 unavailable mechs specified in the exim configuration are still
2947 caught, and the auth.log warnings about OTP are gone."
2949 PH/25 When debugging is enabled, the contents of the command line are added
2950 to the debugging output, even when log_selector=+arguments is not
2953 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2954 answer is "GNU", and only if the return is "GNU/something" is the answer
2957 PH/27 $acl_verify_message is now set immediately after the failure of a
2958 verification in an ACL, and so is available in subsequent modifiers. In
2959 particular, the message can be preserved by coding like this:
2961 warn !verify = sender
2962 set acl_m0 = $acl_verify_message
2964 Previously, $acl_verify_message was set only while expanding "message"
2965 and "log_message" when a very denied access.
2967 PH/28 Modified OS/os.c-Linux with
2969 -#ifndef OS_LOAD_AVERAGE
2970 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2972 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2973 nomenclature these days.)
2975 PH/29 Installed patch from the Sieve maintainer that adds the options
2976 sieve_useraddress and sieve_subaddress to the redirect router.
2978 PH/30 In these circumstances:
2979 . Two addresses routed to the same list of hosts;
2980 . First host does not offer TLS;
2981 . First host accepts first address;
2982 . First host gives temporary error to second address;
2983 . Second host offers TLS and a TLS session is established;
2984 . Second host accepts second address.
2985 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2986 and peerdn, if requested) that were in fact used only for the second
2989 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2990 attention to any local part prefix or suffix that was matched by the
2991 router that accepted the address. It now behaves in the same way as it
2992 does for delivery: the affixes are removed from the local part unless
2993 rcpt_include_affixes is set on the transport.
2995 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2996 timeout during the DATA phase of an incoming message.
2998 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2999 applied a patch sanctioned by the Sieve maintainer.
3001 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3002 the uid or gid is negative. A case of a negative gid caused this to be
3003 noticed. The fix allows for either to be negative.
3005 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3006 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3009 PH/36 The change PH/12 above was broken. Fixed it.
3011 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3012 the grounds that routing the same address twice would always produce the
3013 same answer. This might have been true once, but it is certainly no
3014 longer true now. Routing a child address may depend on the previous
3015 routing that produced that child. Some complicated redirection strategies
3016 went wrong when messages had multiple recipients, and made Exim's
3017 behaviour dependent on the order in which the addresses were given.
3019 I have moved the duplicate checking until after the routing is complete.
3020 Exim scans the addresses that are assigned to local and remote
3021 transports, and removes any duplicates. This means that more work will be
3022 done, as duplicates will always all be routed, but duplicates are
3023 presumably rare, so I don't expect this is of any significance.
3025 For deliveries to pipes, files, and autoreplies, the duplicate checking
3026 still happens during the routing process, since they are not going to be
3029 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3030 It corrects a timeout issue with spamd. This is Ian's comment: "The
3031 background is that sometimes spamd either never reads data from a
3032 connection it has accepted, or it never writes response data. The exiscan
3033 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3034 blindly assumes that writes won't block so it may never time out."
3036 PH/39 Allow G after quota size as well as K and M.
3038 PH/40 The value set for $authenticated_id in an authenticator may not contain
3039 binary zeroes or newlines because the value is written to log lines and
3040 to spool files. There was no check on this. Now the value is run through
3041 the string_printing() function so that such characters are converted to
3042 printable escape sequences.
3044 PH/41 $message_linecount is a new variable that contains the total number of
3045 lines in the message. Compare $body_linecount, which is the count for the
3048 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3049 bounce and delay warning messages, except in certain special cases, which
3052 (a) An SMTP error message from a remote host;
3053 (b) A message specified in a :fail: redirection;
3054 (c) A message specified in a "fail" command in a system filter;
3055 (d) A message specified in a FAIL return from the queryprogram router;
3056 (e) A message specified by the cannot_route_message router option.
3058 In these cases only, Exim does include the error details in bounce and
3059 warning messages. There are also a few cases where bland messages such
3060 as "unrouteable address" or "local delivery error" are given.
3062 PH/43 $value is now also set for the "else" part of a ${run expansion.
3064 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3065 being worked on, but at least Exim now implements the latest version to
3068 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3069 process to complete was treated as a delivery failure, a timeout while
3070 writing the message to the pipe was logged, but erroneously treated as a
3071 successful delivery. Such timeouts include transport filter timeouts. For
3072 consistency with the overall process timeout, these timeouts are now
3073 treated as errors, giving rise to delivery failures by default. However,
3074 there is now a new Boolean option for the pipe transport called
3075 timeout_defer, which, if set TRUE, converts the failures into defers for
3076 both kinds of timeout. A transport filter timeout is now identified in
3079 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3080 systems where "make" and "gmake" are different, calling "gmake" at top
3081 level broke things. I've arranged for the value of $(MAKE) to be passed
3082 from the Makefile to this script so that it can call the same version of
3086 A note about Exim versions 4.44 and 4.50
3087 ----------------------------------------
3089 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3090 changes of various kinds. As a consequence, a big documentation update was
3091 needed. This delayed the release for rather longer than seemed good, especially
3092 in the light of a couple of (minor) security issues. Therefore, the changes
3093 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3094 release. So 4.44 and 4.50 are in effect two different branches that both start
3097 I have left the 4.50 change log unchanged; it contains all the changes since
3098 4.43. The change log for 4.44 is below; many of its items are identical to
3099 those for 4.50. This seems to be the most sensible way to preserve the
3100 historical information.
3106 1. Minor wording change to the doc/README.SIEVE file.
3108 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3109 computation of the current number of files was incorrect.
3111 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3112 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3115 4. Give more explanation in the error message when the command for a transport
3116 filter fails to execute.
3118 5. There are several places where Exim runs a non-Exim command in a
3119 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3120 was being done only for the command run by the queryprogram router. It is
3121 now done for all such subprocesses. The other cases are: ${run, transport
3122 filters, and the commands run by the lmtp and pipe transports.
3124 6. Added CONFIGURE_GROUP build-time option.
3126 7. Some older OS have a limit of 256 on the maximum number of file
3127 descriptors. Exim was using setrlimit() to set 1000 as a large value
3128 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3129 systems. I've change it so that if it can't get 1000, it tries for 256.
3131 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3132 was an oversight, and furthermore, ever since the addition of extra
3133 controls (e.g. 4.43/32), the checks on when to allow different forms of
3134 "control" were broken. There should now be diagnostics for all cases when a
3135 control that does not make sense is encountered.
3137 9. Added the /retain_sender option to "control=submission".
3139 10. $recipients is now available in the predata ACL (oversight).
3141 11. Tidy the search cache before the fork to do a delivery from a message
3142 received from the command line. Otherwise the child will trigger a lookup
3143 failure and thereby defer the delivery if it tries to use (for example) a
3144 cached ldap connection that the parent has called unbind on.
3146 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3147 of $address_data from the recipient verification was clobbered by the
3148 sender verification.
3150 13. The value of address_data from a sender verification is now available in
3151 $sender_address_data in subsequent conditions in the ACL statement.
3153 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3155 15. Added a new option "connect=<time>" to callout options, to set a different
3158 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3159 was its contents. (It was OK if the option was not defined at all.)
3161 17. A "Completed" log line is now written for messages that are removed from
3162 the spool by the -Mrm option.
3164 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3165 information about exactly what failed.
3167 19. Added -dd to debug only the daemon process.
3169 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3170 handles timeouts, both on the server side and network timeouts. Renamed the
3171 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3173 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3174 It is now set to "smtps".
3176 22. $host_address is now set to the target address during the checking of
3177 ignore_target_hosts.
3179 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3180 being passed; this would have caused $sender_host_name to have been used if
3181 matching the list had actually called for a host name (not very likely,
3182 since this list is usually IP addresses). A host name is now passed as
3185 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3186 code that specifies a non-restarting handler (typically sigaction() in
3187 modern systems) in an attempt to fix a rare and obscure crash bug.
3189 25. Narrowed the window for a race in the daemon that could cause it to ignore
3190 SIGCHLD signals. This is not a major problem, because they are used only to
3191 wake it up if nothing else does.
3193 26. A malformed maildirsize file could cause Exim to calculate negative values
3194 for the mailbox size or file count. Odd effects could occur as a result.
3195 The maildirsize information is now recalculated if the size or filecount
3198 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3199 support for a long time. Removed HAVE_SYS_VFS_H.
3201 28. Installed the latest version of exipick from John Jetmore.
3203 29. In an address list, if the pattern was not a regular expression, an empty
3204 subject address (from a bounce message) matched only if the pattern was an
3205 empty string. Non-empty patterns were not even tested. This was the wrong
3206 because it is perfectly reasonable to use an empty address as part of a
3207 database query. An empty address is now tested by patterns that are
3208 lookups. However, all the other forms of pattern expect the subject to
3209 contain a local part and a domain, and therefore, for them, an empty
3210 address still always fails if the pattern is not itself empty.
3212 30. Exim went into a mad DNS loop when attempting to do a callout where the
3213 host was specified on an smtp transport, and looking it up yielded more
3214 than one IP address.
3216 31. Re-factored the code for checking spool and log partition space into a
3217 function that finds that data and another that does the check. The former
3218 is then used to implement four new variables: $spool_space, $log_space,
3219 $spool_inodes, and $log_inodes.
3221 32. The RFC2047 encoding function was originally intended for short strings
3222 such as real names; it was not keeping to the 75-character limit for
3223 encoded words that the RFC imposes. It now respects the limit, and
3224 generates multiple encoded words if necessary. To be on the safe side, I
3225 have increased the buffer size for the ${rfc2047: expansion operator from
3228 33. It is now permitted to omit both strings after an "if" condition; if the
3229 condition is true, the result is "true". As before, when the second string
3230 is omitted, a false condition yields an empty string. This makes it less
3231 cumbersome to write custom ACL and router conditions.
3233 34. Failure to deliver a bounce message always caused it to be frozen, even if
3234 there was an errors_to setting on the router. The errors_to setting is now
3237 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3238 canonical form (fully expanded) before being placed in
3239 $sender_host_address.
3241 36. The table in the code that translates DNS record types into text (T_A to
3242 "A" for instance) was missing entries for NS and CNAME. It is just possible
3243 that this could have caused confusion if both these types were looked up
3244 for the same domain, because the text type is used as part of Exim's
3245 per-process caching. But the chance of anyone hitting this buglet seems
3248 37. The dnsdb lookup has been extended in a number of ways.
3250 (1) There is a new type, "zns", which walks up the domain tree until it
3251 finds some nameserver records. It should be used with care.
3253 (2) There is a new type, "mxh", which is like "mx" except that it returns
3254 just the host names, not the priorities.
3256 (3) It is now possible to give a list of domains (or IP addresses) to be
3257 looked up. The behaviour when one of the lookups defers can be
3258 controlled by a keyword.
3260 (4) It is now possible to specify the separator character for use when
3261 multiple records are returned.
3263 38. The dnslists ACL condition has been extended: it is now possible to supply
3264 a list of IP addresses and/or domains to be looked up in a particular DNS
3267 39. Added log_selector=+queue_time_overall.
3269 40. When running the queue in the test harness, wait just a tad after forking a
3270 delivery process, to get repeatability of debugging output.
3272 41. Include certificate and key file names in error message when GnuTLS fails
3273 to set them up, because the GnuTLS error message doesn't include the name
3274 of the failing file when there is a problem reading it.
3276 42. Allow both -bf and -bF in the same test run.
3278 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3280 44. The "Exiscan patch" is now merged into the mainline Exim source.
3282 45. Sometimes the final signoff response after QUIT could fail to get
3283 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3284 < 0 before doing a fflush(). This bug looks as though it goes back to the
3285 introduction of TLS in release 3.20, but "sometimes" must have been rare
3286 because the tests only now provoked it.
3288 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3289 (this can affect the format of dates).
3291 47. exim_tidydb, when checking for the continued existence of a message for
3292 which it has found a message-specific retry record, was not finding
3293 messages that were in split spool directories. Consequently, it was
3294 deleting retry records that should have stayed in existence.
3296 48. Steve fixed some bugs in eximstats.
3298 49. The SPA authentication driver was not abandoning authentication and moving
3299 on to the next authenticator when an expansion was forced to fail,
3300 contradicting the general specification for all authenticators. Instead it
3301 was generating a temporary error. It now behaves as specified.
3303 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3304 (the order specifies the preference for clients). The order is now AES256,
3305 AES128, 3DES, ARCFOUR128.
3307 51. Small patch to Sieve code - explicitly set From: when generating an
3310 52. Exim crashed if a remote delivery caused a very long error message to be
3311 recorded - for instance if somebody sent an entire SpamAssassin report back
3312 as a large number of 550 error lines. This bug was coincidentally fixed by
3313 increasing the size of one of Exim's internal buffers (big_buffer) that
3314 happened as part of the Exiscan merge. However, to be on the safe side, I
3315 have made the code more robust (and fixed the comments that describe what
3318 53. Now that there can be additional text after "Completed" in log lines (if
3319 the queue_time_overall log selector is set), a one-byte patch to exigrep
3320 was needed to allow it to recognize "Completed" as not the last thing in
3323 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3324 patch that reportedly fixes this has been added. I am not expert enough to
3325 create a test for it. This is what the patch creator wrote:
3327 "I found a little strange behaviour of ldap code when working with
3328 Windows 2003 AD Domain, where users was placed in more than one
3329 Organization Units. When I tried to give exim partial DN, the exit code
3330 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3331 But simultaneously result of request was absolutely normal ldap result,
3332 so I produce this patch..."
3334 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3335 REFERENCE, so I have modified the code to exclude the patch when that macro
3338 55. Some experimental protocols are using DNS PTR records for new purposes. The
3339 keys for these records are domain names, not reversed IP addresses. The
3340 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3341 leaves it alone. Component reversal etc. now happens only for IP addresses.
3344 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3346 57. Double the size of the debug message buffer (to 2048) so that more of very
3347 long debug lines gets shown.
3349 58. The exicyclog utility now does better if the number of log files to keep
3350 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3352 59. Two changes related to the smtp_active_hostname option:
3354 (1) $smtp_active_hostname is now available as a variable.
3355 (2) The default for smtp_banner uses $smtp_active_hostname instead
3356 of $primary_hostname.
3358 60. The host_aton() function is supposed to be passed a string that is known
3359 to be a valid IP address. However, in the case of IPv6 addresses, it was
3360 not checking this. This is a hostage to fortune. Exim now panics and dies
3361 if the condition is not met. A case was found where this could be provoked
3362 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3363 components; fortuitously, this particular loophole had already been fixed
3364 by change 4.50/55 above.
3366 If there are any other similar loopholes, the new check in host_aton()
3367 itself should stop them being exploited. The report I received stated that
3368 data on the command line could provoke the exploit when Exim was running as
3369 exim, but did not say which command line option was involved. All I could
3370 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3371 running as the user.
3374 61. There was a buffer overflow vulnerability in the SPA authentication code
3375 (which came originally from the Samba project). I have added a test to the
3376 spa_base64_to_bits() function which I hope fixes it.
3379 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3380 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3382 63. The daemon start-up calls getloadavg() while still root for those OS that
3383 need the first call to be done as root, but it missed one case: when
3384 deliver_queue_load_max is set with deliver_drop_privilege. This is
3385 necessary for the benefit of the queue runner, because there is no re-exec
3386 when deliver_drop_privilege is set.
3388 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3389 This has been fixed.
3391 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3392 was in use, was not putting the data itself into the right store pool;
3393 consequently, it could be overwritten for a subsequent message in the same
3394 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3397 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3399 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3400 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3401 the function were treating the return as a boolean value, which happened to
3402 work because 0=false and not-0=true, but is not correct code.
3404 68. The host_aton() function was not handling scoped IPv6 addresses (those
3405 with, for example, "%eth0" on the end) correctly.
3407 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3408 negated items (that is, ~something) in unsigned ints. Some compilers
3409 apparently mutter when there is no cast.
3411 70. If an address verification called from an ACL failed, and did not produce a
3412 user-specific message (i.e. there was only a "system" message), nothing was
3413 put in $acl_verify_message. In this situation, it now puts the system
3416 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3417 session; change 4.31/43 added the unwanted input to the log line - except
3418 that it did not do this in the start of session case. It now does.
3420 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3421 This is wrong and can cause the other end to generate a synchronization
3422 error if it is another Exim or anything else that does the synchronization
3423 check. A QUIT command is no longer sent after a timeout.
3425 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3426 during host lookups.
3428 74. The defer_ok option of callout verification was not working if it was used
3429 when verifying addresses in header lines, that is, for this case:
3431 verify = header_sender/callout=defer_ok
3433 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3434 those file descriptors could be used for SMTP connections. If anything
3435 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3436 could be sent to the SMTP client, causing chaos. The daemon now opens
3437 stdin, stdout, and stderr to /dev/null when it puts itself into the
3440 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3441 log by default. The user can override this with suitable Perl magic.
3443 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3444 the log message when discard triggers, was not working for the DATA ACL or
3445 for the non-SMTP ACL.
3447 78. Error message wording change in sieve.c.
3449 79. If smtp_accept_max_per_host was set, the number of connections could be
3450 restricted to fewer than expected, because the daemon was trying to set up
3451 a new connection before checking whether the processes handling previous
3452 connections had finished. The check for completed processes is now done
3453 earlier. On busy systems, this bug wouldn't be noticed because something
3454 else would have woken the daemon, and it would have reaped the completed
3457 80. If a message was submitted locally by a user whose login name contained one
3458 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3459 It caused a spool format error. I have fixed the spool reading code. A
3460 related problem was that the "from" clause in the Received: line became
3461 illegal because of the space(s). It is now covered by ${quote_local_part.
3463 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3466 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3468 83. Patch from Sieve maintainer to fix unterminated string problem in
3469 "vacation" handling.
3471 84. Some minor changes to the Linux configuration files to help with other
3472 OS variants using glibc.
3474 85. One more patch for Sieve to update vacation handling to latest spec.
3477 ----------------------------------------------------
3478 See the note above about the 4.44 and 4.50 releases.
3479 ----------------------------------------------------
3485 1. Change 4.43/35 introduced a bug that caused file counts to be
3486 incorrectly computed when quota_filecount was set in an appendfile
3489 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3490 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3493 3. Give more explanation in the error message when the command for a transport
3494 filter fails to execute.
3496 4. There are several places where Exim runs a non-Exim command in a
3497 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3498 was being done only for the command run by the queryprogram router. It is
3499 now done for all such subprocesses. The other cases are: ${run, transport
3500 filters, and the commands run by the lmtp and pipe transports.
3502 5. Some older OS have a limit of 256 on the maximum number of file
3503 descriptors. Exim was using setrlimit() to set 1000 as a large value
3504 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3505 systems. I've change it so that if it can't get 1000, it tries for 256.
3507 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3508 was an oversight, and furthermore, ever since the addition of extra
3509 controls (e.g. 4.43/32), the checks on when to allow different forms of
3510 "control" were broken. There should now be diagnostics for all cases when a
3511 control that does not make sense is encountered.
3513 7. $recipients is now available in the predata ACL (oversight).
3515 8. Tidy the search cache before the fork to do a delivery from a message
3516 received from the command line. Otherwise the child will trigger a lookup
3517 failure and thereby defer the delivery if it tries to use (for example) a
3518 cached ldap connection that the parent has called unbind on.
3520 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3521 of $address_data from the recipient verification was clobbered by the
3522 sender verification.
3524 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3525 was its contents. (It was OK if the option was not defined at all.)
3527 11. A "Completed" log line is now written for messages that are removed from
3528 the spool by the -Mrm option.
3530 12. $host_address is now set to the target address during the checking of
3531 ignore_target_hosts.
3533 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3534 being passed; this would have caused $sender_host_name to have been used if
3535 matching the list had actually called for a host name (not very likely,
3536 since this list is usually IP addresses). A host name is now passed as
3539 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3540 code that specifies a non-restarting handler (typically sigaction() in
3541 modern systems) in an attempt to fix a rare and obscure crash bug.
3543 15. Narrowed the window for a race in the daemon that could cause it to ignore
3544 SIGCHLD signals. This is not a major problem, because they are used only to
3545 wake it up if nothing else does.
3547 16. A malformed maildirsize file could cause Exim to calculate negative values
3548 for the mailbox size or file count. Odd effects could occur as a result.
3549 The maildirsize information is now recalculated if the size or filecount
3552 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3553 support for a long time. Removed HAVE_SYS_VFS_H.
3555 18. Updated exipick to current release from John Jetmore.
3557 19. Allow an empty sender to be matched against a lookup in an address list.
3558 Previously the only cases considered were a regular expression, or an
3561 20. Exim went into a mad DNS lookup loop when doing a callout where the
3562 host was specified on the transport, if the DNS lookup yielded more than
3565 21. The RFC2047 encoding function was originally intended for short strings
3566 such as real names; it was not keeping to the 75-character limit for
3567 encoded words that the RFC imposes. It now respects the limit, and
3568 generates multiple encoded words if necessary. To be on the safe side, I
3569 have increased the buffer size for the ${rfc2047: expansion operator from
3572 22. Failure to deliver a bounce message always caused it to be frozen, even if
3573 there was an errors_to setting on the router. The errors_to setting is now
3576 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3577 canonical form (fully expanded) before being placed in
3578 $sender_host_address.
3580 24. Updated eximstats to version 1.33
3582 25. Include certificate and key file names in error message when GnuTLS fails
3583 to set them up, because the GnuTLS error message doesn't include the name
3584 of the failing file when there is a problem reading it.
3586 26. Expand error message when OpenSSL has problems setting up cert/key files.
3589 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3590 (this can affect the format of dates).
3592 28. exim_tidydb, when checking for the continued existence of a message for
3593 which it has found a message-specific retry record, was not finding
3594 messages that were in split spool directories. Consequently, it was
3595 deleting retry records that should have stayed in existence.
3597 29. eximstats updated to version 1.35
3598 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3599 1.35 - bugfix such that pie charts by volume are generated correctly
3601 30. The SPA authentication driver was not abandoning authentication and moving
3602 on to the next authenticator when an expansion was forced to fail,
3603 contradicting the general specification for all authenticators. Instead it
3604 was generating a temporary error. It now behaves as specified.
3606 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3607 (the order specifies the preference for clients). The order is now AES256,
3608 AES128, 3DES, ARCFOUR128.
3610 31. Small patch to Sieve code - explicitly set From: when generating an
3613 32. Exim crashed if a remote delivery caused a very long error message to be
3614 recorded - for instance if somebody sent an entire SpamAssassin report back
3615 as a large number of 550 error lines. This bug was coincidentally fixed by
3616 increasing the size of one of Exim's internal buffers (big_buffer) that
3617 happened as part of the Exiscan merge. However, to be on the safe side, I
3618 have made the code more robust (and fixed the comments that describe what
3621 33. Some experimental protocols are using DNS PTR records for new purposes. The
3622 keys for these records are domain names, not reversed IP addresses. The
3623 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3624 leaves it alone. Component reversal etc. now happens only for IP addresses.
3627 34. The host_aton() function is supposed to be passed a string that is known
3628 to be a valid IP address. However, in the case of IPv6 addresses, it was
3629 not checking this. This is a hostage to fortune. Exim now panics and dies
3630 if the condition is not met. A case was found where this could be provoked
3631 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3632 components; fortuitously, this particular loophole had already been fixed
3633 by change 4.50/55 or 4.44/33 above.
3635 If there are any other similar loopholes, the new check in host_aton()
3636 itself should stop them being exploited. The report I received stated that
3637 data on the command line could provoke the exploit when Exim was running as
3638 exim, but did not say which command line option was involved. All I could
3639 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3640 running as the user.
3643 35. There was a buffer overflow vulnerability in the SPA authentication code
3644 (which came originally from the Samba project). I have added a test to the
3645 spa_base64_to_bits() function which I hope fixes it.
3648 36. The daemon start-up calls getloadavg() while still root for those OS that
3649 need the first call to be done as root, but it missed one case: when
3650 deliver_queue_load_max is set with deliver_drop_privilege. This is
3651 necessary for the benefit of the queue runner, because there is no re-exec
3652 when deliver_drop_privilege is set.
3654 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3655 was in use, was not putting the data itself into the right store pool;
3656 consequently, it could be overwritten for a subsequent message in the same
3657 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3660 38. Sometimes the final signoff response after QUIT could fail to get
3661 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3662 < 0 before doing a fflush(). This bug looks as though it goes back to the
3663 introduction of TLS in release 3.20, but "sometimes" must have been rare
3664 because the tests only now provoked it.
3670 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3671 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3672 Now it returns an integer. A number of calls were still expecting a T/F
3673 return. Fortuitously, in all cases, the tests worked in OK situations,
3674 which is the norm. However, things would have gone wrong on any write
3675 failures on the smtp file descriptor. This function is used when sending
3676 messages over SMTP and also when doing verify callouts.
3678 2. When Exim is called to do synchronous delivery of a locally submitted
3679 message (the -odf or -odi options), it no longer closes stderr before doing
3682 3. Implemented the mua_wrapper option.
3684 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3686 5. Implemented the functions header_remove(), header_testname(),
3687 header_add_at_position(), and receive_remove_recipient(), and exported them
3690 6. If an ACL "warn" statement specified the addition of headers, Exim already
3691 inserted X-ACL-Warn: at the start if there was no header name. However, it
3692 was not making this test for the second and subsequent header lines if
3693 there were newlines in the string. This meant that an invalid header could
3694 be inserted if Exim was badly configured.
3696 7. Allow an ACL "warn" statement to add header lines at the start or after all
3697 the Received: headers, as well as at the end.
3699 8. Added the rcpt_4xx retry error code.
3701 9. Added postmaster_mailfrom=xxx to callout verification option.
3703 10. Added mailfrom=xxxx to the callout verification option, for verify=
3706 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3707 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3709 12. Inserted some casts to stop certain compilers warning when using pointer
3710 differences as field lengths or precisions in printf-type calls (mostly
3711 affecting debugging statements).
3713 13. Added optional readline() support for -be (dynamically loaded).
3715 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3716 same clock tick as a message's arrival, so that its received time was the
3717 same as the "first fail" time on the retry record, and that message
3718 remained on the queue past the ultimate address timeout, every queue runner
3719 would try a delivery (because it was past the ultimate address timeout) but
3720 after another failure, the ultimate address timeout, which should have then
3721 bounced the address, did not kick in. This was a "< instead of <=" error;
3722 in most cases the first failure would have been in the next clock tick
3723 after the received time, and all would be well.
3725 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3726 being recognized when the domain list was tested by the match_domain
3727 condition in an expansion string.
3729 16. Added the ${str2b64: operator.
3731 17. Exim was always calling setrlimit() to set a large limit for the number of
3732 processes, without checking whether the existing limit was already
3733 adequate. (It did check for the limit on file descriptors.) Furthermore,
3734 errors from getrlimit() and setrlimit() were being ignored. Now they are
3735 logged to the main and panic logs, but Exim does carry on, to try to do its
3736 job under whatever limits there are.
3738 18. Imported PCRE 5.0.
3740 19. Trivial typo in log message " temporarily refused connection" (the leading
3743 20. If the log selector return_path_on_delivery was set and an address was
3744 redirected to /dev/null, the delivery process crashed because it assumed
3745 that a return path would always be set for a "successful" delivery. In this
3746 case, the whole delivery is bypassed as an optimization, and therefore no
3749 21. Internal re-arrangement: the function for sending a challenge and reading
3750 a response while authentication was assuming a zero-terminated challenge
3751 string. It's now changed to take a pointer and a length, to allow for
3752 binary data in such strings.
3754 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3756 23. Exim was not respecting finduser_retries when seeking the login of the
3757 uid under which it was called; it was always trying 10 times. (The default
3758 setting of finduser_retries is zero.) Also, it was sleeping after the final
3759 failure, which is pointless.
3761 24. Implemented tls_on_connect_ports.
3763 25. Implemented acl_smtp_predata.
3765 26. If the domain in control=submission is set empty, Exim assumes that the
3766 authenticated id is a complete email address when it generates From: or
3767 Sender: header lines.
3769 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3770 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3771 chown and chgrp in /bin and hostname in /usr/bin.
3773 28. Exim was keeping the "process log" file open after each use, just as it
3774 does for the main log. This opens the possibility of it remaining open for
3775 long periods when the USR1 signal hits a daemon. Occasional processlog
3776 errors were reported, that could have been caused by this. Anyway, it seems
3777 much more sensible not to leave this file open at all, so that is what now
3780 29. The long-running daemon process does not normally write to the log once it
3781 has entered its main loop, and it closes the log before doing so. This is
3782 so that log files can straightforwardly be renamed and moved. However,
3783 there are a couple of unusual error situations where the daemon does write
3784 log entries, and I had neglected to close the log afterwards.
3786 30. The text of an SMTP error response that was received during a remote
3787 delivery was being truncated at 512 bytes. This is too short for some of
3788 the long messages that one sometimes sees. I've increased the limit to
3791 31. It is now possible to make retry rules that apply only when a message has a
3792 specific sender, in particular, an empty sender.
3794 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3795 it possible to be selective about when SMTP synchronization is enforced.
3797 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3799 32. Implemented hosts_connection_nolog.
3801 33. Added an ACL for QUIT.
3803 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3806 35. Added mailbox_size and mailbox_filecount to appendfile.
3808 36. Added control = no_multiline_responses to ACLs.
3810 37. There was a bug in the logic of the code that waits for the clock to tick
3811 in the case where the clock went backwards by a substantial amount such
3812 that the microsecond fraction of "now" was more than the microsecond
3813 fraction of "then" (but the whole seconds number was less).
3815 38. Added support for the libradius Radius client library this is found on
3816 FreeBSD (previously only the radiusclient library was supported).
3822 1. When certain lookups returned multiple values in the form name=value, the
3823 quoting of the values was not always being done properly. Specifically:
3824 (a) If the value started with a double quote, but contained no whitespace,
3826 (b) If the value contained whitespace other than a space character (i.e.
3827 tabs or newlines or carriage returns) it was not quoted.
3828 This fix has been applied to the mysql and pgsql lookups by writing a
3829 separate quoting function and calling it from the lookup code. The fix
3830 should probably also be applied to nisplus, ibase and oracle lookups, but
3831 since I cannot test any of those, I have not disturbed their existing code.
3833 2. A hit in the callout cache for a specific address caused a log line with no
3834 reason for rejecting RCPT. Now it says "Previous (cached) callout
3835 verification failure".
3837 3. There was an off-by-one bug in the queryprogram router. An over-long
3838 return line was truncated at 256 instead of 255 characters, thereby
3839 overflowing its buffer with the terminating zero. As well as fixing this, I
3840 have increased the buffer size to 1024 (and made a note to document this).
3842 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3843 when Exim is waiting for an SMTP response from a remote server, Exim
3844 restarts its select() call on the socket, thereby resetting its timeout.
3845 This is not a problem when such interrupts are rare. Somebody set up a cron
3846 job to run exiwhat every 2 minutes, which is less than the normal select()
3847 timeout (5 or 10 minutes). This meant that the select() timeout never
3848 kicked in because it was always reset. I have fixed this by comparing the
3849 time when an interrupt arrives with the time at the start of the first call
3850 to select(). If more time than the timeout has elapsed, the interrupt is
3851 treated as a timeout.
3853 5. Some internal re-factoring in preparation for the addition of Sieve
3854 extensions (by MH). In particular, the "personal" test is moved to a
3855 separate function, and given an option for scanning Cc: and Bcc: (which is
3856 not set for Exim filters).
3858 6. When Exim created an email address using the login of the caller as the
3859 local part (e.g. when creating a From: or Sender: header line), it was not
3860 quoting the local part when it contained special characters such as @.
3862 7. Installed new OpenBSD configuration files.
3864 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3865 try to make them clearer.
3867 9. Callout options, other than the timeout value, were being ignored when
3868 verifying sender addresses in header lines. For example, when using
3870 verify = header_sender/callout=no_cache
3872 the cache was (incorrectly) being used.
3874 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3875 only the Cygwin environment.
3877 11. When return_path_on_delivery was set as a log selector, if different remote
3878 addresses in the same message used different return paths and parallel
3879 remote delivery occurred, the wrong values would sometimes be logged.
3880 (Whenever a remote delivery process finished, the return path value from
3881 the most recently started remote delivery process was logged.)
3883 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3884 lines when AUTH and/or TLS are in use. This is the "received protocol"
3885 field. Exim used to use "asmtp" for authenticated SMTP, without any
3886 indication (in the protocol name) for TLS use. Now it follows the RFC and
3887 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3888 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3889 names appear in log lines as well as in Received: header lines.
3891 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3892 extensions, and comparison tests, and to fix some bugs.
3894 14. Changes to the "personal" filter test:
3896 (1) The test was buggy in that it was just doing the equivalent of
3897 "contains" tests on header lines. For example, if a user's address was
3898 anne@some.where, the "personal" test would incorrectly be true for
3900 To: susanne@some.where
3902 This test is now done by extracting each address from the header in turn,
3903 and checking the entire address. Other tests that are part of "personal"
3904 are now done using regular expressions (for example, to check local parts
3905 of addresses in From: header lines).
3907 (2) The list of non-personal local parts in From: addresses has been
3908 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3909 taken from the Sieve specification recommendations.
3911 (3) If the message contains any header line starting with "List-" it is
3912 treated as non-personal.
3914 (4) The test for "circular" in the Subject: header line has been removed
3915 because it now seems ill-conceived.
3917 15. Minor typos in src/EDITME comments corrected.
3919 16. Installed latest exipick from John Jetmore.
3921 17. If headers_add on a router specified a text string that was too long for
3922 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3923 of string_sprintf() is now avoided.
3925 18. $message_body_size was not set (it was always zero) when running the DATA
3926 ACL and the local_scan() function.
3928 19. For the "mail" command in an Exim filter, no default was being set for
3929 the once_repeat time, causing a random time value to be used if "once" was
3930 specified. (If the value happened to be <= 0, no repeat happened.) The
3931 default is now 0s, meaning "never repeat". The "vacation" command was OK
3932 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3933 (I found it when inspecting the code).
3935 20. There is now an overall timeout for performing a callout verification. It
3936 defaults to 4 times the callout timeout, which applies to individual SMTP
3937 commands during the callout. The overall timeout applies when there is more
3938 than one host that can be tried. The timeout is checked before trying the
3939 next host. This prevents very long delays if there are a large number of
3940 hosts and all are timing out (e.g. when the network connections are timing
3941 out). The value of the overall timeout can be changed by specifying an
3942 additional sub-option for "callout", called "maxwait". For example:
3944 verify = sender/callout=5s,maxwait=20s
3946 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3947 to the end before writing, but this should make it even safer).
3949 22. Exim was forgetting that it had advertised PIPELINING for the second and
3950 subsequent messages on an SMTP connection. It was also not resetting its
3951 memory on STARTTLS and an internal HELO.
3953 23. When Exim logs an SMTP synchronization error within a session, it now
3954 records whether PIPELINING has been advertised or not.
3956 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3957 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3958 rather than long int.
3960 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3962 26. Added the never_mail option to autoreply.
3968 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3969 crash if the getsockname() call failed; this can happen if a connection is
3970 closed very soon after it is established. The problem was simply in the
3971 order in which certain operations were done, causing Exim to try to write
3972 to the SMTP stream before it had set up the file descriptor. The bug has
3973 been fixed by making things happen in the correct order.
3979 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3980 before the connection was closed, thus losing the rejection response.
3982 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3983 some early Solaris releases, but causes trouble in current releases where
3984 socklen_t is defined.
3986 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3989 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3992 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3993 that releasing the top of it at the end releases what was used for sub-
3994 expansions (unless the block got too big). However, discard this block if
3995 the first thing is a variable or header, so that we can use its block when
3996 it is dynamic (useful for very large $message_headers, for example).
3998 6. Lookups now cache *every* query, not just the most recent. A new, separate
3999 store pool is used for this. It can be recovered when all lookup caches are
4000 flushed. Lookups now release memory at the end of their result strings.
4001 This has involved some general refactoring of the lookup sources.
4003 7. Some code has been added to the store_xxx() functions to reduce the amount
4004 of flapping under certain conditions.
4006 8. log_incoming_interface used to affect only the <= reception log lines. Now
4007 it causes the local interface and port to be added to several more SMTP log
4008 lines, for example "SMTP connection from", and rejection lines.
4010 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4012 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4014 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4015 could be overwritten at the end of the current message (or the start of a
4016 new message if it was set in a HELO ACL). The value is now preserved for
4017 the duration of the SMTP connection.
4019 12. If a transport had a headers_rewrite setting, and a matching header line
4020 contained an unqualified address, that address was qualified, even if it
4021 did not match any rewriting rules. The underlying bug was that the values
4022 of the flags that permit the existence of unqualified sender and recipient
4023 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4024 non-local messages, and by -bnq for local messages) were not being
4025 preserved with the message after it was received.
4027 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4028 "next input=" as part of the text comprising the host identity instead of
4029 the correct text. The code was using the same buffer for two different
4030 strings. However, depending on which order the printing function evaluated
4031 its arguments, the bug did not always show up. Under Linux, for example, my
4032 test suite worked just fine.
4034 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4035 doesn't work with some older versions of Perl. It has been changed to "my",
4036 which in any case is probably the better facility to use.
4038 15. A really picky compiler found some instances of statements for creating
4039 error messages that either had too many or two few arguments for the format
4042 16. The size of the buffer for calls to the DNS resolver has been increased
4043 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4044 for addresses that have a lot of PTR records. This alleviates a problem; it
4045 does not fully solve it.
4047 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4048 buffer now truncates the list and logs the incident, which is the same
4049 action as happens when Exim is looking up a host name and its aliases.
4050 Previously in this situation something unpredictable would happen;
4051 sometimes it was "internal error: store_reset failed".
4053 18. If a server dropped the connection unexpectedly when an Exim client was
4054 using GnuTLS and trying to read a response, the client delivery process
4055 crashed while trying to generate an error log message.
4057 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4058 string, for example:
4060 warn message = H1: something\nH2: something
4062 the text was added as a single header line from Exim's point of view
4063 though it ended up OK in the delivered message. However, searching for the
4064 second and subsequent header lines using $h_h2: did not work. This has been
4065 fixed. Similarly, if a system filter added multiple headers in this way,
4066 the routers could not see them.
4068 20. Expanded the error message when iplsearch is called with an invalid key to
4069 suggest using net-iplsearch in a host list.
4071 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4072 ACLs are no longer actually imposed (and a message to that effect is
4075 22. If a "gecos" field in a passwd entry contained escaped characters, in
4076 particular, if it contained a \" sequence, Exim got it wrong when building
4077 a From: or a Sender: header from that name. A second bug also caused
4078 incorrect handling when an unquoted " was present following a character
4079 that needed quoting.
4081 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4082 was not being matched caselessly.
4084 24. Arranged for all hyphens in the exim.8 source to be escaped with
4087 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4088 properly. Recipient callout cache records were still being keyed to include
4089 the sender, even when use_sender was set false. This led to far more
4090 callouts that were necessary. The sender is no longer included in the key
4091 when use_sender is false.
4093 26. Added "control = submission" modifier to ACLs.
4095 27. Added the ${base62d: operator to decode base 62 numbers.
4097 28. dnsdb lookups can now access SRV records.
4099 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4100 the configuration file.
4102 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4103 (-v) mode. This makes the output for a verbose queue run more intelligible.
4105 31. Added a use_postmaster feature to recipient callouts.
4107 32. Added the $body_zerocount variable, containing the number of binary zero
4108 bytes in the message body.
4110 33. The time of last modification of the "new" subdirectory is now used as the
4111 "mailbox time last read" when there is a quota error for a maildir
4114 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4116 35. Added +ignore_unknown as a special item in host lists.
4118 36. Code for decoding IPv6 addresses in host lists is now included, even if
4119 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4120 address was recognized as an IP address, but was then not correctly decoded
4121 into binary, causing unexpected and incorrect effects when compared with
4128 1. Very minor rewording of debugging text in manualroute to say "list of
4129 hosts" instead of "hostlist".
4131 2. If verify=header_syntax was set, and a header line with an unqualified
4132 address (no domain) and a large number of spaces between the end of the
4133 name and the colon was received, the reception process suffered a buffer
4134 overflow, and (when I tested it) crashed. This was caused by some obsolete
4135 code that should have been removed. The fix is to remove it!
4137 3. When running in the test harness, delay a bit after writing a bounce
4138 message to get a bit more predictability in the log output.
4140 4. Added a call to search_tidyup() just before forking a reception process. In
4141 theory, someone could use a lookup in the expansion of smtp_accept_max_
4142 per_host which, without the tidyup, could leave open a database connection.
4144 5. Added the variables $recipient_data and $sender_data which get set from a
4145 lookup success in an ACL "recipients" or "senders" condition, or a router
4146 "senders" option, similar to $domain_data and $local_part_data.
4148 6. Moved the writing of debug_print from before to after the "senders" test
4151 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4152 problems for message scanning, either using a data ACL, or using
4153 local_scan() because the Received: header was not generated till after they
4154 were called (in order to set the time as the time of reception completion).
4155 I have revised the way this works. The header is now generated after the
4156 body is received, but before the ACL or local_scan() are called. After they
4157 are run, the timestamp in the header is updated.
4163 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4164 before starting a queue runner without re-exec. This happened only when
4165 deliver_drop_privilege was set or when the Exim user was set to root. The
4166 effect of the bug was that timeouts during subsequent deliveries caused
4167 crashes instead of being properly handled. The handler is now left at its
4168 default (and expected) setting.
4170 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4171 message, again when deliver_drop_privilege is set or Exim is run as root.
4172 The bug described in (1) was not present in this case, but the tidying up
4173 of the other signals was missing. I have made the two cases consistent.
4175 3. The ignore_target_hosts setting on a manualroute router was being ignored
4176 for hosts that were looked up using the /MX notation.
4178 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4181 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4182 operated on the sender address. After changing the $sender_address to <>
4183 for the sender address verify, Exim was re-instated it as the original
4184 (before rewriting) address, but remembering that it had rewritten it, so it
4185 wasn't rewriting it again. This bug also had the effect of breaking the
4186 sender address verification caching when the sender address was rewritten.
4188 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4189 This has been changed so that if the ip literal address matches
4190 ignore_target_hosts, the router declines.
4192 7. Added expansion conditions match_domain, match_address, and match_local_
4193 part (NOT match_host).
4195 8. The placeholder for the Received: header didn't have a length field set.
4197 9. Added code to Exim itself and to exim_lock to test for a specific race
4198 condition that could lead to file corruption when using MBX delivery. The
4199 issue is with the lockfile that is created in /tmp. If this file is removed
4200 after a process has opened it but before that process has acquired a lock,
4201 there is the potential for a second process to recreate the file and also
4202 acquire a lock. This could lead to two Exim processes writing to the file
4203 at the same time. The added code performs the same test as UW imapd; it
4204 checks after acquiring the lock that its file descriptor still refers to
4205 the same named file.
4207 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4208 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4211 11. Added the smtp_active_hostname option. If used, this will typically be made
4212 to depend on the incoming interface address. Because $interface_address is
4213 not set up until the daemon has forked a reception process, error responses
4214 that can happen earlier (such as "too many connections") no longer contain
4217 12. If an expansion in a condition on a "warn" statement fails because a lookup
4218 defers, the "warn" statement is abandoned, and the next ACL statement is
4219 processed. Previously this caused the whole ACL to be aborted.
4221 13. Added the iplsearch lookup type.
4223 14. Added ident_timeout as a log selector.
4225 15. Added tls_certificate_verified as a log selector.
4227 16. Added a global option tls_require_ciphers (compare the smtp transport
4228 option of the same name). This controls incoming TLS connections.
4230 17. I finally figured out how to make tls_require_ciphers do a similar thing
4231 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4232 before starting the TLS session.
4234 18. Tabs are now shown as \t in -bP output.
4236 19. If the log selector return_path_on_delivery was set, Exim crashed when
4237 bouncing a message because it had too many Received: header lines.
4239 20. If two routers both had headers_remove settings, and the first one included
4240 a superfluous trailing colon, the final name in the first list and the
4241 first name in the second list were incorrectly joined into one item (with a
4242 colon in the middle).
4248 1. Added -C and -D options to the exinext utility, mainly to make it easier
4249 to include in the automated testing, but these could be helpful when
4250 multiple configurations are in use.
4252 2. The exinext utility was not formatting the output nicely when there was
4253 an alternate port involved in the retry record key, nor when there was a
4254 message id as well (for retries that were specific to a specific message
4255 and a specific host). It was also confused by IPv6 addresses, because of
4256 the additional colons they contain. I have fixed the IPv4 problem, and
4257 patched it up to do a reasonable job for IPv6.
4259 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4260 delivery, the log line now contains "pipelined" if PIPELINING was used.
4262 4. An SMTP transport process used to panic and die if the bind() call to set
4263 an explicit outgoing interface failed. This has been changed; it is now
4264 treated in the same way as a connect() failure.
4266 5. A reference to $sender_host_name in the part of a conditional expansion
4267 that was being skipped was still causing a DNS lookup. This no longer
4270 6. The def: expansion condition was not recognizing references to header lines
4271 that used bh_ and bheader_.
4273 7. Added the _cache feature to named lists.
4275 8. The code for checking quota_filecount in the appendfile transport was
4276 allowing one more file than it should have been.
4278 9. For compatibility with Sendmail, the command line option
4286 and sets the incoming protocol and host name (for trusted callers). The
4287 host name and its colon can be omitted when only the protocol is to be set.
4288 Note the Exim already has two private options, -pd and -ps, that refer to
4289 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4290 "s", but I don't think that's a major issue.
4292 10. A number of refactoring changes to the code, none of which should affect
4295 (a) The number of logging options was getting close to filling up the
4296 32-bit word that was used as a bit map. I have split them into two classes:
4297 those that are passed in the argument to log_write(), and those that are
4298 only ever tested independently outside of that function. These are now in
4299 separate 32-bit words, so there is plenty of room for expansion again.
4300 There is no change in the user interface or the logging behaviour.
4302 (b) When building, for example, log lines, the code previously used a
4303 macro that called string_cat() twice, in order to add two strings. This is
4304 not really sufficiently general. Furthermore, there was one instance where
4305 it was actually wrong because one of the argument was used twice, and in
4306 one call a function was used. (As it happened, calling the function twice
4307 did not affect the overall behaviour.) The macro has been replaced by a
4308 function that can join an arbitrary number of extra strings onto a growing
4311 (c) The code for expansion conditions now uses a table and a binary chop
4312 instead of a serial search (which was left over from when there were very
4313 few conditions). Also, it now recognizes conditions like "pam" even when
4314 the relevant support is not compiled in: a suitably worded error message is
4315 given if an attempt is made to use such a condition.
4317 11. Added ${time_interval:xxxxx}.
4319 12. A bug was causing one of the ddress fields not to be passed back correctly
4320 from remote delivery subprocesses. The field in question was not being
4321 subsequently used, so this caused to problems in practice.
4323 13. Added new log selectors queue_time and deliver_time.
4325 14. Might have fixed a bug in maildirsizefile handling that threw up
4326 "unexpected character" debug warnings, and recalculated the data
4327 unnecessarily. In any case, I expanded the warning message to give more
4330 15. Added the message "Restricted characters in address" to the statements in
4331 the default ACL that block characters like @ and % in local parts.
4333 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4334 Three changes have been made:
4336 (a) There was a serious bug; a negative response to MAIL caused the whole
4337 recipient domain to be cached as invalid, thereby blocking all messages
4338 to all local parts at the same domain, from all senders. This bug has
4339 been fixed. The domain is no longer cached after a negative response to
4340 MAIL if the sender used is not empty.
4342 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4345 (c) A new callout option, "use_sender" has been added for people who want
4346 the modified behaviour.
4352 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4355 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4356 indeed breaks things for older releases.
4358 3. Added additional logging to the case where there is a problem reading data
4359 from a filter that is running in a subprocess using a pipe, in order to
4360 try to track down a specific problem.
4362 4. Testing facility fudge: when running in the test harness and attempting
4363 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4364 getting "No route to host". Convert this to a timeout.
4366 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4369 6. Some OS don't have socklen_t but use size_t instead. This affects the
4370 fifth argument of getsockopt() amongst other things. This is now
4371 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4372 can be set for individual OS. I have set it for SunOS5, OSF1, and
4373 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4374 some earlier ones do not.
4376 7. Change 4.30/15 was not doing the test caselessly.
4378 8. The standard form for an IPv6 address literal was being rejected by address
4379 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4380 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4381 this, as well as the form without the "IPv6" on the front (but only when
4382 address literals are enabled, of course).
4384 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4386 10. Exim crashed if a message with an empty sender address specified by -f
4387 encountered a router with an errors_to setting. This could be provoked only
4388 by a command such as
4392 where an empty string was supplied; "<>" did not hit this bug.
4394 11. Installed PCRE release 4.5.
4396 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4397 remained set. It is now erased.
4399 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4400 times from message ids (which are base 36 rather than the normal 62).
4402 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4403 were being counted as actual protocol errors, and logged if the log
4404 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4405 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4406 and DATA following a set of rejected RCPTs do not count as protocol errors.
4407 In other words, Exim assumes they were pipelined, though this may not
4408 actually be the case. Of course, in all cases the client gets an
4409 appropriate error code.
4411 15. If a lookup fails in an ACL condition, a message about the failure may
4412 be available; it is used if testing the ACL cannot continue, because most
4413 such messages specify what the cause of the deferral is. However, some
4414 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4415 that caused an old message to be retained and used if a later statement
4416 caused a defer, replacing the real cause of the deferral.
4418 16. If an IP address had so many PTR records that the DNS lookup buffer
4419 was not large enough to hold them, Exim could crash while trying to process
4420 the truncated data. It now detects and logs this case.
4422 17. Further to 4.21/58, another change has been made: if (and only if) the
4423 first line of a message (the first header line) ends with CRLF, a bare LF
4424 in a subsequent header line has a space inserted after it, so as not to
4425 terminate the header.
4427 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4428 unnecessarily, used atoi() instead of strtol(), and didn't check the
4429 termination when getting file sizes from file names by regex.
4431 19. Completely re-implemented the support for maildirsize files, in the light
4432 of a number of problems with the previous contributed implementation
4433 (4.30/29). In particular:
4435 . If the quota is zero, the maildirsize file is maintained, but no quota is
4438 . If the maildir directory does not exist, it is created before any attempt
4439 to write a maildirsize file.
4441 . The quota value in the file is just a cache; if the quota is changed in
4442 the transport, the new value overrides.
4444 . A regular expression is available for excluding directories from the
4447 20. The autoreply transport checks the characters in options that define the
4448 message's headers; it allows continued headers, but it was checking with
4449 isspace() after an embedded newline instead of explicitly looking for a
4452 21. If all the "regular" hosts to which an address was routed had passed their
4453 expiry times, and had not reached their retry times, the address was
4454 bounced, even if fallback hosts were defined. Now Exim should go on to try
4457 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4458 equivalent code in the SMTP transport. Some hosts send humungous responses
4459 to HELO/EHLO, more than 1024 it seems.
4461 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4462 gives compiler warnings in some environments. I've now done it "properly",
4465 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4466 (because of problems with the built-in one) was declared to return uschar *
4467 instead of char *, causing compiler failure.
4469 25. Fixed a file descriptor leak when processing alias/forward files.
4471 26. Fixed a minor format string issue in dbfn.c.
4473 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4475 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4476 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4479 29. When a sender address is verified, it is cached, to save repeating the test
4480 when there is more than one recipient in a message. However, when the
4481 verification involves a callout, it is possible for different callout
4482 options to be set for different recipients. It is too complicated to keep
4483 track of this in the cache, so now Exim always runs a verification when a
4484 callout is required, relying on the callout cache for the optimization.
4485 The overhead is duplication of the address routing, but this should not be
4488 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4489 to be verified with callout=postmaster, and the main callout worked but the
4490 postmaster check failed, the verification correctly failed. However, if a
4491 subsequent RCPT command asked for sender verification *without* the
4492 postmaster check, incorrect caching caused this verification also to fail,
4495 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4496 it was not caching the DNS options (qualify_single, search_parents) that
4497 were used when the lookup failed. A subsequent lookup with different
4498 options therefore always gave the same answer, though there were cases
4499 where it should not have. (Example: a "domains = !$mx_any" option on a
4500 dnslookup router: the "domains" option is always processed without any
4501 widening, but the router might have qualify_single set.) Now Exim uses the
4502 cached value only when the same options are set.
4504 32. Added John Jetmore's "exipick" utility to the distribution.
4506 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4507 than a timeout (e.g. a certificate is required, and is not provided), an
4508 Exim server now closes the connection immediately. Previously it waited for
4509 the client to close - but if the client is SSL, it seems that they each
4510 wait for each other, leading to a delay before one of them times out.
4512 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4513 maintained 0.8.x compatibility because I don't think many are using it, and
4514 it is clearly obsolete.
4516 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4519 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4520 client certificate was expired. A simple patch fixes this, though I don't
4521 understand the full logic of why the verify callback is called multiple
4524 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4525 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4526 which causes problems with some clients (such as the Certicom SSL Plus
4527 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4528 disables the coutermeasure allowing Eudora to connect."
4530 38. Exim was not checking that a write() to a log file succeeded. This could
4531 lead to Bad Things if a log got too big, in particular if it hit a file
4532 size limit. Exim now panics and dies if it cannot write to a log file, just
4533 as it does if it cannot open a log file.
4535 39. Modified OS/Makefile-Linux so that it now contains
4537 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4539 The two -D definitions ensure that Exim is compiled with large file
4540 support, which makes it possible to handle log files that are bigger than
4543 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4544 instance) a domain was checked against a named list that involved a lookup,
4545 causing $domain_data to be set, then another domain was checked against the
4546 same list, then the first domain was re-checked, the value of $domain_data
4547 after the final check could be wrong. In particular, if the second check
4548 failed, it could be set empty. This bug probably also applied to
4551 41. The strip_trailing_dot option was not being applied to the address given
4552 with the -f command-line option.
4554 42. The code for reading a message's header from the spool was incrementing
4555 $received_count, but never initializing it. This meant that the value was
4556 incorrect (doubled) while delivering a message in the same process in which
4557 it was received. In the most common configuration of Exim, this never
4558 happens - a fresh exec is done - but it can happen when
4559 deliver_drop_privilege is set.
4561 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4562 it now includes up to 150 characters of the unexpected data in the log
4565 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4566 and building data strings. The size of both of these buffers was 10 000
4567 bytes - far larger than anybody would *ever* want, thought I. Needless to
4568 say, somebody hit the limit. I have increased the maximum line length to
4569 20 000 and the maximum data length of concatenated lines to 100 000. I have
4570 also fixed two bugs, because there was no checking on these buffers. Tsk,
4571 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4572 buffer is too small.
4574 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4575 lsearch lookups. Now it does.
4577 46. When parsing a route_list item in a manualroute router, a fixed-length
4578 buffer was used for the list of hosts. I made this 1024 bytes long,
4579 thinking that nobody would ever have a list of hosts that long. Wrong.
4580 Somebody had a whole pile of complicated expansion conditions, and the
4581 string was silently truncated, leading to an expansion error. It turns out
4582 that it is easier to change to an unlimited length (owing to other changes
4583 that have happened since this code was originally written) than to build
4584 structure for giving a limitation error. The length of the item that
4585 expands into the list of hosts is now unlimited.
4587 47. The lsearch lookup could not handle data where the length of text line was
4588 more than 4095 characters. Such lines were truncated, leading to shortened
4589 data being returned. It should now handle lines of any length.
4591 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4592 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4595 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4596 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4597 no longer happens in comment lines. A list of replacements is now placed
4598 at the head of all of the source files, except those whose only change is
4599 to replace PERL_COMMAND in the very first #! line.
4601 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4602 messages on the queue, with a bottom-up merge sort, using code contributed
4603 by Michael Haardt. This should make operations like -bp somewhat faster on
4604 large queues. It won't affect queue runners, except when queue_run_in_order
4607 51. Installed eximstats 1.31 in the distribution.
4609 52. Added support for SRV lookups to the dnslookup router.
4611 53. If an ACL referred to $message_body or $message_body_end, the value was not
4612 reset for any messages that followed in the same SMTP session.
4614 54. The store-handling optimization for building very long strings was not
4615 differentiating between the different store pools. I don't think this
4616 actually made any difference in practice, but I've tidied it.
4618 55. While running the routers to verify a sender address, $sender_address
4619 was still set to the sender address. This is wrong, because when routing to
4620 send a bounce to the sender, it would be empty. Therefore, I have changed
4621 it so that, while verifying a sender address, $sender_address is set to <>.
4622 (There is no change to what happens when verifying a recipient address.)
4624 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4625 target A or AAAA records (if not already returned) without resetting the
4626 qualify_single or search_parents options of the DNS resolver. These are
4627 inappropriate in this case because the targets of MX and SRV records must
4628 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4629 target that, when qualified, matched something in the local domain. These
4630 two options are now turned off when doing these lookups.
4632 57. It seems that at least some releases of Reiserfs (which does not have the
4633 concept of a fixed number of inodes) returns zero and not -1 for the
4634 number of available inodes. This interacted badly with check_spool_inodes,
4635 which assumed that -1 was the "no such thing" setting. What I have done is
4636 to check that the total number of inodes is greater than zero before doing
4637 the test of how many are available.
4639 58. When a "warn" ACL statement has a log_message modifier, the message is
4640 remembered, and not repeated. This is to avoid a lot of repetition when a
4641 message has many recipients that cause the same warning to be written.
4642 Howewer, Exim was preserving the list of already written lines for an
4643 entire SMTP session, which doesn't seem right. The memory is now reset if a
4644 new message is started.
4646 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4647 debugging output unless log_rewrite was also set.
4649 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4650 of (int)(handle) when we know that handle contains (void *)(-1).
4652 61. The Exim daemon panic-logs an error return when it closes the incoming
4653 connection. However "connection reset by peer" seems to be common, and
4654 isn't really an error worthy of noting specially, so that particular error
4657 62. When Exim is trying to find all the local interfaces, it used to panic and
4658 die if the ioctl to get the interface flags failed. However, it seems that
4659 on at least one OS (Solaris 9) it is possible to have an interface that is
4660 included in the list of interfaces, but for which you get a failure error
4661 for this call. This happens when the interface is not "plumbed" into a
4662 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4663 failure of the "get flags" call assumes that the interface is down.
4665 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4666 makes life easier for people who are doing arithmetic on fields extracted
4667 from dates, where you often get leading zeros that should not be
4668 interpreted as octal.
4670 64. Added qualify_domain to the redirect router, to override the global
4673 65. If a pathologically long header line contained very many addresses (the
4674 report of this problem mentioned 10 000) and each of them was rewritten,
4675 Exim could use up a very large amount of memory. (It kept on making new
4676 copies of the header line as it rewrote, and never released the old ones.)
4677 At the expense of a bit more processing, the header rewriting function has
4678 been changed so that it no longer eats memory in this way.
4680 66. The generation of the Received: header has been moved from the time that a
4681 message starts to be received, to the time that it finishes. The timestamp
4682 in the Received: header should now be very close to that of the <= log
4683 line. There are two side-effects of this change:
4685 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4686 logged header lines no longer include the local Received: line, because
4687 it has not yet been created. The same applies to a copy of the message
4688 that is returned to a non-SMTP sender when a message is rejected.
4690 (b) When a filter file is tested using -bf, no additional Received: header
4691 is added to the test message. After some thought, I decided that this
4694 This change does not affect the value of $received_for. It is still set
4695 after address rewriting, but before local_scan() is called.
4697 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4699 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4700 gave an unhelpful panic error message, and a defer error. I have managed to
4701 change this behaviour so that it now rejects any supplied certificate,
4702 which seems right, as the list of acceptable certificates is empty.
4704 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4705 gave an unhelpful defer error. I have not managed to make this reject any
4706 supplied certificates, but the error message it gives is "no certificate
4707 supplied", which is not helpful.
4709 70. exigrep's output now also includes lines that are not associated with any
4710 message, but which match the given pattern. Implemented by a patch from
4711 Martin Sluka, which also tidied up the Perl a bit.
4713 71. Recipient callout verification, like sender verification, was using <> in
4714 the MAIL FROM command. This isn't really the right thing, since the actual
4715 sender may affect whether the remote host accepts the recipient or not. I
4716 have changed it to use the actual sender in the callout; this means that
4717 the cache record is now keyed on a recipient/sender pair, not just the
4718 recipient address. There doesn't seem to be a real danger of callout loops,
4719 since a callout by the remote host to check the sender would use <>.
4720 [SEE ABOVE: changed after hitting problems.]
4722 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4723 temporary errors. However, in the case of such a code being given after
4724 the end of a data transmission (i.e. after ".") Exim was failing to write
4725 a retry record for the message. (Yes, there was some broken host that was
4726 actually sending 8xx at this point.)
4728 73. An unknown lookup type in a host list could cause Exim to panic-die when
4729 the list was checked. (An example that provoked this was putting <; in the
4730 middle of a list instead of at the start.) If this happened during a DATA
4731 ACL check, a -D file could be left lying around. This kind of configuration
4732 error no longer causes Exim to die; instead it causes a defer errror. The
4733 incident is still logged to the main and panic logs.
4735 74. Buglet left over from Exim 3 conversion. The message "too many messages
4736 in one connection" was written to the rejectlog but not the mainlog, except
4737 when address rewriting (yes!) was being logged.
4739 75. Added write_rejectlog option.
4741 76. When a system filter was run not as root (that is, when system_filter_user
4742 was set), the values of the $n variables were not being returned to the
4743 main process; thus, they were not subsequently available in the $sn
4746 77. Added +return_path_on_delivery log selector.
4748 78. A connection timeout was being treated differently from recipients deferred
4749 when testing hosts_max_try with a message that was older than the host's
4750 retry timeout. (The host should not be counted, thus allowing all hosts to
4751 be tried at least once before bouncing.) This may have been the cause of an
4752 occasionally reported bug whereby a message would remain on the queue
4753 longer than the retry timeout, but would be bounced if a delivery was
4754 forced. I say "may" because I never totally pinned down the problem;
4755 setting up timeout/retry tests is difficult. See also the next item.
4757 79. The ultimate address timeout was not being applied to errors that involved
4758 a combination of host plus message (for example, a timeout on a MAIL
4759 command). When an address resolved to a number of possible hosts, and they
4760 were not all tried for each delivery (e.g. because of hosts_max_try), a
4761 message could remain on the queue longer than the retry timeout.
4763 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4766 81. Fixed an obscure SMTP outgoing bug which required at least the following
4767 conditions: (a) there was another message waiting for the same server;
4768 (b) the server returned 5xx to all RCPT commands in the first message so
4769 that the message was not completed; (c) the server dropped the connection
4770 or gave a negative response to the RSET that Exim sends to abort the
4771 transaction. The observed case was a dropped connection after DATA that had
4772 been sent in pipelining mode. That is, the server had advertised PIPELINING
4773 but was not implementing it correctly. The effect of the bug was incorrect
4774 behaviour, such as trying another host, and this could lead to a crash.
4780 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4781 and daemon.c were passed as pointers to ints; they should have been
4782 pointers to socklen_t variables (which are typically unsigned ints).
4784 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4787 3. Fixed a really odd bug that affected only the testing scheme; patching a
4788 certain fixed string in the binary changed the value of another string that
4789 happened to be identical to the end of the original first string.
4791 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4792 name", it returns that address as the IP address. On some operating
4793 systems (e.g. Solaris), it also passes back the IP address string as the
4794 "host name". However, on others (e.g. Linux), it passes back an empty
4795 string. Exim wasn't checking for this, and was changing the host name to an
4796 empty string, assuming it had been canonicized.
4798 5. Although rare, it is permitted to have more than one PTR record for a given
4799 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4800 all the names associated with an address, because they do in Solaris.
4801 However, it seems that they do not in Linux for data that comes from the
4802 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4803 I found this out when I moved to a new Linux workstation and tried to run
4804 the Exim test suite.
4806 To get round this problem I have changed the code so that it now does its
4807 own call to the DNS to look up PTR records when searching for a host name.
4808 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4809 addresses that are only in /etc/hosts are still found.
4811 This behaviour is, however, controlled by an option called host_lookup_
4812 order, which defaults to "bydns:byaddr". If people want to use the other
4813 order, or indeed, just use one or the other means of lookup, they can
4814 specify it in this variable.
4816 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4817 some operating systems, this comes back from gethostbyaddr() as an empty
4818 string, and this is what Exim used to test for. However, it seems that in
4819 other systems, "." is yielded. Exim now tests for this case too.
4821 7. The values of check_spool_space and check_log_space are now held internally
4822 as a number of kilobytes instead of an absolute number of bytes. If a
4823 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4824 kilobyte. This means that much larger values can be stored.
4826 8. Exim monitor: an attempt to get the action menu when not actually pointing
4827 at a message produces an empty menu entitled "No message selected". This
4828 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4829 no entries in it ("Shell widget menu has zero width and/or height"). So I
4830 have added a single, blank menu entry in this case.
4832 9. Added ${quote_local_part.
4834 10. MIME decoding is now applied to the contents of Subject: header lines when
4837 11. Now that a reference to $sender_host_address automatically causes a reverse
4838 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4839 host lookup before query-style lookups in lists that might use this
4840 variable. This has therefore been abolished, and the "net-" prefix is no
4841 longer necessary for query-style lookups.
4843 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4844 have been a typo for LFLAGS, so it has been changed.
4846 13. The install script calls Exim with "-C /dev/null" in order to find the
4847 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4848 to be output. Howeve, since Exim outputs its version number before the
4849 error, it didn't break the script. It just looked ugly. I fixed this by
4850 always allowing "-C /dev/null" if the caller is root.
4852 14. Ignore overlarge ACL variable number when reading spool file - insurance
4853 against a later release with more variables having written the file.
4855 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4856 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4857 this, as well as the form without the "IPv6" on the front.
4859 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4860 OS/Makefile-Darwin file.
4862 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4863 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4865 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4866 variable. However, some OS use ints for inodes. I've added cast to long int
4867 to get rid of the compiler warning.
4869 19. I had forgotten to lock out "/../" in configuration file names when
4870 ALT_CONFIG_PREFIX was set.
4872 20. Routers used for verification do not need to specify transports. However,
4873 if such a router generated a host list, and callout was configured, Exim
4874 crashed, because it could not find a port number from the (non-existent)
4875 transport. It now assumes port 25 in this circumstance.
4877 21. Added the -t option to exigrep.
4879 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4880 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4881 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4882 them is set, LOOKUP_LSEARCH is forced.
4884 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4886 24. Added sender and host information to the "rejected by local_scan()" log
4887 line; previously there was no indication of these.
4889 25. Added .include_if_exists.
4891 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4892 Linux. It turns out that not all file systems support this. Apparently some
4893 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4894 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4895 not supported on the file descriptor, is now ignored when Exim is trying to
4896 sync a directory. This applies only to Linux.
4898 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4900 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4901 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4904 29. Added support for maildirsize files from supplied patch (modified a bit).
4906 30. The use of :fail: followed by an empty string could lead Exim to respond to
4907 sender verification failures with (e.g.):
4909 550 Verification failed for <xxx>
4910 550 Sender verify failed
4912 where the first response line was missing the '-' that indicates it is not
4913 the final line of the response.
4915 31. The loop for finding the name of the user that called Exim had a hardwired
4916 limit of 10; it now uses the value of finduser_retries, which is used for
4917 all other user lookups.
4919 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4922 33. Exim was neglecting to zero errno before one call of strtol() when
4923 expanding a string and expecting an integer value. On some systems this
4924 resulted in spurious "integer overflow" errors. Also, it was casting the
4925 result into an int without checking.
4927 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4928 did not work. The code looks as if it has *never* worked, though it appears
4929 to have been documented since at least releast 1.62. I have made it work.
4931 35. The "timeout_DNS" error in retry rules, also documented since at least
4932 1.62, also never worked. As it isn't clear exactly what this means, and
4933 clearly it isn't a major issue, I have abolished the feature by treating it
4934 as "timeout", and writing a warning to the main and panic logs.
4936 36. The display of retry rules for -brt wasn't always showing the error code
4939 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4940 timeout_connect_A, timeout_connect_MX.
4942 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4943 to the empty sender.
4945 39. The daemon was not analysing the content of -oX till after it had closed
4946 stderr and disconnected from the controlling terminal. This meant that any
4947 syntax errors were only noted on the panic log, and the return code from
4948 the command was 0. By re-arranging the code a little, I've made the
4949 decoding happen first, so such errors now appear on stderr, and the return
4950 code is 1. However, the actual setting up of the sockets still happens in
4951 the disconnected process, so errors there are still only recorded on the
4954 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4955 connections (as happens on some IP stacks) was logged at start up time as
4956 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4957 it from "IPv6 and IPv4", which means that two separate sockets are being
4960 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4961 says whether AF_INET or AF_INET6 was passed as an argument.
4963 42. Exiwhat output was messed up when time zones were included in log
4966 43. Exiwhat now gives more information about the daemon's listening ports,
4967 and whether -tls-on-connect was used.
4969 44. The "port" option of the smtp transport is now expanded.
4971 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4972 silently ignored. Now an error message is written to the main and panic
4975 46. There's a new ACL modifier called "logwrite" which writes to a log file
4976 as soon as it is encountered.
4978 47. Added $local_user_uid and $local_user_gid at routing time.
4980 48. Exim crashed when trying to verify a sender address that was being
4983 49. Exim was recognizing only a space character after ".include". It now also
4984 recognizes a tab character.
4986 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4987 extracting the relevant information from the specification. The man page no
4988 longer contains scrambled data for the -d option, and I've added a section
4989 at the front about calling Exim under different names.
4991 51. Added "extra_headers" argument to the "mail" command in filter files.
4993 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4996 53. Installed eximstats 1.29.
4998 54. Added transport_filter_timeout as a generic transport option.
5000 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5001 Cc: header lines. This was required by RFC 822, but it not required by RFC
5004 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5005 incoming messages that did not have them. Now it does so only if the
5006 message originates locally, that is, if there is no associated remote host
5007 address. When Resent- header lines are present, this applies to the Resent-
5008 lines rather than the non-Resent- lines.
5010 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5011 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5013 58. Messages for configuration errors now include the name of the main
5014 configuration file - useful now that there may be more than one file in a
5015 list (.included file names were always shown).
5017 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5018 for those rare installations that do not start the daemon as root or run it
5019 setuid root. I've cut out the call to initgroups() if the daemon is not
5022 60. The Exim user and group can now be bound into the binary as text strings
5023 that are looked up at the start of Exim's processing.
5025 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5027 62. Added $mailstore_basename variable.
5029 63. Installed patch to sieve.c from Michael Haardt.
5031 64. When Exim failed to open the panic log after failing to open the main log,
5032 the original message it was trying to log was written to stderr and debug
5033 output, but if they were not available (the usual case in production), it
5034 was lost. Now it is written to syslog before the two lines that record the
5035 failures to open the logs.
5037 65. Users' Exim filters run in subprocesses under the user's uid. It is
5038 possible for a "deliver" command or an alias in a "personal" command to
5039 provoke an address rewrite. If logging of address rewriting is configured,
5040 this fails because the process is not running as root or exim. There may be
5041 a better way of dealing with this, but for the moment (because 4.30 needs
5042 to be released), I have disabled address rewrite logging when running a
5043 filter in a non-root, non-exim process.
5049 1. The buildconfig auxiliary program wasn't quoting the value set for
5050 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5051 not defined. This bug was masked in 4.22 by the effect that was fixed in
5054 2. Some messages that were rejected after a message id was allocated were
5055 shown as "incomplete" by exigrep. It no longer does this for messages that
5056 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5058 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5059 have allow_domain_literals set, the ID did not get logged in the <= line.
5060 Domain literals are now always recognized in Message-ID: header lines.
5062 4. The first argument for a ${extract expansion item is the key name or field
5063 number. Leading and trailing spaces in this item were not being ignored,
5064 causing some misleading effects.
5066 5. When deliver_drop_privilege was set, single queue runner processes started
5067 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5068 same command in the process it spins off) were not dropping privilege.
5070 6. When the daemon running as "exim" started a queue runner, it always
5071 re-executed Exim in the spun-off process. This is a waste of effort when
5072 deliver_drop_privilege is set. The new process now just calls the
5073 queue-runner function directly.
5079 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5082 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5083 ignored. Though the use of -oP was forcing the writing of a pid file, it
5084 was always written to the default place.
5086 3. If the message "no IP address found for host xxxx" is generated during
5087 incoming verification, it is now followed by identification of the incoming
5088 connection (so you can more easily find what provoked it).
5090 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5092 5. Added some features to "harden" Exim a bit more against certain attacks:
5094 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5095 be put in Local/Makefile. This is like the never_users runtime option,
5096 but it cannot be overridden. The default setting is "root".
5098 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5099 prefix string with which any file named in a -C command line option
5102 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5103 is retained for -C and -D only if the caller of Exim is root. Without
5104 it, the exim user may also use -C and -D and retain privilege.
5106 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5107 command line option is disabled.
5109 6. Macro names set by the -D option must start with an upper case letter, just
5110 like macro names defined in the configuration file.
5112 7. Added "dereference=" facility to LDAP.
5114 8. Two instances of the typo "uknown" in the source files are fixed.
5116 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5117 the Configure-Makefile script screwed up while processing it.
5119 10. Incorporated PCRE 4.4.
5121 11. The SMTP synchronization check was not operating right at the start of an
5122 SMTP session. For example, it could not catch a HELO sent before the client
5123 waited for the greeting. There is now a check for outstanding input at the
5124 point when the greeting is written. Because of the duplex, asynchronous
5125 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5126 way, but not yet received, when the check is performed.
5128 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5129 on TCP/IP sockets, because this apparently causes some broken clients to
5132 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5133 unchanged) from the Cygwin maintainer.
5135 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5136 was supported instead of testing for mbx. Effectively a typo.
5138 15. The spa authenticator server code was not checking that the input it
5139 received was valid base64.
5141 16. The debug output line for the "set" modifier in ACLs was not showing the
5142 name of the variable that was being set.
5144 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5146 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5147 lookup on its own. Something else was needed to trigger the lookup. For
5148 example, a match in host_lookup or the need for a host name in a host list.
5149 Now, if $sender_host_name is referenced and the host name has not yet been
5150 looked up, a lookup is performed. If the lookup fails, the variable remains
5151 empty, and $host_lookup_failed is set to "1".
5153 19. Added "eqi" as a case-independent comparison operator.
5155 20. The saslauthd authentication condition could segfault if neither service
5156 nor realm was specified.
5158 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5159 error message that was logged was misleading, and incoming SMTP
5160 connections were dropped. The message is now more accurate, and temporary
5161 errors are given to SMTP connections.
5163 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5164 (except RSET) with a 421 error, until QUIT is received. However, it was
5165 failing to send a response to QUIT.
5167 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5168 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5169 ACL is now run after the helo_try_verify_hosts code.
5171 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5172 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5173 case-independent, so other case variants are also recognized). Apparently
5174 some systems use these upper case variants.
5176 25. If more than two messages were waiting for the same host, and a transport
5177 filter was specified for the transport, Exim sent two messages over the
5178 same TCP/IP connection, and then failed with "socket operation on non-
5179 socket" when it tried to send the third.
5181 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5183 27. The extern definition of crypt16() in expand.c was not being excluded when
5184 the OS had its own crypt16() function.
5186 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5187 as a preferred synonym for return_size_limit, both as an option and as an
5190 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5192 30. Changed the default configuration ACL to relax the local part checking rule
5193 for addresses that are not in any local domains. For these addresses,
5194 slashes and pipe symbols are allowed within local parts, but the sequence
5195 /../ is explicitly forbidden.
5197 31. SPA server authentication was not clearing the challenge buffer before
5200 32. log_message in a "warn" ACL statement was writing to the reject log as
5201 well as to the main log, which contradicts the documentation and doesn't
5202 seem right (because no rejection is happening). So I have stopped it.
5204 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5205 However, I am unable to do any testing of this.
5207 34. Fixed an infelicity in the appendfile transport. When checking directories
5208 for a mailbox, to see if any needed to be created, it was accidentally
5209 using path names with one or more superfluous leading slashes; tracing
5210 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5212 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5213 contents are added to the log line that is written for every discarded
5214 recipient. (Previously a log_message setting was ignored.)
5216 36. The ${quote: operator now quotes the string if it is empty.
5218 37. The install script runs exim in order to find its version number. If for
5219 some reason other than non-existence or emptiness, which it checks, it
5220 could not run './exim', it was installing it with an empty version number,
5221 i.e. as "exim-". This error state is now caught, and the installation is
5224 38. An argument was missing from the function that creates an error message
5225 when Exim fails to connect to the socket for saslauthd authentication.
5226 This could cause Exim to crash, or give a corrupted message.
5228 39. Added isip, isip4, and isip6 to ${if conditions.
5230 40. The ACL variables $acl_xx are now saved with the message, and can be
5231 accessed later in routers, transports, and filters.
5233 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5234 strings in the file are not string-expanded.
5236 42. If a MAIL command specified a SIZE value that was too large to fit into an
5237 int variable, the check against message_size_limit failed. Such values are
5238 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5239 day this will have to be increased, but I don't think I want to be around
5240 when emails are that large.
5247 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5248 iconv() is not standard in FreeBSD.
5250 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5251 IPv6 enabled. The observed symptom was a segmentation fault on return from
5252 the function os_common_find_running_interfaces() in src/os.c.
5254 3. In the check_special_case() function in daemon.c I had used "errno" as an
5255 argument name, which causes warnings on some systems. This was basically a
5256 typo, since it was named "eno" in the comments!
5258 4. The code that waits for the clock to tick (at a resolution of some fraction
5259 of a second) so as to ensure message-id uniqueness was always waiting for
5260 at least one whole tick, when it could have waited for less. [This is
5261 almost certainly not relevant at current processor speeds, where it is
5262 unlikely to ever wait at all. But we try to future-proof.]
5264 5. The function that sleeps for a time interval that includes fractions of a
5265 second contained a race. It did not block SIGALRM between setting the
5266 timer, and suspending (a couple of lines later). If the interval was short
5267 and the sigsuspend() was delayed until after it had expired, the suspension
5268 never ended. On busy systems this could lead to processes getting stuck for
5271 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5272 process, before it forks any delivery processes. The open lookup caching
5273 mechanism meant that the open file or database connection was passed into
5274 the delivery process. The problem was that delivery processes always tidy
5275 up cached lookup data. This could cause a problem for the next delivery
5276 process started by the queue runner, because the external queue runner
5277 process does not know about the closure. So the next delivery process
5278 still has data in the lookup cache. In the case of a file lookup, there was
5279 no problem because closing a file descriptor in a subprocess doesn't affect
5280 the parent. However, if the lookup was caching a connection to a database,
5281 the connection was closed, and the second delivery process was likely to
5282 see errors such as "PGSQL: query failed: server closed the connection
5283 unexpectedly". The problem has been fixed by closing all cached lookups
5284 in a queue runner before running a delivery process.
5286 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5287 seem to have the "const" qualifier which it has on other OS. I've
5290 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5291 *of the same type* (client or server) with the same public name that an
5292 error should be diagnosed.
5294 9. When Exim looked up a host name for an IP address, but failed to find the
5295 original IP address when looking up the host name (a safety check), it
5296 output the message "<ip address> does not match any IP for NULL", which was
5297 confusing, to say the least. The bug was that the host name should have
5298 appeared instead of "NULL".
5300 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5301 user that is built into the binary, and the -C or -D options is used, root
5302 privilege is dropped before the configuration file is read. In addition,
5303 logging is switched to stderr instead of the normal log files. If the
5304 configuration then re-defines the Exim user, the unprivileged environment
5305 is probably not what is expected, so Exim logs a panic warning message (but
5308 However, if deliver_drop_privilege is set, the unprivileged state may well
5309 be exactly what is intended, so the warning has been cut out in that case,
5310 and Exim is allowed to try to write to its normal log files.
5316 1. smtp_return_error_details was not giving details for temporary sender
5317 or receiver verification errors.
5319 2. Diagnose a configuration error if two authenticators have the same public
5322 3. Exim used not to create the message log file for a message until the first
5323 delivery attempt. This could be confusing when incoming messages were held
5324 for policy or load reasons. The message log file is now created at the time
5325 the message is received, and an initial "Received" line is written to it.
5327 4. The automatically generated man page for command line options had a minor
5328 bug that caused no ill effects; however, a more serious problem was that
5329 the procedure for building the man page automatically didn't always
5330 operate. Consequently, release 4.20 contains an out-of-date version. This
5331 shouldn't happen again.
5333 5. When building Exim with embedded Perl support, the script that builds the
5334 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5335 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5337 6. The freeze_tell option was not being used for messages that were frozen on
5338 arrival, either by an ACL or by local_scan().
5340 7. Added the smtp_incomplete_transaction log selector.
5342 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5343 was accepting AUTH without a new EHLO.
5345 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5346 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5349 10. Logging of TCP/IP connections (when configured) now happens in the main
5350 daemon process instead of the child process, so that the TCP/IP connection
5351 count is more accurate (but it can never be perfect).
5353 11. The use of "drop" in a nested ACL was not being handled correctly in the
5354 outer ACL. Now, if condition failure induced by the nested "drop" causes
5355 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5356 or "require"), the connection is dropped.
5358 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5359 that yield "discard" can now be used with an "accept" or a "discard" verb,
5360 but an error is generated for any others (because I can't see a useful way
5361 to define what should happen).
5363 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5364 are now processed in the same way as lines in the Exim configuration file.
5365 In particular, continuation lines are supported.
5367 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5369 15. Added -ti meaning -t -i.
5371 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5372 domains, and warn by logging if others are found.
5374 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5375 returned from GIFCONF when Exim is trying to find the list of interfaces on
5376 a host. The code in os.c has been modified to copy each ifreq to an aligned
5377 structure in all cases.
5379 Also, in some cases, the returned ifreq's were being copied to a 'struct
5380 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5381 means the last couple of bytes of an IPv6 address could be chopped if the
5382 ifreq contained only a normal sockaddr (14 bytes storage).
5384 18. Named domain lists were not supported in the hosts_treat_as_local option.
5385 An entry such as +xxxx was not recognized, and was treated as a literal
5388 19. Ensure that header lines added by a DATA ACL are included in the reject log
5389 if the ACL subsequently rejects the message.
5391 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5392 MD5 (which is deprecated).
5394 21. When testing a filter file using -bf, Exim was writing a message when it
5395 took the sender from a "From " line in the message, but it was not doing so
5396 when it took $return_path from a Return-Path: header line. It now does.
5398 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5399 with a valid header line field name (a series of printing characters
5400 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5402 23. Changed "disc" in the source to "disk" to conform to the documentation and
5403 the book and for uniformity.
5405 24. Ignore Sendmail's -Ooption=value command line item.
5407 25. When execve() failed while trying to run a command in a pipe transport,
5408 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5409 could be confused with a return value of 69 from the command itself. This
5410 has been changed to 127, the value the shell returns if it is asked to run
5411 a non-existent command. The wording for the related log line suggests a
5412 non-existent command as the problem.
5414 26. If received_header_text expands to an empty string, do not add a Received:
5415 header line to the message. (Well, it adds a token one on the spool, but
5416 marks it "old" so that it doesn't get used or transmitted.)
5418 27. Installed eximstats 1.28 (addition of -nt option).
5420 28. There was no check for failure on the call to getsockname() in the daemon
5421 code. This can fail if there is a shortage of resources on the system, with
5422 ENOMEM, for example. A temporary error is now given on failure.
5424 29. Contrary to the C standard, it seems that in some environments, the
5425 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5426 program. Exim now does this explicitly; it affects the formatting of
5427 timestamps using strftime().
5429 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5430 complaints. I've now initialized all the variables, to avoid this.
5432 32. Header lines added by a system filter were not being "seen" during
5433 transport-time rewrites.
5435 33. The info_callback() function passed to OpenSSL is set up with type void
5436 (*)(SSL *, int, int), as described somewhere. However, when calling the
5437 function (actually a macro) that sets it up, the type void(*)() is
5438 expected. I've put in a cast to prevent warnings from picky compilers.
5440 34. If a DNS black list lookup found a CNAME record, but there were no A
5441 records associated with the domain it pointed at, Exim crashed.
5443 35. If a DNS black list lookup returned more than one A record, Exim ignored
5444 all but the first. It now scans all returned addresses if a particular IP
5445 value is being sought. In this situation, the contents of the
5446 $dnslist_value variable are a list of all the addresses, separated by a
5449 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5450 to accept a host name and all its aliases if the forward lookup for any of
5451 them yielded the IP address of the incoming connection. Now it accepts only
5452 those names whose forward lookup yields the correct IP address. Any other
5453 names are discarded. This closes a loophole whereby a rogue DNS
5454 administrator could create reverse DNS records to break through a
5455 wildcarded host restriction in an ACL.
5457 37. If a user filter or a system filter that ran in a subprocess used any of
5458 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5459 the wrong values were passed to the pipe command ($thisaddress had the
5460 value of $0, $0 had the value of $1, etc). This bug was introduced by
5461 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5463 38. Improved the line breaking for long SMTP error messages from ACLs.
5464 Previously, if there was no break point between 40 and 75 characters, Exim
5465 left the rest of the message alone. Two changes have been made: (a) I've
5466 reduced the minimum length to 35 characters; (b) if it can't find a break
5467 point between 35 and 75 characters, it looks ahead and uses the first one
5468 that it finds. This may give the occasional overlong line, but at least the
5469 remaining text gets split now.
5471 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5472 file descriptors might be low, and that setting 1000 would always raise it.
5473 It turns out that in some environments, the limit is already over 1000 and
5474 that lowering it causes trouble. So now Exim takes care not to decrease it.
5476 40. When delivering a message, the value of $return_path is set to $sender_
5477 address at the start of routing (routers may change the value). By an
5478 oversight, this default was not being set up when an address was tested by
5479 -bt or -bv, which affected the outcome if any router or filter referred to
5482 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5483 log only when "message" or "log_message" are set. However, if one of the
5484 conditions was an address verification, or a call to a nested ACL, the
5485 messages generated by the underlying test were being passed through. This
5486 no longer happens. The underlying message is available in $acl_verify_
5487 message for both "message" and "log_message" expansions, so it can be
5488 passed through if needed.
5490 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5491 new expansion $bh_ to give the encoded byte string without charset
5492 translation. Translation happens only if iconv() is available; HAVE_ICONV
5493 indicates this at build time. HEADERS_CHARSET gives the charset to
5494 translate to; headers_charset can change it in the configuration, and
5495 "headers charset" can change it in an individual filter file.
5497 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5498 that creates RFC 2047 encoded "words" labels them as that charset instead
5499 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5500 expansion operator; (ii) when Exim creates a From: line for a local
5501 message; (iii) when a header line is rewritten to include a "phrase" part.
5503 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5504 buggy, causing it to skip the first lines of messages whose message ID
5505 ended in 'D'. This would not have bitten before Exim release 4.14, because
5506 message IDs were unlikely to end in 'D' before then. The effect was to have
5507 incorrect size information for certain domains.
5509 45. #include "config.h" was missing at the start of the crypt16.c module. This
5510 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5513 46. If there was a timeout during a "random" callout check, Exim treated it as
5514 a failure of the random address, and carried on sending RSET and the real
5515 address. If the delay was just some slowness somewhere, the response to the
5516 original RCPT would be taken as a response to RSET and so on, causing
5517 mayhem of various kinds.
5519 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5520 when I implemented it. It didn't allow for the fact that some option values
5521 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5522 the right test for positive values.
5524 48. Domain names in DNS records are case-independent. Exim always looks them up
5525 in lower case. Some resolvers return domain names in exactly the case they
5526 appear in the zone file, that is, they may contain uppercase letters. Not
5527 all resolvers do this - some return always lower case. Exim was treating a
5528 change of case by a resolver as a change of domain, similar to a widening
5529 of a domain abbreviation. This triggered its re-routing code and so it was
5530 trying to route what was effectively the same domain again. This normally
5531 caused routing to fail (because the router wouldn't handle the domain
5532 twice). Now Exim checks for this case specially, and just changes the
5533 casing of the domain that it ultimately uses when it transmits the message
5536 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5539 50. If a filter generated a file delivery with a non-absolute name (possible if
5540 no home directory exists for the router), the forbid_file option was not
5543 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5544 the existing equality matching.
5546 52. Exim was using ints instead of ino_t variables in some places where it was
5547 dealing with inode numbers.
5549 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5550 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5551 called TMPDIR, and if it finds it is different, it changes its value.
5553 54. The smtp_printf() function is now made available to local_scan() so
5554 additional output lines can be written before returning. There is also an
5555 smtp_fflush() function to enable the detection of a dropped connection.
5556 The variables smtp_input and smtp_batched_input are exported to
5559 55. Changed the default runtime configuration: the message "Unknown user"
5560 has been removed from the ACL, and instead placed on the localuser router,
5561 using the cannot_route_message feature. This means that any verification
5562 failures that generate their own messages won't get overridden. Similarly,
5563 the "Unrouteable address" message that was in the ACL for unverifiable
5564 relay addresses has also been removed.
5566 56. Added hosts_avoid_esmtp to the smtp transport.
5568 57. The exicyclog script was not checking for the esoteric option
5569 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5570 will work only if exicyclog is run under the appropriate euid.
5572 58. Following a discussion on the list, the rules by which Exim recognises line
5573 endings on incoming messages have been changed. The -dropcr and drop_cr
5574 options are now no-ops, retained only for backwards compatibility. The
5575 following line terminators are recognized: LF CRLF CR. However, special
5576 processing applies to CR:
5578 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5579 nor a local message in the state where . is a terminator.
5581 (ii) If a bare CR is encountered in a header line, an extra space is added
5582 after the line terminator so as not to end the header. The reasoning
5583 behind this is that bare CRs in header lines are most likely either
5584 to be mistakes, or people trying to play silly games.
5586 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5587 was being incorrectly given as 18 bytes larger than it should have been.
5588 This is a VOB (very old bug).
5590 60. This may never have affected anything current, but just in case it has:
5591 When the local host is found other than at the start of a list of hosts,
5592 the local host, those with the same MX, and any that follow, are discarded.
5593 When the list in question was part of a longer list of hosts, the following
5594 hosts (not currently being processed) were also being discarded. This no
5595 longer happens. I'm not sure if this situation could ever has previously
5598 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5601 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5602 header. This is something that is recommended in a new Internet Draft, and
5603 is something that is documented as being done by Sendmail. There are two
5604 possible values. For messages generated by the autoreply transport, Exim
5607 Auto-Submitted: auto-replied
5609 whereas for all other generated messages (e.g. bounces) it adds
5611 Auto-Submitted: auto-generated
5613 63. The "personal" condition in filters now includes a test for the
5614 Auto-Submitted: header. If it contains the string "auto-" the message it
5615 not considered personal.
5617 64. Added rcpt_include_affixes as a generic transport option.
5619 65. Added queue_only_override (default true).
5621 66. Added the syslog_duplication option.
5623 67. If what should have been the first header line of a message consisted of
5624 a space followed by a colon, Exim was mis-interpreting it as a header line.
5625 It isn't of course - it is syntactically invalid and should therefore be
5626 treated as the start of the message body. The misbehaviour could have
5627 caused a number of strange effects, including loss of data in subsequent
5628 header lines, and spool format errors.
5630 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5631 client host had authenticated. This control can now be exercised by an ACL
5632 for more flexibility.
5634 69. By default, callouts do not happen when testing with -bh. There is now a
5635 variant, -bhc, which does actually run the callout code, including
5636 consulting and updating the callout cache.
5638 70. Added support for saslauthd authentication, courtesy of Alexander
5641 71. If statvfs() failed on the spool or log directories while checking their
5642 size for availability, Exim confusingly gave the error "space shortage".
5643 Furthermore, in debugging mode it crashed with a floating point exception.
5644 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5645 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5646 serious problem, Exim now writes to the main and panic logs when this
5647 happens, with details of the failure. It then refuses to accept the
5648 incoming message, giving the message "spool directory problem" or "log
5649 directory problem" with a 421 code for SMTP messages.
5651 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5652 0, 1, and 2 exist, because some OS complain for execs without them (see
5653 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5654 descriptors. However, the code omitted to check that the open succeeded,
5655 causing mysterious errors if for some reason the permissions on /dev/null
5656 got screwed. Now Exim writes a message to the main and panic logs, and
5657 bombs out if it can't open /dev/null.
5659 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5660 interact so that it is all more flexible. It is supposed to remain
5661 backwards compatible. Also added extra_local_interfaces.
5663 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5664 to bomb out with an assertion failure - to the client this appears as a
5665 connection drop. This problem occurs in the part of the code that was taken
5666 from the Samba project. Fortunately, the assertion is in a very simple
5667 function, so I have fixed this by reproducing the function inline in the
5668 one place where it is called, and arranging for authentication to fail
5669 instead of killing the process with assert().
5671 75. The SPA client code was not working when the server requested OEM rather
5672 than Unicode encoding.
5674 76. Added code to make require_files with a specific uid setting more usable in
5675 the case where statting the file as root fails - usually a non-root-mounted
5676 NFS file system. When this happens and the failure is EACCES, Exim now
5677 forks a subprocess and does the per-uid checking as the relevant uid.
5679 77. Added process_log_path.
5681 78. If log_file_path was not explicitly set, a setting of check_log_space or
5682 check_log_inodes was ignored.
5684 79. If a space check for the spool or log partitions fails, the incident is now
5685 logged. Of course, in the latter case the data may get lost...
5687 80. Added the %p formatting code to string_format() so that it can be used to
5688 print addresses in debug_print(). Adjusted all the address printing in the
5689 debugging in store.c to use %p rather than %d.
5691 81. There was a concern that a line of code in smtp_in.c could overflow a
5692 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5693 initially expressed, the concern was not well-founded, because trailing
5694 spaces are removed early. However, if the trailing spaces were followed by
5695 a NULL, they did not get removed, so the overflow was possible. Two fixes
5698 (a) I re-wrote the offending code in a cleaner fashion.
5699 (b) If an incoming SMTP command contains a NULL character, it is rejected
5702 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5703 runs initgroups(), so that if the Exim user is in any additional groups,
5704 they will be used during message reception.
5710 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob