1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
9 PP/02 Make -n do something, by making it not do something.
10 When combined with -bP, the name of an option is not output.
12 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
15 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
16 $sender_host_name and config options to manage this, and basic check
19 PP/05 DSCP support for outbound connections and control modifier for inbound.
21 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
22 (Only plugin which currently uses this is kerberos4, which nobody should
23 be using, but we should make it available and other future plugins might
24 conceivably use it, even though it would break NAT; stuff *should* be
25 using channel bindings instead).
27 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
28 name; added for Sendmail compatibility; requires admin caller.
29 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
30 just ignore it); requires trusted caller.
31 Also parse but ignore: -Ac -Am -X<logfile>
34 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
36 TL/02 Add +smtp_confirmation as a default logging option.
38 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
39 Patch by Magnus Holmgren from 2007-02-20.
41 TL/04 Bugzilla 1281 - Spec typo.
42 Bugzilla 1283 - Spec typo.
43 Bugzilla 1290 - Spec grammar fixes.
45 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
47 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
49 JH/02 Support "G" suffix to numbers in ${if comparisons.
51 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
53 NM/01 Bugzilla 1197 - Spec typo
54 Bugzilla 1196 - Spec examples corrections
56 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
58 PP/09 Add gnutls_enable_pkcs11 option.
60 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
61 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
63 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
64 "acl {{name}{arg}...}", and optional args on acl condition
67 JH/05 Permit multiple router/transport headers_add/remove lines.
69 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
71 JH/07 Avoid using a waiting database for a single-message-only transport.
72 Performance patch from Paul Fisher. Bugzilla 1262.
74 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
81 PP/01 Handle short writes when writing local log-files.
82 In practice, only affects FreeBSD (8 onwards).
83 Bugzilla 1053, with thanks to Dmitry Isaikin.
85 NM/01 Bugzilla 949 - Documentation tweak
87 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
90 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
92 PP/02 Implemented gsasl authenticator.
94 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
96 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
97 `pkg-config foo` for cflags/libs.
99 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
100 with rest of GSASL and with heimdal_gssapi.
102 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
103 `pkg-config foo` for cflags/libs for the TLS implementation.
105 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
106 properties get this fed in as external SSF. A number of robustness
107 and debugging improvements to the cyrus_sasl authenticator.
109 PP/08 cyrus_sasl server now expands the server_realm option.
111 PP/09 Bugzilla 1214 - Log authentication information in reject log.
112 Patch by Jeremy Harris.
114 PP/10 Added dbmjz lookup type.
116 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
118 PP/12 MAIL args handles TAB as well as SP, for better interop with
119 non-compliant senders.
120 Analysis and variant patch by Todd Lyons.
122 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
123 Bug report from Lars Müller <lars@samba.org> (via SUSE),
124 Patch from Dirk Mueller <dmueller@suse.com>
126 PP/13 tls_peerdn now print-escaped for spool files.
127 Observed some $tls_peerdn in wild which contained \n, which resulted
128 in spool file corruption.
130 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
131 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
132 or write after TLS renegotiation, which otherwise led to messages
135 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
136 as a tracking header (ie: a signed header comes before the signature).
137 Patch from Wolfgang Breyha.
139 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
140 comma-sep list; embedded commas doubled.
142 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
144 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
146 Report and patch from Dmitry Banschikov.
148 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
149 Removed SSL_clear() after SSL_new() which led to protocol negotiation
150 failures. We appear to now support TLS1.1+ with Exim.
152 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
153 lets Exim select keys and certificates based upon TLS SNI from client.
154 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
155 before an outbound SMTP session. New log_selector, +tls_sni.
157 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
158 NULL dereference. Report and patch from Alun Jones.
160 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
161 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
162 Not seeing resolver debug output on NetBSD, but suspect this is a
163 resolver implementation change.
165 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
166 Left warnings. Added "eximon gdb" invocation mode.
168 PP/21 Defaulting "accept_8bitmime" to true, not false.
170 PP/22 Added -bw for inetd wait mode support.
172 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
173 locate the relevant includes and libraries. Made this the default.
175 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
176 Bugzilla 1246, report and most of solution from Tomasz Kusy.
178 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
179 This may cause build issues on older platforms.
181 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
182 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
183 gnutls_require_mac & gnutls_require_protocols (no longer supported).
184 Added SNI support via GnuTLS too.
185 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
187 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
189 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
190 Applied second patch from Janne, implementing suggestion to default
191 multiple-strings-in-record handling to match SPF spec.
193 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
195 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
196 read-only, out of scope).
197 Patch from Wolfgang Breyha, report from Stuart Northfield.
199 PP/29 Fix three issues highlighted by clang analyser static analysis.
200 Only crash-plausible issue would require the Cambridge-specific
201 iplookup router and a misconfiguration.
202 Report from Marcin Mirosław.
204 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
206 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
207 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
208 As part of this, removing so much warning spew let me fix some minor
209 real issues in debug logging.
211 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
212 assignment on my part. Fixed.
214 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
215 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
216 Janne Snabb (who went above and beyond: thank you).
218 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
219 string otherwise requires a connection and a bunch more work and it's
220 relatively easy to get wrong. Should also expose TLS library linkage
223 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
224 64-bit ${eval} (JH/03).
226 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
227 GNU libc to support some of the 64-bit stuff, should not lead to
228 conflicts. Defined before os.h is pulled in, so if a given platform
229 needs to override this, it can.
231 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
232 protection layer was required, which is not implemented.
233 Bugzilla 1254, patch from Wolfgang Breyha.
235 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
236 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
237 tls_dhparam take prime identifiers. Also unbreak combination of
238 OpenSSL+DH_params+TLSSNI.
240 PP/39 Disable SSLv2 by default in OpenSSL support.
246 PP/01 Solaris build fix for Oracle's LDAP libraries.
247 Bugzilla 1109, patch from Stephen Usher.
249 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
251 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
254 TF/02 Fix a couple more cases where we did not log the error message
255 when unlink() failed. See also change 4.74-TF/03.
257 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
258 lock up or crash if it happened to be inside a call to libc when it
259 got a SIGUSR1 from exiwhat.
261 The SIGUSR1 handler appends the current process status to the process
262 log which is later printed by exiwhat. It used to use the general
263 purpose logging code to do this, but several functions it calls are
264 not safe for signals.
266 The new output code in the SIGUSR1 handler is specific to the process
267 log, and simple enough that it's easy to inspect for signal safety.
268 Removing some special cases also simplifies the general logging code.
269 Removing the spurious timestamps from the process log simplifies
272 TF/04 Improved ratelimit ACL condition.
274 The /noupdate option has been deprecated in favour of /readonly which
275 has clearer semantics. The /leaky, /strict, and /readonly update modes
276 are mutually exclusive. The update mode is no longer included in the
277 database key; it just determines when the database is updated. (This
278 means that when you upgrde Exim will forget old rate measurements.)
280 Exim now checks that the per_* options are used with an update mode that
281 makes sense for the current ACL. For example, when Exim is processing a
282 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
283 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
284 must specify per_mail/readonly. If you omit the update mode it defaults to
285 /leaky where that makes sense (as before) or /readonly where required.
287 The /noupdate option is now undocumented but still supported for
288 backwards compatibility. It is equivalent to /readonly except that in
289 ACLs where /readonly is required you may specify /leaky/noupdate or
290 /strict/noupdate which are treated the same as /readonly.
292 A useful new feature is the /count= option. This is a generalization
293 of the per_byte option, so that you can measure the throughput of other
294 aggregate values. For example, the per_byte option is now equivalent
295 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
297 The per_rcpt option has been generalized using the /count= mechanism
298 (though it's more complicated than the per_byte equivalence). When it is
299 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
300 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
301 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
302 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
303 Note that using per_rcpt with a non-readonly update mode in more than
304 one ACL will cause the recipients to be double-counted. (The per_mail
305 and per_byte options don't have this problem.)
307 The handling of very low rates has changed slightly. If the computed rate
308 is less than the event's count (usually one) then this event is the first
309 after a long gap. In this case the rate is set to the same as this event's
310 count, so that the first message of a spam run is counted properly.
312 The major new feature is a mechanism for counting the rate of unique
313 events. The new per_addr option counts the number of different
314 recipients that someone has sent messages to in the last time period. It
315 behaves like per_rcpt if all the recipient addresses are different, but
316 duplicate recipient addresses do not increase the measured rate. Like
317 the /count= option this is a general mechanism, so the per_addr option
318 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
319 example, measure the rate that a client uses different sender addresses
320 with the options per_mail/unique=$sender_address. There are further
321 details in the main documentation.
323 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
325 TF/06 Removed a few PCRE remnants.
327 TF/07 Automatically extract Exim's version number from tags in the git
328 repository when doing development or release builds.
330 PP/02 Raise smtp_cmd_buffer_size to 16kB.
331 Bugzilla 879. Patch from Paul Fisher.
333 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
334 Heavily based on revision 40f9a89a from Simon Arlott's tree.
337 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
339 PP/05 Variable $av_failed, true if the AV scanner deferred.
340 Bugzilla 1078. Patch from John Horne.
342 PP/06 Stop make process more reliably on build failure.
343 Bugzilla 1087. Patch from Heiko Schlittermann.
345 PP/07 Make maildir_use_size_file an _expandable_ boolean.
346 Bugzilla 1089. Patch from Heiko Schlittermann.
348 PP/08 Handle ${run} returning more data than OS pipe buffer size.
349 Bugzilla 1131. Patch from Holger Weiß.
351 PP/09 Handle IPv6 addresses with SPF.
352 Bugzilla 860. Patch from Wolfgang Breyha.
354 PP/10 GnuTLS: support TLS 1.2 & 1.1.
356 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
359 PP/11 match_* no longer expand right-hand-side by default.
360 New compile-time build option, EXPAND_LISTMATCH_RHS.
361 New expansion conditions, "inlist", "inlisti".
363 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
365 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
367 PP/14 fix log_write() format string regression from TF/03.
368 Bugzilla 1152. Patch from Dmitry Isaikin.
374 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
376 PP/02 Harmonised TLS library version reporting; only show if debugging.
377 Layout now matches that introduced for other libraries in 4.74 PP/03.
379 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
381 PP/04 New "dns_use_edns0" global option.
383 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
386 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
387 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
389 TK/01 Updated PolarSSL code to 0.14.2.
390 Bugzilla 1097. Patch from Andreas Metzler.
392 PP/07 Catch divide-by-zero in ${eval:...}.
395 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
398 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
399 format-string attack -- SECURITY: remote arbitrary code execution.
401 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
402 time unintentionally subject to list matching rules, letting the header
403 cause arbitrary Exim lookups (of items which can occur in lists, *not*
404 arbitrary string expansion). This allowed for information disclosure.
406 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
407 INT_MIN/-1 -- value coerced to INT_MAX.
413 NM/01 Workround for PCRE version dependency in version reporting
416 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
417 This fixes portability to compilers other than gcc, notably
418 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
420 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
421 makefiles for portability to HP-UX and POSIX correctness.
423 PP/01 Permit LOOKUP_foo enabling on the make command-line.
424 Also via indented variable definition in the Makefile.
425 (Debugging by Oliver Heesakkers).
427 PP/02 Restore caching of spamd results with expanded spamd_address.
428 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
430 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
431 Improves build reliability. Fix from: Frank Elsner
433 NM/02 Fix wide character breakage in the rfc2047 coding
434 Fixes bug 1064. Patch from Andrey N. Oktyabrski
436 NM/03 Allow underscore in dnslist lookups
437 Fixes bug 1026. Patch from Graeme Fowler
439 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
440 Code patches from Adam Ciarcinski of NetBSD.
442 NM/04 Fixed exiqgrep to cope with mailq missing size issue
445 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
446 is logged, to avoid truncation. Patch from John Horne.
448 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
449 Patch from Jakob Hirsch.
451 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
452 SQL string expansion failure details.
453 Patch from Andrey Oktyabrski.
455 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
456 Patch from Simon Arlott.
458 PP/09 New lookups functionality failed to compile on old gcc which rejects
459 extern declarations in function scope.
460 Patch from Oliver Fleischmann
462 PP/10 Use sig_atomic_t for flags set from signal handlers.
463 Check getgroups() return and improve debugging.
464 Fixed developed for diagnosis in bug 927 (which turned out to be
467 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
468 Patch from Mark Zealey.
470 PP/12 Bugzilla 1056: Improved spamd server selection.
471 Patch from Mark Zealey.
473 PP/13 Bugzilla 1086: Deal with maildir quota file races.
474 Based on patch from Heiko Schlittermann.
476 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
477 Patch from Uwe Doering, sign-off by Michael Haardt.
479 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
480 variable declaration deep within a block. Bug and patch from
483 PP/15 lookups-Makefile IRIX compatibilty coercion.
485 PP/16 Make DISABLE_DKIM build knob functional.
487 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
488 Patch by Simon Arlott
490 TF/03 Fix valgrind.h portability to C89 compilers that do not support
491 variable argument macros. Our copy now differs from upstream.
497 TF/01 Failure to get a lock on a hints database can have serious
498 consequences so log it to the panic log.
500 TF/02 Log LMTP confirmation messages in the same way as SMTP,
501 controlled using the smtp_confirmation log selector.
503 TF/03 Include the error message when we fail to unlink a spool file.
505 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
506 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
507 for maintaining out-of-tree patches for some time.
509 PP/01 Bugzilla 139: Documentation and portability issues.
510 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
511 Handle per-OS dynamic-module compilation flags.
513 PP/02 Let /dev/null have normal permissions.
514 The 4.73 fixes were a little too stringent and complained about the
515 permissions on /dev/null. Exempt it from some checks.
516 Reported by Andreas M. Kirchwitz.
518 PP/03 Report version information for many libraries, including
519 Exim version information for dynamically loaded libraries. Created
520 version.h, now support a version extension string for distributors
521 who patch heavily. Dynamic module ABI change.
523 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
524 privilege escalation vulnerability whereby the Exim run-time user
525 can cause root to append content of the attacker's choosing to
528 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
531 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
532 If dropping privileges for untrusted macros, we disabled normal logging
533 on the basis that it would fail; for the Exim run-time user, this is not
534 the case, and it resulted in successful deliveries going unlogged.
535 Fixed. Reported by Andreas Metzler.
541 PP/01 Date: & Message-Id: revert to normally being appended to a message,
542 only prepend for the Resent-* case. Fixes regression introduced in
543 Exim 4.70 by NM/22 for Bugzilla 607.
545 PP/02 Include check_rfc2047_length in configure.default because we're seeing
546 increasing numbers of administrators be bitten by this.
548 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
550 PP/03 Bugzilla 994: added openssl_options main configuration option.
552 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
554 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
556 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
558 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
559 without a peer certificate, leading to a segfault because of an
560 assumption that peers always have certificates. Be a little more
561 paranoid. Problem reported by Martin Tscholak.
563 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
564 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
565 NB: ClamAV planning to remove STREAM in "middle of 2010".
566 CL also introduces -bmalware, various -d+acl logging additions and
567 more caution in buffer sizes.
569 PP/09 Implemented reverse_ip expansion operator.
571 PP/10 Bugzilla 937: provide a "debug" ACL control.
573 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
575 PP/12 Bugzilla 973: Implement --version.
577 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
579 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
581 PP/15 Bugzilla 816: support multiple condition rules on Routers.
583 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
584 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
585 ignore trailing whitespace.
587 JJ/02 prevent non-panic DKIM error from being sent to paniclog
589 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
592 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
593 Notification from Dr Andrew Aitchison.
595 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
596 ExtendedDetectionInfo response format.
597 Notification from John Horne.
599 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
602 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
603 XSL and documented dependency on system catalogs, with examples of how
606 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
609 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
610 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
611 configuration file which is writeable by the Exim user or group.
613 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
614 of configuration files to cover files specified with the -C option if
615 they are going to be used with root privileges, not just the default
618 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
619 option (effectively making it always true).
621 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
622 files to be used while preserving root privileges.
624 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
625 that rogue child processes cannot use them.
627 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
628 run-time user, instead of root.
630 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
631 Exim run-time user without dropping privileges.
633 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
634 result string, instead of calling string_vformat() twice with the same
637 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
638 for other users. Others should always drop root privileges if they use
639 -C on the command line, even for a whitelisted configure file.
641 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
643 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
649 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
650 $data_path, and $header_path variables; fixed documentation bugs and
653 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
654 exipick to access non-standard spools, including the "frozen" queue
657 NM/01 Bugzilla 965: Support mysql stored procedures.
658 Patch from Alain Williams
660 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
662 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
663 Patch from Andreas Metzler
665 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
666 Patch from Kirill Miazine
668 NM/05 Bugzilla 671: Added umask to procmail example.
670 JJ/03 installed exipick 20100323.0, fixing doc bug
672 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
673 directory. Notification and patch from Dan Rosenberg.
675 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
677 TK/02 Improve log output when DKIM signing operation fails.
679 MH/01 Treat the transport option dkim_domain as a colon separated
680 list, not as a single string, and sign the message with each element,
681 omitting multiple occurences of the same signer.
683 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
684 Bugzilla 985, 986. Patch by Simon Arlott
686 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
687 Patch by Simon Arlott
689 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
690 MBX locking. Notification from Dan Rosenberg.
696 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
698 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
700 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
702 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
704 NM/04 Bugzilla 915: Flag broken perl installation during build.
710 TK/01 Added patch by Johannes Berg that expands the main option
711 "spamd_address" if it starts with a dollar sign.
713 TK/02 Write list of recipients to X-Envelope-Sender header when building
714 the mbox-format spool file for content scanning (suggested by Jakob
717 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
718 (http://www.dcc-servers.net/) support via dccifd. Activated by
719 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
721 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
722 by Mark Daniel Reidel <mr@df.eu>.
724 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
725 When building exim an external PCRE library is now needed -
726 PCRE is a system library on the majority of modern systems.
727 See entry on PCRE_LIBS in EDITME file.
729 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
730 conversation. Added nologin parameter to request.
731 Patch contributed by Kirill Miazine.
733 TF/01 Do not log submission mode rewrites if they do not change the address.
735 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
737 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
738 log files in place. Contributed by Roberto Lima.
740 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
742 TF/03 Bugzilla 615: When checking the local_parts router precondition
743 after a local_part_suffix or local_part_prefix option, Exim now
744 does not use the address's named list lookup cache, since this
745 contains cached lookups for the whole local part.
747 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
748 Robert Millan. Documentation is in experimental-spec.txt.
750 TF/04 Bugzilla 668: Fix parallel build (make -j).
752 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
754 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
755 Patch provided by Jan Srzednicki.
757 TF/05 Leading white space used to be stripped from $spam_report which
758 wrecked the formatting. Now it is preserved.
760 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
761 that they are available at delivery time.
763 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
765 TF/08 TLS error reporting now respects the incoming_interface and
766 incoming_port log selectors.
768 TF/09 Produce a more useful error message if an SMTP transport's hosts
769 setting expands to an empty string.
771 NM/06 Bugzilla 744: EXPN did not work under TLS.
772 Patch provided by Phil Pennock.
774 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
775 Patch provided by Richard Godbee.
777 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
778 acl_smtp_notquit, added index entry.
780 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
781 Patch provided by Eugene Bujak.
783 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
784 accept(). Patch provided by Maxim Dounin.
786 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
787 Patch provided by Phil Pennock.
789 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
791 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
792 Patch provided by Brad "anomie" Jorsch.
794 NM/14 Bugzilla 622: Added timeout setting to transport filter.
795 Patch provided by Dean Brooks.
797 TK/05 Add native DKIM support (does not depend on external libraries).
799 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
800 Patch provided by Graeme Fowler.
802 NM/16 Bugzilla 851: Documentation example syntax fix.
804 NM/17 Changed NOTICE file to remove references to embedded PCRE.
806 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
809 NM/19 Bugzilla 745: TLS version reporting.
810 Patch provided by Phil Pennock.
812 NM/20 Bugzilla 167: bool: condition support.
813 Patch provided by Phil Pennock.
815 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
816 clients. Patch provided by Phil Pennock.
818 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
819 Patch provided by Brad "anomie" Jorsch.
821 NM/23 Bugzilla 687: Fix misparses in eximstats.
822 Patch provided by Heiko Schlittermann.
824 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
825 Patch provided by Heiko Schlittermann.
827 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
828 plus update to original patch.
830 NM/26 Bugzilla 799: Documentation correction for ratelimit.
832 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
833 Patch provided by David Brownlee.
835 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
837 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
839 NM/30 Bugzilla 888: TLS documentation bugfixes.
841 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
843 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
844 Unlike the original bugzilla I have changed all shell scripts in src tree.
846 NM/33 Bugzilla 898: Transport filter timeout fix.
847 Patch by Todd Rinaldo.
849 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
850 Patch by Serge Demonchaux.
852 NM/35 Bugzilla 39: Base64 decode bug fixes.
853 Patch by Jakob Hirsch.
855 NM/36 Bugzilla 909: Correct connect() call in dcc code.
857 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
859 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
861 NM/39 Bugzilla 911: Fixed MakeLinks build script.
867 TK/01 Add preliminary DKIM support. Currently requires a forked version of
868 ALT-N's libdkim that I have put here:
869 http://duncanthrax.net/exim-experimental/
871 Note to Michael Haardt: I had to rename some vars in sieve.c. They
872 were called 'true' and it seems that C99 defines that as a reserved
873 keyword to be used with 'bool' variable types. That means you could
874 not include C99-style headers which use bools without triggering
875 build errors in sieve.c.
877 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
878 as mailq or other aliases. Changed the --help handling significantly
879 to do whats expected. exim_usage() emits usage/help information.
881 SC/01 Added the -bylocaldomain option to eximstats.
883 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
885 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
887 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
893 PH/01 Another patch from the Sieve maintainer.
895 PH/02 When an IPv6 address is converted to a string for single-key lookup
896 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
897 dots are used instead of colons so that keys in lsearch files need not
898 contain colons. This was done some time before quoting was made available
899 in lsearch files. However, iplsearch files do require colons in IPv6 keys
900 (notated using the quote facility) so as to distinguish them from IPv4
901 keys. This meant that lookups for IP addresses in host lists did not work
902 for iplsearch lookups.
904 This has been fixed by arranging for IPv6 addresses to be expressed with
905 colons if the lookup type is iplsearch. This is not incompatible, because
906 previously such lookups could never work.
908 The situation is now rather anomolous, since one *can* have colons in
909 ordinary lsearch keys. However, making the change in all cases is
910 incompatible and would probably break a number of configurations.
912 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
915 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
916 conversion specification without a maximum field width, thereby enabling
917 a rogue spamd server to cause a buffer overflow. While nobody in their
918 right mind would setup Exim to query an untrusted spamd server, an
919 attacker that gains access to a server running spamd could potentially
920 exploit this vulnerability to run arbitrary code as the Exim user.
922 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
923 $primary_hostname instead of what libspf2 thinks the hosts name is.
925 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
926 a directory entry by the name of the lookup key. Previously, if a
927 symlink pointed to a non-existing file or a file in a directory that
928 Exim lacked permissions to read, a lookup for a key matching that
929 symlink would fail. Now it is enough that a matching directory entry
930 exists, symlink or not. (Bugzilla 503.)
932 PH/03 The body_linecount and body_zerocount variables are now exported in the
935 PH/04 Added the $dnslist_matched variable.
937 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
938 This means they are set thereafter only if the connection becomes
941 PH/06 Added the client_condition to authenticators so that some can be skipped
942 by clients under certain conditions.
944 PH/07 The error message for a badly-placed control=no_multiline_responses left
945 "_responses" off the end of the name.
947 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
949 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
950 (without spaces) instead of just copying the configuration text.
952 PH/10 Added the /noupdate option to the ratelimit ACL condition.
954 PH/11 Added $max_received_linelength.
956 PH/12 Added +ignore_defer and +include_defer to host lists.
958 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
959 way in which PCRE > 7.0 is built.
961 PH/14 Implemented queue_only_load_latch.
963 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
964 MAIL command. The effect was to mangle the value on 64-bit systems.
966 PH/16 Another patch from the Sieve maintainer.
968 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
970 PH/18 If a system quota error occurred while trying to create the file for
971 a maildir delivery, the message "Mailbox is full" was not appended to the
972 bounce if the delivery eventually timed out. Change 4.67/27 below applied
973 only to a quota excession during the actual writing of the file.
975 PH/19 It seems that peer DN values may contain newlines (and other non-printing
976 characters?) which causes problems in log lines. The DN values are now
977 passed through string_printing() before being added to log lines.
979 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
980 and InterBase are left for another time.)
982 PH/21 Added message_body_newlines option.
984 PH/22 Guard against possible overflow in moan_check_errorcopy().
986 PH/23 POSIX allows open() to be a macro; guard against that.
988 PH/24 If the recipient of an error message contained an @ in the local part
989 (suitably quoted, of course), incorrect values were put in $domain and
990 $local_part during the evaluation of errors_copy.
996 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
997 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1000 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1001 issue a MAIL command.
1003 PH/02 In an ACL statement such as
1005 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1007 if a client was not listed at all, or was listed with a value other than
1008 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1009 the condition was not true (as it should be), so access was not denied.
1010 The bug was that the ! inversion was incorrectly passed on to the second
1011 item. This has been fixed.
1013 PH/03 Added additional dnslists conditions == and =& which are different from
1014 = and & when the dns lookup returns more than one IP address.
1016 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1017 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1019 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1020 FSYNC, which compiles an option called disable_fsync that allows for
1021 bypassing fsync(). The documentation is heavily laced with warnings.
1023 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1025 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1026 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1027 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1028 including adding "make clean"; (3) Added -fPIC when compiling the test
1029 dynamically loaded module, to get rid of a warning.
1031 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1032 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1033 = 0s. The bug is otherwise harmless.
1035 PH/07 There was a bug in the dovecot authenticator such that the value of
1036 $auth1 could be overwritten, and so not correctly preserved, after a
1037 successful authentication. This usually meant that the value preserved by
1038 the server_setid option was incorrect.
1040 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1042 PH/09 Installed PCRE release 7.0.
1044 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1045 run for batched SMTP input. It is now run at the start of every message
1046 in the batch. While fixing this I discovered that the process information
1047 (output by running exiwhat) was not always getting set for -bs and -bS
1048 input. This is fixed, and it now also says "batched" for BSMTP.
1050 PH/11 Added control=no_pipelining.
1052 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1053 patch, slightly modified), and move the expansion of helo_data till after
1054 the connection is made in the smtp transport (so it can use these
1057 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1059 PH/14 Added log_selector = +pid.
1061 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1063 PH/16 Add ${if forany and ${if forall.
1065 PH/17 Added dsn_from option to vary the From: line in DSNs.
1067 PH/18 Flush SMTP output before performing a callout, unless control =
1068 no_callout_flush is set.
1070 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1071 was true (the default) a successful delivery failed to delete the retry
1072 item, thus causing premature timeout of the address. The bug is now
1075 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1077 PH/21 Long custom messages for fakedefer and fakereject are now split up
1078 into multiline reponses in the same way that messages for "deny" and
1079 other ACL rejections are.
1081 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1082 with slight modification.
1084 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1085 draft, changing the syntax and factoring some duplicate code".
1087 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1088 for deliveries of the second and subsequent messages over the same SMTP
1091 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1092 ${reduce, with only minor "tidies".
1094 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1096 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1097 expansion side effects.
1099 PH/27 When a message times out after an over-quota error from an Exim-imposed
1100 quota, the bounce message says "mailbox is full". This message was not
1101 being given when it was a system quota that was exceeded. It now should
1104 MH/03 Made $recipients available in local_scan(). local_scan() already has
1105 better access to the recipient list through recipients_list[], but
1106 $recipients can be useful in postmaster-provided expansion strings.
1108 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1109 in the case of a MAIL command with additional options following the
1110 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1111 were accidentally chopped off.
1113 PH/29 SMTP synchronization checks are implemented when a command is read -
1114 there is a check that no more input is waiting when there shouldn't be
1115 any. However, for some commands, a delay in an ACL can mean that it is
1116 some time before the response is written. In this time, more input might
1117 arrive, invalidly. So now there are extra checks after an ACL has run for
1118 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1119 pipelining has not been advertised.
1121 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1123 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1124 returned in the SMTP response when smtp_return_error_details was false.
1125 This has been fixed.
1127 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1128 instead of the C I/O that was originally supplied, because problems were
1129 reported on Solaris.
1131 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1132 Exim which did not show up earlier: it was assuming that a call to
1133 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1134 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1135 was a macro that became an assignment, so it seemed to work. This has
1136 changed to a proper function call with a void return, hence the compile
1137 error. Exim's code has been fixed.
1139 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1142 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1144 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1146 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1147 because it turns out to be too big or there is a timeout) there is no
1148 "Completed" line in the log. When some messages of this type were
1149 selected by exigrep, they were listed as "not completed". Others were
1150 picked up by some special patterns. I have improved the selection
1151 criteria to be more general.
1153 PH/38 The host_find_failed option in the manualroute router can now be set
1154 to "ignore", to completely ignore a host whose IP address cannot be
1155 found. If all hosts are ignored, the behaviour is controlled by the new
1156 host_all_ignored option.
1158 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1159 homing or because of multiple MX records with /mx) generated more than
1160 one IP address, and the following item turned out to be the local host,
1161 all the secondary addresses of the first item were incorrectly removed
1162 from the list, along with the local host and any following hosts (which
1163 is what is supposed to happen).
1165 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1166 whoever called Exim into the -H file. In the case of the daemon it was
1167 behaving confusingly. When first started, it used values for whoever
1168 started the daemon, but after a SIGHUP it used the Exim user (because it
1169 calls itself on a restart). I have changed the code so that it now always
1172 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1173 message are rejected with the same error (e.g. no authentication or bad
1174 sender address), and a DATA command is nevertheless sent (as can happen
1175 with PIPELINING or a stupid MUA), the error message that was given to the
1176 RCPT commands is included in the rejection of the DATA command. This is
1177 intended to be helpful for MUAs that show only the final error to their
1180 PH/42 Another patch from the Sieve maintainer.
1182 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1183 Eximstats - Fixed some broken HTML links and added missing column headers
1185 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1186 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1188 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1190 PH/43 Yet another patch from the Sieve maintainer.
1192 PH/44 I found a way to check for a TCP/IP connection going away before sending
1193 the response to the final '.' that terminates a message, but only in the
1194 case where the client has not sent further data following the '.'
1195 (unfortunately, this is allowed). However, in many cases there won't be
1196 any further data because there won't be any more messages to send. A call
1197 to select() can be used: if it shows that the input is "ready", there is
1198 either input waiting, or the socket has been closed. An attempt to read
1199 the next input character can distinguish the two cases. Previously, Exim
1200 would have sent an OK response which the client would never have see.
1201 This could lead to message repetition. This fix should cure that, at
1202 least in a lot of common cases.
1204 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1205 advertised in response to EHLO.
1211 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1212 fixed by 4.65/MH/01 (is this a record?) are fixed:
1214 (i) An empty string was always treated as zero by the numeric comparison
1215 operators. This behaviour has been restored.
1217 (ii) It is documented that the numeric comparison operators always treat
1218 their arguments as decimal numbers. This was broken in that numbers
1219 starting with 0 were being interpreted as octal.
1221 While fixing these problems I realized that there was another issue that
1222 hadn't been noticed. Values of message_size_limit (both the global option
1223 and the transport option) were treated as octal if they started with 0.
1224 The documentation was vague. These values are now always treated as
1225 decimal, and I will make that clear in the documentation.
1231 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1232 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1235 MH/01 Don't check that the operands of numeric comparison operators are
1236 integers when their expansion is in "skipping" mode (fixes bug
1237 introduced by 4.64-PH/07).
1239 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1240 child addresses, Exim now panics and dies. Previously, because the count
1241 is held in a short int, deliveries were likely to be lost. As such a
1242 large number of recipients for a single message is ridiculous
1243 (performance will be very, very poor), I have chosen to impose a limit
1244 rather than extend the field.
1250 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1251 leftover -K file (the existence of which was triggered by #402).
1252 While we were at it, introduced process PID as part of the -K
1253 filename. This should rule out race conditions when creating
1256 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1257 processing considerably. Previous code took too long for large mails,
1258 triggering a timeout which in turn triggers #401.
1260 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1261 in the DK code in transports.c. sendfile() is not really portable,
1262 hence the _LINUX specificness.
1264 TF/01 In the add_headers option to the mail command in an Exim filter,
1265 there was a bug that Exim would claim a syntax error in any
1266 header after the first one which had an odd number of characters
1269 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1270 callout verification, Exim cached a "reject" for the entire domain. This
1271 is correct for most verifications, but it is not correct for a recipient
1272 verification with use_sender or use_postmaster set, because in that case
1273 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1274 case of MAIL FROM:<> rejection from other early rejections (e.g.
1275 rejection of HELO). When verifying a recipient using a non-null MAIL
1276 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1277 Whatever the result of the callout, the value of the domain cache is
1278 left unchanged (for any other kind of callout, getting as far as trying
1279 RCPT means that the domain itself is ok).
1281 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1284 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1285 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1286 session, the daemon ignores ECONNECT errors and logs others; it now
1287 ignores EPIPE as well.
1289 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1290 (quoted-printable decoding).
1292 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1293 later the small subsequent patch to fix an introduced bug.
1295 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1297 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1299 PH/08 An error is now given if message_size_limit is specified negative.
1301 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1302 to be given (somewhat) arbitrary names.
1304 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1307 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1308 miscellaneous code fixes
1310 PH/10 Added the log_reject_target ACL modifier to specify where to log
1313 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1314 hostname. This is wrong, because it relates to the incoming message (and
1315 probably the interface on which it is arriving) and not to the outgoing
1316 callout (which could be using a different interface). This has been
1317 changed to use the value of the helo_data option from the smtp transport
1318 instead - this is what is used when a message is actually being sent. If
1319 there is no remote transport (possible with a router that sets up host
1320 addresses), $smtp_active_hostname is used.
1322 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1323 tweaks were necessary in order to get it to work (see also 21 below):
1324 (a) The code assumed that strncpy() returns a negative number on buffer
1325 overflow, which isn't the case. Replaced with Exim's string_format()
1327 (b) There were several signed/unsigned issues. I just did the minimum
1328 hacking in of casts. There is scope for a larger refactoring.
1329 (c) The code used strcasecmp() which is not a standard C function.
1330 Replaced with Exim's strcmpic() function.
1331 (d) The code set only $1; it now sets $auth1 as well.
1332 (e) A simple test gave the error "authentication client didn't specify
1333 service in request". It would seem that Dovecot has changed its
1334 interface. Fortunately there's a specification; I followed it and
1335 changed what the client sends and it appears to be working now.
1337 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1340 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1341 address A is aliased to B and C, where B exists and C does not. Without
1342 -v the output is "A verified" because verification stops after a
1343 successful redirection if more than one address is generated. However,
1344 with -v the child addresses are also verified. Exim was outputting "A
1345 failed to verify" and then showing the successful verification for C,
1346 with its parentage. It now outputs "B failed to verify", showing B's
1347 parentage before showing the successful verification of C.
1349 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1350 look up a TXT record in a specific list after matching in a combined
1353 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1354 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1355 they consult the DNS. I had assumed they would set it the way they
1356 wanted; and indeed my experiments on Linux seem to show that in some
1357 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1358 To be on the safe side, however, I have now made the interface to
1359 host_find_byname() similar to host_find_bydns(), with an argument
1360 containing the DNS resolver options. The host_find_byname() function now
1361 sets these options at its start, just as host_find_bydns() does. The smtp
1362 transport options dns_qualify_single and dns_search_parents are passed to
1363 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1364 of host_find_byname() use the default settings of RES_DEFNAMES
1365 (qualify_single) but not RES_DNSRCH (search_parents).
1367 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1368 spool_read_header() do less string testing, by means of a preliminary
1369 switch on the second character of optional "-foo" lines. (This is
1370 overdue, caused by the large number of possibilities that now exist.
1371 Originally there were few.) While I was there, I also converted the
1372 str(n)cmp tests so they don't re-test the leading "-" and the first
1373 character, in the hope this might squeeze out yet more improvement.
1375 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1376 flag allowing group syntax was set by the header_syntax check but not
1377 turned off, possible causing trouble later; (2) The flag was not being
1378 set at all for the header_verify test, causing "group"-style headers to
1379 be rejected. I have now set it in this case, and also caused header_
1380 verify to ignore an empty address taken from a group. While doing this, I
1381 came across some other cases where the code for allowing group syntax
1382 while scanning a header line wasn't quite right (mostly, not resetting
1383 the flag correctly in the right place). These bugs could have caused
1384 trouble for malformed header lines. I hope it is now all correct.
1386 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1387 with the "reply" argument non-NULL. The code, however (which originally
1388 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1389 but it didn't always do it. This confused somebody who was copying the
1390 code for some other use. I have removed all the tests.
1392 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1393 feature that was used to support insecure browsers during the U.S. crypto
1394 embargo. It requires special client support, and Exim is probably the
1395 only MTA that supported it -- and would never use it because real RSA is
1396 always available. This code has been removed, because it had the bad
1397 effect of slowing Exim down by computing (never used) parameters for the
1398 RSA_EXPORT functionality.
1400 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1401 authenticator to fail if there's a tab character in the incoming data
1402 (there should never be unless someone is messing about, as it's supposed
1403 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1404 if the connection is using TLS or if the remote IP is the same as the
1405 local IP, and the "valid-client-cert option" if a client certificate has
1408 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1409 authenticators. This can be used for authorization after authentication
1410 succeeds. (In the case of plaintext, it servers for both authentication
1413 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1414 if any retry times were supplied.
1416 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1417 connection, where there is no client IP address to check. In this
1418 situation, the verify now always succeeds.
1420 PH/25 Applied John Jetmore's -Mset patch.
1422 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1424 PH/27 In a string expansion for a processed (not raw) header when multiple
1425 headers of the same name were present, leading whitespace was being
1426 removed from all of them, but trailing whitespace was being removed only
1427 from the last one. Now trailing whitespace is removed from each header
1428 before concatenation. Completely empty headers in a concatenation (as
1429 before) are ignored.
1431 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1432 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1434 PH/29 [Removed. This was a change that I later backed out, and forgot to
1435 correct the ChangeLog entry (that I had efficiently created) before
1436 committing the later change.]
1438 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1439 address errors (4xx response to RCPT) over the same connection as other
1440 messages routed to the same hosts. Such deliveries are always "forced",
1441 so retry times are not inspected. This resulted in far too many retries
1442 for the affected addresses. The effect occurred only when there were more
1443 hosts than the hosts_max_try setting in the smtp transport when it had
1444 the 4xx errors. Those hosts that it had tried were not added to the list
1445 of hosts for which the message was waiting, so if all were tried, there
1446 was no problem. Two fixes have been applied:
1448 (i) If there are any address or message errors in an SMTP delivery, none
1449 of the hosts (tried or untried) are now added to the list of hosts
1450 for which the message is waiting, so the message should not be a
1451 candidate for sending over the same connection that was used for a
1452 successful delivery of some other message. This seems entirely
1453 reasonable: after all the message is NOT "waiting for some host".
1454 This is so "obvious" that I'm not sure why it wasn't done
1455 previously. Hope I haven't missed anything, but it can't do any
1456 harm, as the worst effect is to miss an optimization.
1458 (ii) If, despite (i), such a delivery is accidentally attempted, the
1459 routing retry time is respected, so at least it doesn't keep
1460 hammering the server.
1462 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1463 in ${readsocket because some servers need this prod.
1465 PH/32 Added some extra debug output when updating a wait-xxx database.
1467 PH/33 The hint "could be header name not terminated by colon", which has been
1468 given for certain expansion errors for a long time, was not being given
1469 for the ${if def:h_colon_omitted{... case.
1471 PH/34 The spec says: "With one important exception, whenever a domain list is
1472 being scanned, $domain contains the subject domain." There was at least
1473 one case where this was not true.
1475 PH/35 The error "getsockname() failed: connection reset by peer" was being
1476 written to the panic log as well as the main log, but it isn't really
1477 panic-worthy as it just means the connection died rather early on. I have
1478 removed the panic log writing for the ECONNRESET error when getsockname()
1481 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1482 runs only) independently of the message's sender address. This meant
1483 that, if the 4xx error was in fact related to the sender, a different
1484 message to the same recipient with a different sender could confuse
1485 things. In particualar, this can happen when sending to a greylisting
1486 server, but other circumstances could also provoke similar problems.
1487 I have changed the default so that the retry time for these errors is now
1488 based a combination of the sender and recipient addresses. This change
1489 can be overridden by setting address_retry_include_sender=false in the
1492 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1493 remote server are returned as part of bounce messages. This was not
1494 happening for LMTP over a pipe (the lmtp transport), but now it is the
1495 same for both kinds of LMTP.
1497 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1498 in header lines that were in fact CNAMEs. This is no longer the case.
1500 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1501 and queue runs started by the daemon processed all messages. This has
1502 been fixed so that -R and -S can now usefully be given with -q<time>.
1504 PH/40 Import PCRE release 6.7 (fixes some bugs).
1506 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1508 PH/42 Give an error if -q is specified more than once.
1510 PH/43 Renamed the variables $interface_address and $interface_port as
1511 $received_ip_address and $received_port, to make it clear that these
1512 values apply to message reception, and not to the outgoing interface when
1513 a message is delivered. (The old names remain recognized, of course.)
1515 PH/44 There was no timeout on the connect() call when using a Unix domain
1516 socket in the ${readsocket expansion. There now is.
1518 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1519 be meaningful with "accept".
1521 SC/01 Eximstats V1.43
1522 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1524 SC/02 Eximstats V1.44
1525 Use a glob alias rather than an array ref in the generated
1526 parser. This improves both readability and performance.
1528 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1529 Collect SpamAssassin and rejection statistics.
1530 Don't display local sender or destination tables unless
1531 there is data to show.
1532 Added average volumes into the top table text output.
1534 SC/04 Eximstats V1.46
1535 Collect data on the number of addresses (recipients)
1536 as well as the number of messages.
1538 SC/05 Eximstats V1.47
1539 Added 'Message too big' to the list of mail rejection
1540 reasons (thanks to Marco Gaiarin).
1542 SC/06 Eximstats V1.48
1543 Mainlog lines which have GMT offsets and are too short to
1544 have a flag are now skipped.
1546 SC/07 Eximstats V1.49 (Alain Williams)
1547 Added the -emptyok flag.
1549 SC/08 Eximstats V1.50
1550 Fixes for obtaining the IP address from reject messages.
1552 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1553 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1554 whitesspace changes from 4.64-PH/27
1556 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1559 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1560 are found, allow negative numbers in numeric criteria)
1562 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1564 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1565 to match changes made in 4.64-PH/43
1567 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1569 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1570 "domain" is actually an IP address, and force a failure. This locks out
1571 those revolvers/nameservers that support "A-for-A" lookups, in
1572 contravention of the specifications.
1574 PH/48 When a host name was looked up from an IP address, and the subsequent
1575 forward lookup of the name timed out, the host name was left in
1576 $sender_host_name, contrary to the specification.
1578 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1579 restricted to single-key lookups, Exim was not diagnosing an error if
1580 * or *@ was used with a query-style lookup.
1582 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1584 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1585 long ago, but noone interested enough thought of it. Let's just say that
1586 the "1.1" means that there are some new functions that weren't there at
1587 some point in the past.
1589 PH/51 Error processing for expansion failure of helo_data from an smtp
1590 transport during callout processing was broken.
1592 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1593 tested/used via the -bh/-bhc/-bs options.
1595 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1596 bug, fixed in subsequent PCRE releases).
1598 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1599 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1601 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1607 SC/01 Use a glob alias rather than an array ref in eximstats generated
1608 parser. This improves both readability and performance.
1610 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1611 Don't display local sender or destination tables in eximstats unless
1612 there is data to show.
1613 Added average volumes into the eximstats top table text output.
1615 SC/03 Collect data on the number of addresses (recipients) as well
1616 as the number of messages in eximstats.
1618 TF/01 Correct an error in the documentation for the redirect router. Exim
1619 does (usually) call initgroups() when daemonizing.
1621 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1622 with consistent privilege compared to when running as a daemon.
1624 TF/03 Note in the spec that $authenticated_id is not set for local
1625 submissions from trusted users.
1627 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1628 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1630 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1631 by adding some example configuration directives to the default
1632 configuration file. A little bit of work is required to uncomment the
1633 directives and define how usernames and passwords are checked, but
1634 there is now a framework to start from.
1636 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1637 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1638 without this. I don't know how relevant this is to other LDAP libraries.
1640 PH/02 Add the verb name to the "unknown ACL verb" error.
1642 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1644 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1646 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1647 directory not expanded when it should be if an expanded home directory
1648 was set for the address (which is overridden by the transport).
1650 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1653 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1654 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1655 because it is too late at that time, and has no effect.
1657 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1658 security issue with \' (bugzilla #107). I could not use the
1659 PQescapeStringConn() function, because it needs a PGconn value as one of
1662 PH/08 When testing addresses using -bt, indicate those final addresses that
1663 are duplicates that would not cause an additional delivery. At least one
1664 person was confused, thinking that -bt output corresponded to deliveries.
1665 (Suppressing duplicates isn't a good idea as you lose the information
1666 about possibly different redirections that led to the duplicates.)
1668 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1669 systems where poll() doesn't work, in particular OS X.
1671 PH/10 Added more information to debugging output for retry time not reached.
1673 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1674 operations in malware.c.
1676 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1679 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1680 syslog_duplication set false, log lines that would normally be written
1681 both the the main log and to the reject log were not written to syslog at
1684 PH/14 In the default configuration, change the use of "message" in ACL warn
1685 statements to "add_header".
1687 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1688 not followed by a command (e.g. "seen endif").
1690 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1691 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1694 PH/17 Added extra conditions to the default value of delay_warning_condition
1698 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1699 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1700 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1703 The Auto-Submitted: and various List- headers are standardised, whereas I
1704 don't think Precedence: ever was.
1706 PH/18 Refactored debugging code in route_finduser() to show more information,
1707 in particular, the error code if getpwnam() issues one.
1709 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1710 This is apparently needed in addition to the PH/07 change above to avoid
1711 any possible encoding problems.
1713 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1714 but not after initializing Perl.
1716 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1717 output them only if debugging. By default they are written stderr,
1718 apparently, which is not desirable.
1720 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1723 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1726 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1728 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1729 authenticated or an ident call has been made. Suppress the default
1730 values for $authenticated_id and $authenticated_sender (but permit -oMai
1731 and -oMas) when testing with -bh.
1733 PH/24 Re-jigged the order of the tests in the default configuration so that the
1734 tests for valid domains and recipients precede the DNS black list and CSA
1735 tests, on the grounds that those ones are more expensive.
1737 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1738 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1739 This bug exists in every version of Exim that I still have, right back to
1742 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1743 However, an attempt to turn on case-sensitivity in a regex key by
1744 including (?-i) didn't work because the subject string was already
1745 lowercased, and the effects were non-intuitive. It turns out that a
1746 one-line patch can be used to allow (?-i) to work as expected.
1752 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1753 other effects) broke the use of negated acl sub-conditions.
1755 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1758 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1759 "Deny" causes Exim to reject the incoming connection with a 554 error.
1760 Unfortunately, if there is a major crisis, such as a disk failure,
1761 tcp-wrappers gives "deny", whereas what one would like would be some
1762 kind of temporary error. A kludge has been added to help with this.
1763 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1764 554 error is used if errno is still zero or contains ENOENT (which occurs
1765 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1768 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1770 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1771 errors. Otherwise a message that provokes a temporary error (when other
1772 messages do not) can cause a whole host to time out.
1774 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1775 addresses were routed directly to files or pipes from a redirect router.
1776 File deliveries just didn't batch; pipe deliveries might have suffered
1779 PH/06 A failure to get a lock for a hints database would erroneously always say
1780 "Failed to get write lock", even when it was really a read lock.
1782 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1783 of 0600. This has been changed to use the value of the lockfile_mode
1784 option (which defaults to 0600).
1786 PH/08 Applied small patch from the Sieve maintainer.
1788 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1789 folder from quota calculations, a direct delivery into this folder messed
1790 up the contents of the maildirsize file. This was because the regex was
1791 used only to exclude .Trash (or whatever) when the size of the mailbox
1792 was calculated. There was no check that a delivery was happening into an
1793 excluded directory. This bug has been fixed by ignoring all quota
1794 processing for deliveries into excluded directories.
1796 PH/10 Added the maildirfolder_create_regex option to appendfile.
1802 PH/01 The code for finding all the local interface addresses on a FreeBSD
1803 system running IPv6 was broken. This may well have applied to all BSD
1804 systems, as well as to others that have similar system calls. The broken
1805 code found IPv4 interfaces correctly, but gave incorrect values for the
1806 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1807 that it would not match correctly against @[] and not recognize the IPv6
1810 PH/02 The ipliteral router was not recognizing addresses of the form user@
1811 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1813 PH/03 Added disable_ipv6.
1815 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1816 decoded form, because it is most often used to construct To: headers
1817 lines in autoreplies, and the decoded form may well be syntactically
1818 invalid. However, $reply_address has leading white space removed, and all
1819 newlines turned into spaces so that the autoreply transport does not
1822 PH/05 If group was specified without a user on a router, and no group or user
1823 was specified on a transport, the group from the router was ignored.
1825 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1826 for visible compile-time settings that can be used to change these
1827 numbers, for those that want even more. Backwards compatibility with old
1828 spool files has been maintained. However, going back to a previous Exim
1829 release will lost any variables that are in spool files.
1831 PH/07 Two small changes when running in the test harness: increase delay when
1832 passing a TCP/IP connection to a new process, in case the original
1833 process has to generate a bounce, and remove special handling of
1834 127.0.0.2 (sic), which is no longer necessary.
1836 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1837 be the same on different OS.
1839 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1842 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1843 whether --show-vars was specified or not
1845 JJ/02 exipick: Added support for new ACL variable spool format introduced
1848 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1849 syntactically invalid From: or Reply-to: line, and a filter used this to
1850 generate an autoreply, and therefore failed to obtain an address for the
1851 autoreply, Exim could try to deliver to a non-existent relative file
1852 name, causing unrelated and misleading errors. What now happens is that
1853 it logs this as a hard delivery error, but does not attempt to create a
1856 PH/11 The exinext utility has a -C option for testing purposes, but although
1857 the given file was scanned by exinext itself; it wasn't being passed on
1858 when Exim was called.
1860 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1861 an end-of-file indication when reading a command response.
1863 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1864 compiled. In many other places in Exim, IPv6 addresses are always
1865 recognized, so I have changed this. It also means that IPv4 domain
1866 literals of the form [IPV4:n.n.n.n] are now always recognized.
1868 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1869 used if the router is not running as root, for example, when verifying at
1870 ACL time, or when using -bh. The debugging output from this situation was
1871 non-existent - all you got was a failure to exec. I have made two
1874 (a) Failures to set uid/gid, the current directory, or a process leader
1875 in a subprocess such as that created by queryprogram now generate
1876 suitable debugging ouput when -d is set.
1878 (b) The queryprogram router detects when it is not running as root,
1879 outputs suitable debugging information if -d is set, and then runs
1880 the subprocess without attempting to change uid/gid.
1882 PH/15 Minor change to Makefile for building test_host (undocumented testing
1885 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1886 additional section of a DNS packet that returns MX or SRV records.
1887 Instead, it always explicitly searches for A/AAAA records. This avoids
1888 major problems that occur when a DNS server includes only records of one
1889 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1890 fixed another bug: if SRV records were looked up and the corresponding
1891 address records were *not* found in the additional section, the port
1892 values from the SRV records were lost.
1894 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1895 using the correct key (the original address) when searching the retry
1896 rules in order to find which one to use for generating the retry hint.
1898 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1899 adding the default one. Similarly, if it contains a Reply-To: header, the
1900 errors_reply_to option, if set, is not used.
1902 PH/19 When calculating a retry time, Exim used to measure the "time since
1903 failure" by looking at the "first failed" field in the retry record. Now
1904 it does not use this if it is later than than the arrival time of the
1905 message. Instead it uses the arrival time. This makes for better
1906 behaviour in cases where some deliveries succeed, thus re-setting the
1907 "first failed" field. An example is a quota failure for a huge message
1908 when small messages continue to be delivered. Without this change, the
1909 "time since failure" will always be short, possible causing more frequent
1910 delivery attempts for the huge message than are intended.
1911 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1913 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1914 $1, $2, $3) because the numerical variables can be reset during some
1915 expansion items (e.g. "match"), thereby losing the authentication data.
1917 PH/21 Make -bV show the size of off_t variables so that the test suite can
1918 decide whether to run tests for quotas > 2G.
1920 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1921 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1922 filecount value is greater than 2G or if a quota value is greater than 2G
1923 on a system where the size of off_t is not greater than 4, a panic error
1926 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1927 never match. The debug and -bh output now contains an explicit error
1928 message indicating a malformed IPv4 address or mask.
1930 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1931 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1932 PH/23 above applies.
1934 PH/25 Do not write to syslog when running in the test harness. The only
1935 occasion when this arises is a failure to open the main or panic logs
1936 (for which there is an explicit test).
1938 PH/26 Added the /no_tell option to "control=freeze".
1940 PH/27 If a host name lookup failed very early in a connection, for example, if
1941 the IP address matched host_lookup and the reverse lookup yielded a name
1942 that did not have a forward lookup, an error message of the form "no IP
1943 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1944 could be logged. Now it outputs the IP address instead of "NULL".
1946 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1947 allows the sender and the authenticated sender to be set when
1948 submitting a message from within Exim. Since child_open_exim() is
1949 documented for local_scan(), the new function should be too.
1951 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1952 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1953 results in an empty string is now treated as unset.
1955 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1957 PH/31 Added sender_verify_fail logging option.
1959 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1960 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1961 tidied the source and removed it altogether.
1963 PH/33 When a queue run was abandoned because the load average was too high, a
1964 log line was always written; now it is written only if the queue_run log
1965 selector is set. In addition, the log line for abandonment now contains
1966 information about the queue run such as the pid. This is always present
1967 in "start" and "stop" lines but was omitted from the "abandon" line.
1969 PH/34 Omit spaces between a header name and the colon in the error message that
1970 is given when verify = headers_syntax fails (if there are lots of them,
1971 the message gets confusing).
1973 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1974 names, as there are now some PTR records that contain slashes. This check
1975 is only to protect against broken name servers that fall over on strange
1976 characters, so the fact that it applies to all lookups doesn't matter.
1978 PH/36 Now that the new test suite is complete, we can remove some of the
1979 special code in Exim that was needed for the old test suite. For example,
1980 sorting DNS records because real resolvers return them in an arbitrary
1981 order. The new test suite's fake resolver always returns records in the
1984 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1985 bounces) except when queue_only is set, to avoid logging races between
1986 the different processes.
1988 PH/38 Panic-die if .include specifies a non-absolute path.
1990 PH/39 A tweak to the "H" retry rule from its user.
1992 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1993 a label. They prevented compilation on older perls.
1995 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1996 a warning to be raised on newish perls.
1998 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1999 on queue. Changes to match documented behaviour of showing count of
2000 messages matching specified criteria.
2002 PH/40 Changed the default ident timeout from 30s to 5s.
2004 PH/41 Added support for the use of login_cap features, on those BSD systems
2005 that have them, for controlling the resources used by pipe deliveries.
2007 PH/42 The content-scanning code uses fopen() to create files in which to put
2008 message data. Previously it was not paying any attention to the mode of
2009 the files. Exim runs with umask(0) because the rest of the code creates
2010 files with open(), and sets the required mode explicitly. Thus, these
2011 files were ending up world-writeable. This was not a big issue, because,
2012 being within the spool directory, they were not world-accessible. I have
2013 created a function called modefopen, which takes an additional mode
2014 argument. It sets umask(777), creates the file, chmods it to the required
2015 mode, then resets the umask. All the relevant calls to fopen() in the
2016 content scanning code have been changed to use this function.
2018 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2019 to 24 hours. This avoids potential overflow problems when processing G
2020 and H retry rules. I suspect nobody ever tinkers with this value.
2022 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2024 PH/45 When the plaintext authenticator is running as a client, the server's
2025 challenges are checked to ensure they are valid base64 strings. By
2026 default, the authentication attempt is cancelled if an invalid string is
2027 received. Setting client_ignore_invalid_base64 true ignores these errors.
2028 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2029 they are received. Thus, the responses can be made to depend on the
2030 challenges. If an invalid string is ignored, an empty string is placed in
2033 PH/46 Messages that are created by the autoreply transport now contains a
2034 References: header, in accordance with RFCs 2822 and 3834.
2036 PH/47 Added authenticated_sender_force to the smtp transport.
2038 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2040 PH/49 Installed latest patch from the Sieve maintainer.
2042 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2043 was also set, the appendfile transport was unnecessarily scanning a
2044 directory of message files (e.g. for maildir delivery) to find the count
2045 of files (along with the size), even though it did not need this
2046 information. It now does the scan only if it needs to find either the
2047 size of the count of files.
2049 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2051 PH/52 Two bugs concerned with error handling when the smtp transport is
2054 (i) Exim was not creating retry information for temporary errors given
2055 for individual recipients after the DATA command when the smtp transport
2056 was used in LMTP mode. This meant that they could be retried too
2057 frequently, and not timed out correctly.
2059 (ii) Exim was setting the flag that allows error details to be returned
2060 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2061 recipients that were returned after the DATA command.
2063 PH/53 This is related to PH/52, but is more general: for any failing address,
2064 when detailed error information was permitted to be returned to the
2065 sender, but the error was temporary, then after the final timeout, only
2066 "retry timeout exceeded" was returned. Now it returns the full error as
2067 well as "retry timeout exceeded".
2069 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2070 do this, and (what is worse) MTAs that accept it.
2072 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2073 will now be deprecated.
2075 PH/56 New os.c-cygwin from the Cygwin maintainer.
2077 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2078 formats (previously only available in exim formats via -bpr, -bpru,
2079 and -bpra. Now also available in native and exiqgrep formats)
2081 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2082 with very large, slow to parse queues
2084 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2086 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2088 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2089 responses to authentication challenges, though it was showing the
2090 challenges; (ii) I've removed the CR characters from the debug output for
2093 PH/58 Allow for the insertion of a newline as well as a space when a string
2094 is turned into more than one encoded-word during RFC 2047 encoding. The
2095 Sieve code now uses this.
2097 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2098 data_4xx, lost_connection, tls_required.
2100 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2101 message was being sent as an SMTP response.
2103 PH/61 Add -l and -k options to exicyclog.
2105 PH/62 When verifying, if an address was redirected to one new address, so that
2106 verification continued, and the new address failed or deferred after
2107 having set something in $address_data, the value of $address_data was not
2108 passed back to the ACL. This was different to the case when no
2109 redirection occurred. The value is now passed back in both cases.
2111 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2112 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2113 use login_cap.h, so on its own it isn't the distinguishing feature. The
2114 new name refers directly to the setclassresources() function.
2116 PH/65 Added configuration files for NetBSD3.
2118 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2120 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2121 is preferred over IPv4.
2123 PH/68 The bounce_return_message and bounce_return_body options were not being
2124 honoured for bounces generated during the reception of non-SMTP messages.
2125 In particular, this applied to messages rejected by the ACL. This bug has
2126 been fixed. However, if bounce_return_message is true and bounce_return_
2127 body is false, the headers that are returned for a non-SMTP message
2128 include only those that have been read before the error was detected.
2129 (In the case of an ACL rejection, they have all been read.)
2131 PH/69 The HTML version of the specification is now built in a directory called
2132 spec_html instead of spec.html, because the latter looks like a path with
2133 a MIME-type, and this confuses some software.
2135 PH/70 Catch two compiler warnings in sieve.c.
2137 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2138 function verify_get_ident() calls ip_connect() to connect a socket, but
2139 if the "connect()" function timed out, ip_connect() used to close the
2140 socket. However, verify_get_ident() also closes the socket later, and in
2141 between Exim writes to the log, which may get opened at this point. When
2142 the socket was closed in ip_connect(), the log could get the same file
2143 descriptor number as the socket. This naturally causes chaos. The fix is
2144 not to close the socket in ip_connect(); the socket should be closed by
2145 the function that creates it. There was only one place in the code where
2146 this was missing, in the iplookup router, which I don't think anybody now
2147 uses, but I've fixed it anyway.
2149 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2150 well as to direct DNS lookups. Otherwise the handling of names in host
2151 lists is inconsistent and therefore confusing.
2157 PH/01 Two changes to the default runtime configuration:
2159 (1) Move the checks for relay_from_hosts and authenticated clients from
2160 after to before the (commented out) DNS black list checks.
2162 (2) Add control=submission to the relay_from_hosts and authenticated
2163 clients checks, on the grounds that messages accepted by these
2164 statements are most likely to be submissions.
2166 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2168 (1) Generate an error if the third argument for the ${prvs expansion is
2171 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2174 (3) Reset the variables that are obtained from the first argument of
2175 ${prvscheck and used in the second argument before leaving the code,
2176 because their memory is reclaimed, so using them afterwards may do
2179 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2180 one (it's much easier than Tom thought :-).
2182 (5) Because of (4), we can now allow for the use of $prvscheck_result
2183 inside the third argument.
2185 PH/03 For some reason, the default setting of PATH when running a command from
2186 a pipe transport was just "/usr/bin". I have changed it to
2189 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2190 anything to be listed in the output from -bV.
2192 PH/05 When a filter generated an autoreply, the entire To: header line was
2193 quoted in the delivery log line, like this:
2195 => >A.N.Other <ano@some.domain> <original@ddress> ...
2197 This has been changed so that it extracts the operative address. There
2198 may be more than one such address. If so, they are comma-separated, like
2201 => >ano@some.domain,ona@other.domain <original@ddress> ...
2203 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2204 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2205 not being looked up in the rDNS to get a host name, Exim was showing the
2206 IP address twice in Received: lines, even though the IP addresses were
2207 identical. For example:
2209 Received: from [1.2.3.4] (helo=[1.2.3.4])
2211 However, if the real host name was known, it was omitting the HELO data
2212 if it matched the actual IP address. This has been tidied up so that it
2213 doesn't show the same IP address twice.
2215 PH/07 When both +timestamp and +memory debugging was on, the value given by
2216 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2217 called by the debug printing, thereby overwriting the timestamp buffer.
2218 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2220 PH/08 When the original message was included in an autoreply transport, it
2221 always said "this is a copy of the message, including all the headers",
2222 even if body_only or headers_only was set. It now gives an appropriate
2225 PH/09 Applied a patch from the Sieve maintainer which:
2227 o fixes some comments
2228 o adds the (disabled) notify extension core
2229 o adds some debug output for the result of if/elsif tests
2230 o points to the current vacation draft in the documentation
2231 and documents the missing references header update
2235 o fixes a bug in processing the envelope test (when testing
2236 multiple envelope elements, the last element determinted the
2239 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2240 Electronic Mail") by including:
2242 Auto-submitted: auto-generated
2244 in the messages that it generates (bounce messages and others, such as
2245 warnings). In the case of bounce messages for non-SMTP mesages, there was
2246 also a typo: it was using "Auto_submitted" (underscore instead of
2247 hyphen). Since every message generated by Exim is necessarily in response
2248 to another message, thes have all been changed to:
2250 Auto-Submitted: auto-replied
2252 in accordance with these statements in the RFC:
2254 The auto-replied keyword:
2256 - SHOULD be used on messages sent in direct response to another
2257 message by an automatic process,
2259 - MUST NOT be used on manually-generated messages,
2261 - MAY be used on Delivery Status Notifications (DSNs) and Message
2262 Disposition Notifications (MDNs),
2264 - MUST NOT be used on messages generated by automatic or periodic
2265 processes, except for messages which are automatic responses to
2268 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2269 to the default Received: header definition.
2271 PH/12 Added log selector acl_warn_skipped (default on).
2273 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2274 variables because (a) they are in the wrong storage pool and (b) even if
2275 they were copied, it wouldn't work properly because of the caching.
2277 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2278 checking when decoding. Apparently there are clients that generate
2279 overlong encoded strings. Why am I not surprised?
2281 PH/15 If the first argument of "${if match_address" was not empty, but did not
2282 contain an "@" character, Exim crashed. Now it writes a panic log message
2283 and treats the condition as false.
2285 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2287 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2288 extension "envelope-auth". The code is finished and in agreement with
2289 other implementations, but there is no documentation so far and in fact,
2290 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2291 not changing the active code.
2293 Print executed "if" and "elsif" statements when debugging is used. This
2294 helps a great deal to understand what a filter does.
2296 Document more things not specified clearly in RFC3028. I had all this
2297 sorted out, when out of a sudden new issues came to my mind. Oops."
2299 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2302 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2303 canonical form (as documented). However, after a host name lookup from
2304 the IP address, check_host() was doing a simple string comparison with
2305 addresses acquired from the DNS when checking that the found name did
2306 have the original IP as one of its addresses. Since any found IPv6
2307 addresses are likely to be in abbreviated form, the comparison could
2308 fail. Luckily, there already exists a function for doing the comparison
2309 by converting both addresses to binary, so now that is used instead of
2310 the text comparison.
2312 PH/20 There was another similar case to PH/19, when a complete host name was
2313 given in a host list; looking up its IP address could give an abbreviated
2314 form, whereas the current host's name might or might not be abbreviated.
2315 The same fix has been applied.
2321 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2322 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2325 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2326 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2328 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2330 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2331 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2332 Also, for "Auto-Submitted", treat anything other than "no" as
2333 non-personal, in accordance with RFC 3834. (Previously it treated
2334 anything starting "auto-" as non-personal.)
2336 TF/01 The control=submission/name=... option had a problem with syntax
2337 errors if the name included a slash character. The /name= option
2338 now slurps the rest of the string, so it can include any characters
2339 but it must come last in the list of options (after /sender_retain
2342 PH/05 Some modifications to the interface to the fake nameserver for the new
2350 TK/01 Added the "success_on_redirect" address verification option. See
2351 NewStuff for rationale and an example.
2353 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2355 PH/02 Patch to exigrep to allow it to work on syslog lines.
2357 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2358 fread() to skip over the body file's header line, because in Cygwin the
2359 header line is locked and is inaccessible.
2361 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2362 co-exist for some time) to make it clear that it is the Exim ID that is
2363 referenced, not the Message-ID: header line.
2365 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2366 string_format() function, because snprintf() does not exist on all
2369 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2370 ${stat: expansion item.
2372 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2373 protocol synchronization error", to keep the pedants happy.
2375 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2376 well as for IRIX systems, when gcc is being used. See the host.c source
2379 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2381 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2384 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2385 transport and to the smtp transport in LMTP mode.
2387 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2389 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2390 run a filter in a subprocess. This could lead to confusion in subsequent
2391 lookups in the parent process. There should also be a search_tidyup() at
2392 the end of the subprocess.
2394 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2395 only if the host matched helo_try_verify_hosts, which caused the
2396 verification to occur when the EHLO/HELO command was issued. The ACL just
2397 tested the remembered result. Now, if a previous verification attempt has
2398 not happened, "verify = helo" does it there and then.
2400 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2402 TK/03 Fix log output including CR from clamd.
2404 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2405 exist provoked a memory error which could cause a segfault.
2407 PH/15 Installed PCRE 6.2
2409 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2411 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2412 of the problem. Specifically, suggested +O2 rather than +O1 for the
2415 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2417 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2418 record, the port was not correctly logged when the outgoing_port log
2419 selector was set (it logged the transort's default port).
2421 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2422 fallback_hosts, and "hosts" in the smtp transport.
2424 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2425 host errors such as "Connection refused".
2427 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2428 authentication with radiusclient 0.4.9:
2430 - Error returned from rc_read_config was caught wrongly
2431 - Username/password not passed on to radius server due to wrong length.
2433 The presumption is that some radiusclient API changes for 4.51/PH/17
2434 were not taken care of correctly. The code is still untested by me (my
2435 Linux distribution still has 0.3.2 of radiusclient), but it was
2436 contributed by a Radius user.
2438 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2439 expanding the "port" option of the smtp transport.
2441 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2442 while reading a MIME header. Thanks to Tom Hughes for a patch.
2444 PH/24 Include config.h inside local_scan.h so that configuration settings are
2447 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2448 that in an ACL for RCPT (for example), you can examine exactly what was
2451 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2452 commands, but it was not correctly comparing the address with the actual
2453 client host address. Thus, it would show the EHLO address in Received:
2454 header lines when this was not necessary.
2456 PH/27 Added the % operator to ${eval:}.
2458 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2459 it should be ignoring failures (because with -C, for example, it has lost
2460 privilege). It wasn't ignoring creation failures other than "already
2463 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2466 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2467 because an input line was too long, either on its own, or by virtue of
2468 too many continuations, the temporary file was not being removed, and the
2469 return code was incorrect.
2471 PH/31 Missing "BOOL" in function definition in filtertest.c.
2473 PH/32 Applied Sieve patches from the maintainer.
2475 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2477 PH/33 Added "verify = not_blind".
2479 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2480 Local/Makefile (with some defaults set). These are used in built scripts
2481 such as exicyclog, but they have never been used in the exim_install
2482 script (though there are many overriding facilities there). I have
2483 arranged that the exim_install script now takes note of these two
2486 PH/35 Installed configuration files for Dragonfly.
2488 PH/36 When a locally submitted message by a trusted user did not contain a
2489 From: header, and the sender address was obtained from -f or from an SMTP
2490 MAIL command, and the trusted user did not use -F to supply a sender
2491 name, $originator_name was incorrectly used when constructing a From:
2492 header. Furthermore, $originator_name was used for submission mode
2493 messages from external hosts without From: headers in a similar way,
2494 which is clearly wrong.
2496 PH/37 Added control=suppress_local_fixups.
2498 PH/38 When log_selector = +received_sender was set, and the addition of the
2499 sender made the log line's construction buffer exactly full, or one byte
2500 less than full, an overflow happened when the terminating "\n" was
2503 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2504 when the result of a list match is failure because a DNS lookup failed.
2506 PH/40 RM_COMMAND is now used in the building process.
2508 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2509 the "build-* directories that it finds.
2511 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2512 address in a domain literal was a prefix of an interface address.
2514 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2515 when verifying a sender address, unless rewrite_headers is false.
2517 PH/44 Wrote a long comment about why errors_to addresses are verified as
2518 recipients, not senders.
2520 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2521 the ratelimit ACL was added.
2523 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2525 PH/46 Added extra information about PostgreSQL errors to the error string.
2527 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2528 avoiding the need to install special zones in a real server. This is
2529 backwards compatible; if it can't find the fake resolver, it drops back.
2530 Thus, both old and new test suites can be run.
2532 TF/02 Added util/ratelimit.pl
2534 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2535 clock is set back in time.
2537 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2538 Candler <B.Candler@pobox.com>.
2540 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2541 OK for addresses that are the result of redirections.
2543 PH/48 A number of further additions for the benefit of the new test suite,
2544 including a fake gethostbyname() that interfaces to the fake DNS resolver
2547 TF/06 The fix for widen_domains has also been applied to qualify_single and
2548 search_parents which are the other dnslookup options that can cause
2551 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2554 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2556 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2557 service" if it encountered an MX record with an empty target hostname.
2558 The message is now "an MX or SRV record indicated no SMTP service".
2560 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2561 if the DNS of the sending site is misconfigured. This is quite a
2562 common situation. This change restores the behaviour of treating a
2563 helo verification defer as a failure.
2565 PH/51 If self=fail was set on a router, the bounce message did not include the
2566 actual error message.
2572 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2574 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2575 command itself ended in error, the underlying message about the transport
2576 filter timeout was being overwritten with the pipe command error. Now the
2577 underlying error message should be appended to the second error message.
2579 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2581 PH/02 Reduce the amount of output that "make" produces by default. Full output
2582 can still be requested.
2584 PH/03 The warning log line about a condition test deferring for a "warn" verb
2585 was being output only once per connection, rather than after each
2586 occurrence (because it was using the same function as for successful
2587 "warn" verbs). This seems wrong, so I have changed it.
2589 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2590 it should not have, which might have caused a crash in the right
2591 circumstances, but probably never did.
2593 PH/04 Installed a modified version of Tony Finch's patch to make submission
2594 mode fix the return path as well as the Sender: header line, and to
2595 add a /name= option so that you can make the user's friendly name appear
2598 TF/03 Added the control = fakedefer ACL modifier.
2600 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2601 Mark Lowes for thorough testing.
2603 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2605 TK/03 Merged latest SRS patch from Miles Wilton.
2607 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2608 with the definition in sysexits.h (which is #included earlier).
2609 Fortunately, Exim does not actually use EX_OK. The code used to try to
2610 preserve the sysexits.h value, by assumimg that macro definitions were
2611 scanned for macro replacements. I have been disabused of this notion,
2612 so now the code just undefines EX_OK before #including unistd.h.
2614 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2615 in the smtp transport. When a block could not be written in a single
2616 write() function, the timeout was being re-applied to each part-write.
2617 This seems wrong - if the receiver was accepting one byte at a time it
2618 would take for ever. The timeout is now adjusted when this happens. It
2619 doesn't have to be particularly precise.
2621 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2622 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2624 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2625 without a domain if the check to <postmaster@domain> fails.
2627 SC/01 Eximstats: added -xls and the ability to specify output files
2628 (patch written by Frank Heydlauf).
2630 SC/02 Eximstats: use FileHandles for outputing results.
2632 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2634 SC/04 Eximstats: fixed display of large numbers with -nvr option
2636 SC/05 Eximstats: fixed merging of reports with empty tables.
2638 SC/06 Eximstats: added the -include_original_destination flag
2640 SC/07 Eximstats: removed tabs and trailing whitespace.
2642 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2644 TK/06 MBOX spool code: Add real "From " MBOX separator line
2645 so the .eml file is really in mbox format (even though
2646 most programs do not really care). Patch from Alex Miller.
2648 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2649 The latter is generated from $received_to and is only set if the
2650 message has one envelope recipient. SA can use these headers,
2651 obviously out-of-the-box. Patch from Alex Miller.
2653 PH/08 The ${def test on a variable was returning false if the variable's
2654 value was "0", contrary to what the specification has always said!
2655 The result should be true unless the variable is empty.
2657 PH/09 The syntax error of a character other than { following "${if
2658 def:variable_name" (after optional whitespace) was not being diagnosed.
2659 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2660 accidental colon was present, for example, could give incorrect results.
2662 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2663 result is used (not including appendfile, where other changes are about
2666 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2667 This involved changing a lot of size variables from int to off_t. It
2668 should work with maildirs and everything.
2670 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2671 spamd dying while we are connected to it.
2673 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2676 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2677 function for BDB 4.3.
2679 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2681 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2682 the sources for PCRE have been re-arranged and more files are now
2685 PH/15 The code I had for printing potentially long long variables in PH/11
2686 above was not the best (it lost precision). The length of off_t variables
2687 is now inspected at build time, and an appropriate printing format (%ld
2688 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2689 to be "long long int" or "long int". This is needed for the internal
2690 formatting function string_vformat().
2692 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2693 the configuration file to be ":syslog", then the script "guesses" where
2694 the logs files are, rather than using the compiled in default. In our
2695 case the guess is not the same as the compiled default, so the script
2696 suddenly stopped working when I started to use syslog. The patch checks
2697 to see if log_file_path is "". If so, it attempts to read it from exim
2698 with no configuration file to get the compiled in version, before it
2699 falls back to the previous guessing code."
2701 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2702 implementing BATV in an Exim configuration. See NewStuff for the gory
2705 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2706 Makefile that are specific to HP-UX.
2708 PH/18 If the "use_postmaster" option was set for a recipient callout together
2709 with the "random" option, the postmaster address was used as the MAIL
2710 FROM address for the random test, but not for the subsequent recipient
2711 test. It is now used for both.
2713 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2714 patch removes a few documentation additions to RFC 3028, because the
2715 latest draft now contains them. It adds the new en;ascii-case comparator
2716 and a new error check for 8bit text in MIME parts. Comparator and
2717 require names are now matched exactly. I enabled the subaddress
2718 extension, but it is not well tested yet (read: it works for me)."
2720 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2721 rework some of the code of TK/09 above to avoid the hardwired use of
2722 "%lld" and "long long". Replaced the call to snprintf() with a call to
2725 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2726 records point to non-existent hosts", "retry timeout exceeded", and
2727 "retry time not reached for any host after a long failure period".
2729 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2730 experimental DomainKeys support:
2732 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2733 (2) On an error such as an illegally used "control", the wrong name for
2734 the control was given.
2736 These problems did NOT occur unless DomainKeys support was compiled.
2738 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2740 PH/24 Added ${if match_ip condition.
2742 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2743 databases so that it will be absolutely obvious if a crash occurs in the
2744 DB library. This is a regular occurrence (often caused by mis-matched
2747 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2748 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2749 picked up on a user's system that detects such things. There doesn't seem
2750 to be a gcc warning option for this - only an attribute that has to be
2751 put on the function's prototype. It seems that in Fedora Core 4 they have
2752 set this on a number of new functions. No doubt there will be more in due
2755 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2756 specify a transport. However, if an address that was verified by such a
2757 router was the subject of a callout, Exim crashed because it tried to
2758 read the rcpt_include_affixes from the non-existent transport. Now it
2759 just assumes that the setting of that option is false. This bug was
2760 introduced by 4.51/PH/31.
2762 PH/28 Changed -d+all to exclude +memory, because that information is very
2763 rarely of interest, but it makes the output a lot bigger. People tend to
2764 do -d+all out of habit.
2766 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2767 code in os-type was giving problems when libc.so lives in lib64, like on
2770 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2771 aren't the modern standard, and it seems that some systems' include files
2772 don't always have them. Exim was already checking for some of the newer
2773 ones like T_AAAA, and defining it itself. I've added checks for all the
2774 record types that Exim uses.
2776 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2777 not automatically generating a new one, as it is supposed to. This
2778 prevented TLS from working. If the file did exist, but contained invalid
2779 data, a new version was generated, as expected. It was only the case of a
2780 non-existent file that was broken.
2782 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2783 with a change in libdomainkeys > 0.64.
2785 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2786 from DNS. If the selector record carries the flag, it now has
2787 precedence over the domain-wide flag.
2789 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2791 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2792 the use of a port name that isn't defined in /etc/services) occurred, the
2793 message was deferred as in a normal delivery, and thus remained on the
2794 spool, instead of being failed because of the mua_wrapper setting. This
2795 is now fixed, and I tidied up some of the mua_wrapper messages at the
2798 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2799 the messages in a hash of arrays rather than using individual hashes.
2800 This is a bit cleaner and results in dramatic memory savings, albeit
2801 at a slight CPU cost.
2803 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2804 as requested by Marc Sherman.
2806 SC/10 Eximstats: added histograms for user specified patterns as requested
2809 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2811 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2812 fopen() in the content-scanning modules that did not already have it.
2818 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2819 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2821 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2823 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2825 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2826 to test Sieve filters that use "vacation".
2828 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2829 that changes the way the GnuTLS parameters are stored in the cache file.
2830 The new format can be generated externally. For backward compatibility,
2831 if the data in the cache doesn't make sense, Exim assumes it has read an
2832 old-format file, and it generates new data and writes a new file. This
2833 means that you can't go back to an older release without removing the
2836 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2837 work if there are any delivery delays because "one_time" forces the
2838 parent to be marked "delivered", so its unseen clone is never tried
2839 again. For this reason, Exim now forbids the simultaneous setting of
2842 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2843 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2844 entry if you want to know the details. The fix, however, neglected to
2845 consider the case when local delivery batching is involved. The test for
2846 "previously delivered" was not happening when checking to see if an
2847 address could be batched with a previous (undelivered) one; under
2848 certain circumstances this could lead to multiple deliveries to the same
2851 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2852 in its include files, and this causes problems building Exim.
2854 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2855 header_syntax) but Exim was just ignoring anything given after a slash.
2856 In particular, this caused confusion with an attempt to use "verify =
2857 reverse_host_lookup/defer_ok". An error is now given when options are
2858 supplied for verify items that do not have them. (Maybe reverse_host_
2859 lookup should have a defer_ok option, but that's a different point.)
2861 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2862 defined by RFC 821) to 2048, because there were problems with some AUTH
2863 commands, and RFC 1869 says the size should be increased for extended
2864 SMTP commands that take arguments.
2866 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2869 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2870 "unknown" error; now it says that the functionality isn't in the binary.
2872 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2873 an address' error message when a string expansion fails (syntax or
2874 whatever). Otherwise the password may appear in the log. Following change
2875 PH/42 below, there is no longer a chance of it appearing in a bounce
2878 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2880 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2881 of its addresses was ever tried. (Bugzilla bug #2.)
2883 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2884 the result incorrectly in the debug output. (It correctly added a newline
2885 to what was transported.)
2887 TF/01 Added $received_time.
2889 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2890 commented out examples of how to interface to a virus scanner and to
2891 SpamAssassin. Also added commented examples of av_scanner and
2892 spamd_address settings.
2894 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2895 and controls are allowed in which ACLs. There were a couple of minor
2896 errors. Some of the entries in the conditions table (which is a table of
2897 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2898 negation of where the condition IS allowed.
2900 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2902 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2903 header file does not have a version number, so I've had to invent a new
2904 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2905 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2906 radiusclient), but it was contributed by a Radius user.
2908 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2909 files or directories, for OpenSSL.
2911 PH/19 When an Exim process that is running as root has to create an Exim log
2912 file, it does so in a subprocess that runs as exim:exim so as to get the
2913 ownership right at creation (otherwise, other Exim processes might see
2914 the file with the wrong ownership). There was no test for failure of this
2915 fork() call, which would lead to the process getting stuck as it waited
2916 for a non-existent subprocess. Forks do occasionally fail when resources
2917 run out. I reviewed all the other calls to fork(); they all seem to check
2920 PH/20 When checking for unexpected SMTP input at connect time (before writing
2921 the banner), Exim was not dealing correctly with a non-positive return
2922 from the read() function. If the client had disconnected by this time,
2923 the result was a log entry for a synchronization error with an empty
2924 string after "input=" when read() returned zero. If read() returned -1
2925 (an event I could not check), uninitialized data bytes were printed.
2926 There were reports of junk text (parts of files, etc) appearing after
2929 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2931 PH/22 Added support for macro redefinition, and (re)definition in between
2932 driver and ACL definitions.
2934 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2935 forgetting to use the resulting value; it was using the unexpanded value.
2937 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2938 hadn't been configured. The fix is from Juergen Kreileder, who
2939 understands it better than I do:
2941 "Here's what I see happening with three configured cyrus_sasl
2942 authenticators configured (plain, login, cram-md5):
2944 On startup auth_cyrus_sasl_init() gets called for each of these.
2945 This means three calls to sasl_listmech() without a specified mech_list.
2946 => SASL tests which mechs of all available mechs actually work
2947 => three warnings about OTP not working
2948 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2950 With the patch, sasl_listmech() also gets called three times. But now
2951 SASL's mech_list option is set to the server_mech specified in the the
2952 authenticator. Or in other words, the answer from sasl_listmech()
2953 gets limited to just the mech you're testing for (which is different
2955 => the return list contains just 'plain' or 'login', 'cram-md5' or
2956 nothing depending on the value of ob->server_mech.
2958 I've just tested the patch: Authentication still works fine,
2959 unavailable mechs specified in the exim configuration are still
2960 caught, and the auth.log warnings about OTP are gone."
2962 PH/25 When debugging is enabled, the contents of the command line are added
2963 to the debugging output, even when log_selector=+arguments is not
2966 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2967 answer is "GNU", and only if the return is "GNU/something" is the answer
2970 PH/27 $acl_verify_message is now set immediately after the failure of a
2971 verification in an ACL, and so is available in subsequent modifiers. In
2972 particular, the message can be preserved by coding like this:
2974 warn !verify = sender
2975 set acl_m0 = $acl_verify_message
2977 Previously, $acl_verify_message was set only while expanding "message"
2978 and "log_message" when a very denied access.
2980 PH/28 Modified OS/os.c-Linux with
2982 -#ifndef OS_LOAD_AVERAGE
2983 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2985 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2986 nomenclature these days.)
2988 PH/29 Installed patch from the Sieve maintainer that adds the options
2989 sieve_useraddress and sieve_subaddress to the redirect router.
2991 PH/30 In these circumstances:
2992 . Two addresses routed to the same list of hosts;
2993 . First host does not offer TLS;
2994 . First host accepts first address;
2995 . First host gives temporary error to second address;
2996 . Second host offers TLS and a TLS session is established;
2997 . Second host accepts second address.
2998 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2999 and peerdn, if requested) that were in fact used only for the second
3002 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3003 attention to any local part prefix or suffix that was matched by the
3004 router that accepted the address. It now behaves in the same way as it
3005 does for delivery: the affixes are removed from the local part unless
3006 rcpt_include_affixes is set on the transport.
3008 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3009 timeout during the DATA phase of an incoming message.
3011 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3012 applied a patch sanctioned by the Sieve maintainer.
3014 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3015 the uid or gid is negative. A case of a negative gid caused this to be
3016 noticed. The fix allows for either to be negative.
3018 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3019 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3022 PH/36 The change PH/12 above was broken. Fixed it.
3024 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3025 the grounds that routing the same address twice would always produce the
3026 same answer. This might have been true once, but it is certainly no
3027 longer true now. Routing a child address may depend on the previous
3028 routing that produced that child. Some complicated redirection strategies
3029 went wrong when messages had multiple recipients, and made Exim's
3030 behaviour dependent on the order in which the addresses were given.
3032 I have moved the duplicate checking until after the routing is complete.
3033 Exim scans the addresses that are assigned to local and remote
3034 transports, and removes any duplicates. This means that more work will be
3035 done, as duplicates will always all be routed, but duplicates are
3036 presumably rare, so I don't expect this is of any significance.
3038 For deliveries to pipes, files, and autoreplies, the duplicate checking
3039 still happens during the routing process, since they are not going to be
3042 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3043 It corrects a timeout issue with spamd. This is Ian's comment: "The
3044 background is that sometimes spamd either never reads data from a
3045 connection it has accepted, or it never writes response data. The exiscan
3046 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3047 blindly assumes that writes won't block so it may never time out."
3049 PH/39 Allow G after quota size as well as K and M.
3051 PH/40 The value set for $authenticated_id in an authenticator may not contain
3052 binary zeroes or newlines because the value is written to log lines and
3053 to spool files. There was no check on this. Now the value is run through
3054 the string_printing() function so that such characters are converted to
3055 printable escape sequences.
3057 PH/41 $message_linecount is a new variable that contains the total number of
3058 lines in the message. Compare $body_linecount, which is the count for the
3061 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3062 bounce and delay warning messages, except in certain special cases, which
3065 (a) An SMTP error message from a remote host;
3066 (b) A message specified in a :fail: redirection;
3067 (c) A message specified in a "fail" command in a system filter;
3068 (d) A message specified in a FAIL return from the queryprogram router;
3069 (e) A message specified by the cannot_route_message router option.
3071 In these cases only, Exim does include the error details in bounce and
3072 warning messages. There are also a few cases where bland messages such
3073 as "unrouteable address" or "local delivery error" are given.
3075 PH/43 $value is now also set for the "else" part of a ${run expansion.
3077 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3078 being worked on, but at least Exim now implements the latest version to
3081 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3082 process to complete was treated as a delivery failure, a timeout while
3083 writing the message to the pipe was logged, but erroneously treated as a
3084 successful delivery. Such timeouts include transport filter timeouts. For
3085 consistency with the overall process timeout, these timeouts are now
3086 treated as errors, giving rise to delivery failures by default. However,
3087 there is now a new Boolean option for the pipe transport called
3088 timeout_defer, which, if set TRUE, converts the failures into defers for
3089 both kinds of timeout. A transport filter timeout is now identified in
3092 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3093 systems where "make" and "gmake" are different, calling "gmake" at top
3094 level broke things. I've arranged for the value of $(MAKE) to be passed
3095 from the Makefile to this script so that it can call the same version of
3099 A note about Exim versions 4.44 and 4.50
3100 ----------------------------------------
3102 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3103 changes of various kinds. As a consequence, a big documentation update was
3104 needed. This delayed the release for rather longer than seemed good, especially
3105 in the light of a couple of (minor) security issues. Therefore, the changes
3106 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3107 release. So 4.44 and 4.50 are in effect two different branches that both start
3110 I have left the 4.50 change log unchanged; it contains all the changes since
3111 4.43. The change log for 4.44 is below; many of its items are identical to
3112 those for 4.50. This seems to be the most sensible way to preserve the
3113 historical information.
3119 1. Minor wording change to the doc/README.SIEVE file.
3121 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3122 computation of the current number of files was incorrect.
3124 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3125 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3128 4. Give more explanation in the error message when the command for a transport
3129 filter fails to execute.
3131 5. There are several places where Exim runs a non-Exim command in a
3132 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3133 was being done only for the command run by the queryprogram router. It is
3134 now done for all such subprocesses. The other cases are: ${run, transport
3135 filters, and the commands run by the lmtp and pipe transports.
3137 6. Added CONFIGURE_GROUP build-time option.
3139 7. Some older OS have a limit of 256 on the maximum number of file
3140 descriptors. Exim was using setrlimit() to set 1000 as a large value
3141 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3142 systems. I've change it so that if it can't get 1000, it tries for 256.
3144 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3145 was an oversight, and furthermore, ever since the addition of extra
3146 controls (e.g. 4.43/32), the checks on when to allow different forms of
3147 "control" were broken. There should now be diagnostics for all cases when a
3148 control that does not make sense is encountered.
3150 9. Added the /retain_sender option to "control=submission".
3152 10. $recipients is now available in the predata ACL (oversight).
3154 11. Tidy the search cache before the fork to do a delivery from a message
3155 received from the command line. Otherwise the child will trigger a lookup
3156 failure and thereby defer the delivery if it tries to use (for example) a
3157 cached ldap connection that the parent has called unbind on.
3159 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3160 of $address_data from the recipient verification was clobbered by the
3161 sender verification.
3163 13. The value of address_data from a sender verification is now available in
3164 $sender_address_data in subsequent conditions in the ACL statement.
3166 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3168 15. Added a new option "connect=<time>" to callout options, to set a different
3171 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3172 was its contents. (It was OK if the option was not defined at all.)
3174 17. A "Completed" log line is now written for messages that are removed from
3175 the spool by the -Mrm option.
3177 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3178 information about exactly what failed.
3180 19. Added -dd to debug only the daemon process.
3182 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3183 handles timeouts, both on the server side and network timeouts. Renamed the
3184 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3186 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3187 It is now set to "smtps".
3189 22. $host_address is now set to the target address during the checking of
3190 ignore_target_hosts.
3192 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3193 being passed; this would have caused $sender_host_name to have been used if
3194 matching the list had actually called for a host name (not very likely,
3195 since this list is usually IP addresses). A host name is now passed as
3198 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3199 code that specifies a non-restarting handler (typically sigaction() in
3200 modern systems) in an attempt to fix a rare and obscure crash bug.
3202 25. Narrowed the window for a race in the daemon that could cause it to ignore
3203 SIGCHLD signals. This is not a major problem, because they are used only to
3204 wake it up if nothing else does.
3206 26. A malformed maildirsize file could cause Exim to calculate negative values
3207 for the mailbox size or file count. Odd effects could occur as a result.
3208 The maildirsize information is now recalculated if the size or filecount
3211 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3212 support for a long time. Removed HAVE_SYS_VFS_H.
3214 28. Installed the latest version of exipick from John Jetmore.
3216 29. In an address list, if the pattern was not a regular expression, an empty
3217 subject address (from a bounce message) matched only if the pattern was an
3218 empty string. Non-empty patterns were not even tested. This was the wrong
3219 because it is perfectly reasonable to use an empty address as part of a
3220 database query. An empty address is now tested by patterns that are
3221 lookups. However, all the other forms of pattern expect the subject to
3222 contain a local part and a domain, and therefore, for them, an empty
3223 address still always fails if the pattern is not itself empty.
3225 30. Exim went into a mad DNS loop when attempting to do a callout where the
3226 host was specified on an smtp transport, and looking it up yielded more
3227 than one IP address.
3229 31. Re-factored the code for checking spool and log partition space into a
3230 function that finds that data and another that does the check. The former
3231 is then used to implement four new variables: $spool_space, $log_space,
3232 $spool_inodes, and $log_inodes.
3234 32. The RFC2047 encoding function was originally intended for short strings
3235 such as real names; it was not keeping to the 75-character limit for
3236 encoded words that the RFC imposes. It now respects the limit, and
3237 generates multiple encoded words if necessary. To be on the safe side, I
3238 have increased the buffer size for the ${rfc2047: expansion operator from
3241 33. It is now permitted to omit both strings after an "if" condition; if the
3242 condition is true, the result is "true". As before, when the second string
3243 is omitted, a false condition yields an empty string. This makes it less
3244 cumbersome to write custom ACL and router conditions.
3246 34. Failure to deliver a bounce message always caused it to be frozen, even if
3247 there was an errors_to setting on the router. The errors_to setting is now
3250 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3251 canonical form (fully expanded) before being placed in
3252 $sender_host_address.
3254 36. The table in the code that translates DNS record types into text (T_A to
3255 "A" for instance) was missing entries for NS and CNAME. It is just possible
3256 that this could have caused confusion if both these types were looked up
3257 for the same domain, because the text type is used as part of Exim's
3258 per-process caching. But the chance of anyone hitting this buglet seems
3261 37. The dnsdb lookup has been extended in a number of ways.
3263 (1) There is a new type, "zns", which walks up the domain tree until it
3264 finds some nameserver records. It should be used with care.
3266 (2) There is a new type, "mxh", which is like "mx" except that it returns
3267 just the host names, not the priorities.
3269 (3) It is now possible to give a list of domains (or IP addresses) to be
3270 looked up. The behaviour when one of the lookups defers can be
3271 controlled by a keyword.
3273 (4) It is now possible to specify the separator character for use when
3274 multiple records are returned.
3276 38. The dnslists ACL condition has been extended: it is now possible to supply
3277 a list of IP addresses and/or domains to be looked up in a particular DNS
3280 39. Added log_selector=+queue_time_overall.
3282 40. When running the queue in the test harness, wait just a tad after forking a
3283 delivery process, to get repeatability of debugging output.
3285 41. Include certificate and key file names in error message when GnuTLS fails
3286 to set them up, because the GnuTLS error message doesn't include the name
3287 of the failing file when there is a problem reading it.
3289 42. Allow both -bf and -bF in the same test run.
3291 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3293 44. The "Exiscan patch" is now merged into the mainline Exim source.
3295 45. Sometimes the final signoff response after QUIT could fail to get
3296 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3297 < 0 before doing a fflush(). This bug looks as though it goes back to the
3298 introduction of TLS in release 3.20, but "sometimes" must have been rare
3299 because the tests only now provoked it.
3301 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3302 (this can affect the format of dates).
3304 47. exim_tidydb, when checking for the continued existence of a message for
3305 which it has found a message-specific retry record, was not finding
3306 messages that were in split spool directories. Consequently, it was
3307 deleting retry records that should have stayed in existence.
3309 48. Steve fixed some bugs in eximstats.
3311 49. The SPA authentication driver was not abandoning authentication and moving
3312 on to the next authenticator when an expansion was forced to fail,
3313 contradicting the general specification for all authenticators. Instead it
3314 was generating a temporary error. It now behaves as specified.
3316 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3317 (the order specifies the preference for clients). The order is now AES256,
3318 AES128, 3DES, ARCFOUR128.
3320 51. Small patch to Sieve code - explicitly set From: when generating an
3323 52. Exim crashed if a remote delivery caused a very long error message to be
3324 recorded - for instance if somebody sent an entire SpamAssassin report back
3325 as a large number of 550 error lines. This bug was coincidentally fixed by
3326 increasing the size of one of Exim's internal buffers (big_buffer) that
3327 happened as part of the Exiscan merge. However, to be on the safe side, I
3328 have made the code more robust (and fixed the comments that describe what
3331 53. Now that there can be additional text after "Completed" in log lines (if
3332 the queue_time_overall log selector is set), a one-byte patch to exigrep
3333 was needed to allow it to recognize "Completed" as not the last thing in
3336 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3337 patch that reportedly fixes this has been added. I am not expert enough to
3338 create a test for it. This is what the patch creator wrote:
3340 "I found a little strange behaviour of ldap code when working with
3341 Windows 2003 AD Domain, where users was placed in more than one
3342 Organization Units. When I tried to give exim partial DN, the exit code
3343 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3344 But simultaneously result of request was absolutely normal ldap result,
3345 so I produce this patch..."
3347 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3348 REFERENCE, so I have modified the code to exclude the patch when that macro
3351 55. Some experimental protocols are using DNS PTR records for new purposes. The
3352 keys for these records are domain names, not reversed IP addresses. The
3353 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3354 leaves it alone. Component reversal etc. now happens only for IP addresses.
3357 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3359 57. Double the size of the debug message buffer (to 2048) so that more of very
3360 long debug lines gets shown.
3362 58. The exicyclog utility now does better if the number of log files to keep
3363 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3365 59. Two changes related to the smtp_active_hostname option:
3367 (1) $smtp_active_hostname is now available as a variable.
3368 (2) The default for smtp_banner uses $smtp_active_hostname instead
3369 of $primary_hostname.
3371 60. The host_aton() function is supposed to be passed a string that is known
3372 to be a valid IP address. However, in the case of IPv6 addresses, it was
3373 not checking this. This is a hostage to fortune. Exim now panics and dies
3374 if the condition is not met. A case was found where this could be provoked
3375 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3376 components; fortuitously, this particular loophole had already been fixed
3377 by change 4.50/55 above.
3379 If there are any other similar loopholes, the new check in host_aton()
3380 itself should stop them being exploited. The report I received stated that
3381 data on the command line could provoke the exploit when Exim was running as
3382 exim, but did not say which command line option was involved. All I could
3383 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3384 running as the user.
3387 61. There was a buffer overflow vulnerability in the SPA authentication code
3388 (which came originally from the Samba project). I have added a test to the
3389 spa_base64_to_bits() function which I hope fixes it.
3392 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3393 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3395 63. The daemon start-up calls getloadavg() while still root for those OS that
3396 need the first call to be done as root, but it missed one case: when
3397 deliver_queue_load_max is set with deliver_drop_privilege. This is
3398 necessary for the benefit of the queue runner, because there is no re-exec
3399 when deliver_drop_privilege is set.
3401 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3402 This has been fixed.
3404 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3405 was in use, was not putting the data itself into the right store pool;
3406 consequently, it could be overwritten for a subsequent message in the same
3407 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3410 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3412 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3413 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3414 the function were treating the return as a boolean value, which happened to
3415 work because 0=false and not-0=true, but is not correct code.
3417 68. The host_aton() function was not handling scoped IPv6 addresses (those
3418 with, for example, "%eth0" on the end) correctly.
3420 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3421 negated items (that is, ~something) in unsigned ints. Some compilers
3422 apparently mutter when there is no cast.
3424 70. If an address verification called from an ACL failed, and did not produce a
3425 user-specific message (i.e. there was only a "system" message), nothing was
3426 put in $acl_verify_message. In this situation, it now puts the system
3429 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3430 session; change 4.31/43 added the unwanted input to the log line - except
3431 that it did not do this in the start of session case. It now does.
3433 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3434 This is wrong and can cause the other end to generate a synchronization
3435 error if it is another Exim or anything else that does the synchronization
3436 check. A QUIT command is no longer sent after a timeout.
3438 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3439 during host lookups.
3441 74. The defer_ok option of callout verification was not working if it was used
3442 when verifying addresses in header lines, that is, for this case:
3444 verify = header_sender/callout=defer_ok
3446 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3447 those file descriptors could be used for SMTP connections. If anything
3448 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3449 could be sent to the SMTP client, causing chaos. The daemon now opens
3450 stdin, stdout, and stderr to /dev/null when it puts itself into the
3453 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3454 log by default. The user can override this with suitable Perl magic.
3456 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3457 the log message when discard triggers, was not working for the DATA ACL or
3458 for the non-SMTP ACL.
3460 78. Error message wording change in sieve.c.
3462 79. If smtp_accept_max_per_host was set, the number of connections could be
3463 restricted to fewer than expected, because the daemon was trying to set up
3464 a new connection before checking whether the processes handling previous
3465 connections had finished. The check for completed processes is now done
3466 earlier. On busy systems, this bug wouldn't be noticed because something
3467 else would have woken the daemon, and it would have reaped the completed
3470 80. If a message was submitted locally by a user whose login name contained one
3471 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3472 It caused a spool format error. I have fixed the spool reading code. A
3473 related problem was that the "from" clause in the Received: line became
3474 illegal because of the space(s). It is now covered by ${quote_local_part.
3476 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3479 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3481 83. Patch from Sieve maintainer to fix unterminated string problem in
3482 "vacation" handling.
3484 84. Some minor changes to the Linux configuration files to help with other
3485 OS variants using glibc.
3487 85. One more patch for Sieve to update vacation handling to latest spec.
3490 ----------------------------------------------------
3491 See the note above about the 4.44 and 4.50 releases.
3492 ----------------------------------------------------
3498 1. Change 4.43/35 introduced a bug that caused file counts to be
3499 incorrectly computed when quota_filecount was set in an appendfile
3502 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3503 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3506 3. Give more explanation in the error message when the command for a transport
3507 filter fails to execute.
3509 4. There are several places where Exim runs a non-Exim command in a
3510 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3511 was being done only for the command run by the queryprogram router. It is
3512 now done for all such subprocesses. The other cases are: ${run, transport
3513 filters, and the commands run by the lmtp and pipe transports.
3515 5. Some older OS have a limit of 256 on the maximum number of file
3516 descriptors. Exim was using setrlimit() to set 1000 as a large value
3517 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3518 systems. I've change it so that if it can't get 1000, it tries for 256.
3520 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3521 was an oversight, and furthermore, ever since the addition of extra
3522 controls (e.g. 4.43/32), the checks on when to allow different forms of
3523 "control" were broken. There should now be diagnostics for all cases when a
3524 control that does not make sense is encountered.
3526 7. $recipients is now available in the predata ACL (oversight).
3528 8. Tidy the search cache before the fork to do a delivery from a message
3529 received from the command line. Otherwise the child will trigger a lookup
3530 failure and thereby defer the delivery if it tries to use (for example) a
3531 cached ldap connection that the parent has called unbind on.
3533 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3534 of $address_data from the recipient verification was clobbered by the
3535 sender verification.
3537 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3538 was its contents. (It was OK if the option was not defined at all.)
3540 11. A "Completed" log line is now written for messages that are removed from
3541 the spool by the -Mrm option.
3543 12. $host_address is now set to the target address during the checking of
3544 ignore_target_hosts.
3546 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3547 being passed; this would have caused $sender_host_name to have been used if
3548 matching the list had actually called for a host name (not very likely,
3549 since this list is usually IP addresses). A host name is now passed as
3552 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3553 code that specifies a non-restarting handler (typically sigaction() in
3554 modern systems) in an attempt to fix a rare and obscure crash bug.
3556 15. Narrowed the window for a race in the daemon that could cause it to ignore
3557 SIGCHLD signals. This is not a major problem, because they are used only to
3558 wake it up if nothing else does.
3560 16. A malformed maildirsize file could cause Exim to calculate negative values
3561 for the mailbox size or file count. Odd effects could occur as a result.
3562 The maildirsize information is now recalculated if the size or filecount
3565 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3566 support for a long time. Removed HAVE_SYS_VFS_H.
3568 18. Updated exipick to current release from John Jetmore.
3570 19. Allow an empty sender to be matched against a lookup in an address list.
3571 Previously the only cases considered were a regular expression, or an
3574 20. Exim went into a mad DNS lookup loop when doing a callout where the
3575 host was specified on the transport, if the DNS lookup yielded more than
3578 21. The RFC2047 encoding function was originally intended for short strings
3579 such as real names; it was not keeping to the 75-character limit for
3580 encoded words that the RFC imposes. It now respects the limit, and
3581 generates multiple encoded words if necessary. To be on the safe side, I
3582 have increased the buffer size for the ${rfc2047: expansion operator from
3585 22. Failure to deliver a bounce message always caused it to be frozen, even if
3586 there was an errors_to setting on the router. The errors_to setting is now
3589 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3590 canonical form (fully expanded) before being placed in
3591 $sender_host_address.
3593 24. Updated eximstats to version 1.33
3595 25. Include certificate and key file names in error message when GnuTLS fails
3596 to set them up, because the GnuTLS error message doesn't include the name
3597 of the failing file when there is a problem reading it.
3599 26. Expand error message when OpenSSL has problems setting up cert/key files.
3602 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3603 (this can affect the format of dates).
3605 28. exim_tidydb, when checking for the continued existence of a message for
3606 which it has found a message-specific retry record, was not finding
3607 messages that were in split spool directories. Consequently, it was
3608 deleting retry records that should have stayed in existence.
3610 29. eximstats updated to version 1.35
3611 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3612 1.35 - bugfix such that pie charts by volume are generated correctly
3614 30. The SPA authentication driver was not abandoning authentication and moving
3615 on to the next authenticator when an expansion was forced to fail,
3616 contradicting the general specification for all authenticators. Instead it
3617 was generating a temporary error. It now behaves as specified.
3619 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3620 (the order specifies the preference for clients). The order is now AES256,
3621 AES128, 3DES, ARCFOUR128.
3623 31. Small patch to Sieve code - explicitly set From: when generating an
3626 32. Exim crashed if a remote delivery caused a very long error message to be
3627 recorded - for instance if somebody sent an entire SpamAssassin report back
3628 as a large number of 550 error lines. This bug was coincidentally fixed by
3629 increasing the size of one of Exim's internal buffers (big_buffer) that
3630 happened as part of the Exiscan merge. However, to be on the safe side, I
3631 have made the code more robust (and fixed the comments that describe what
3634 33. Some experimental protocols are using DNS PTR records for new purposes. The
3635 keys for these records are domain names, not reversed IP addresses. The
3636 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3637 leaves it alone. Component reversal etc. now happens only for IP addresses.
3640 34. The host_aton() function is supposed to be passed a string that is known
3641 to be a valid IP address. However, in the case of IPv6 addresses, it was
3642 not checking this. This is a hostage to fortune. Exim now panics and dies
3643 if the condition is not met. A case was found where this could be provoked
3644 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3645 components; fortuitously, this particular loophole had already been fixed
3646 by change 4.50/55 or 4.44/33 above.
3648 If there are any other similar loopholes, the new check in host_aton()
3649 itself should stop them being exploited. The report I received stated that
3650 data on the command line could provoke the exploit when Exim was running as
3651 exim, but did not say which command line option was involved. All I could
3652 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3653 running as the user.
3656 35. There was a buffer overflow vulnerability in the SPA authentication code
3657 (which came originally from the Samba project). I have added a test to the
3658 spa_base64_to_bits() function which I hope fixes it.
3661 36. The daemon start-up calls getloadavg() while still root for those OS that
3662 need the first call to be done as root, but it missed one case: when
3663 deliver_queue_load_max is set with deliver_drop_privilege. This is
3664 necessary for the benefit of the queue runner, because there is no re-exec
3665 when deliver_drop_privilege is set.
3667 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3668 was in use, was not putting the data itself into the right store pool;
3669 consequently, it could be overwritten for a subsequent message in the same
3670 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3673 38. Sometimes the final signoff response after QUIT could fail to get
3674 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3675 < 0 before doing a fflush(). This bug looks as though it goes back to the
3676 introduction of TLS in release 3.20, but "sometimes" must have been rare
3677 because the tests only now provoked it.
3683 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3684 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3685 Now it returns an integer. A number of calls were still expecting a T/F
3686 return. Fortuitously, in all cases, the tests worked in OK situations,
3687 which is the norm. However, things would have gone wrong on any write
3688 failures on the smtp file descriptor. This function is used when sending
3689 messages over SMTP and also when doing verify callouts.
3691 2. When Exim is called to do synchronous delivery of a locally submitted
3692 message (the -odf or -odi options), it no longer closes stderr before doing
3695 3. Implemented the mua_wrapper option.
3697 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3699 5. Implemented the functions header_remove(), header_testname(),
3700 header_add_at_position(), and receive_remove_recipient(), and exported them
3703 6. If an ACL "warn" statement specified the addition of headers, Exim already
3704 inserted X-ACL-Warn: at the start if there was no header name. However, it
3705 was not making this test for the second and subsequent header lines if
3706 there were newlines in the string. This meant that an invalid header could
3707 be inserted if Exim was badly configured.
3709 7. Allow an ACL "warn" statement to add header lines at the start or after all
3710 the Received: headers, as well as at the end.
3712 8. Added the rcpt_4xx retry error code.
3714 9. Added postmaster_mailfrom=xxx to callout verification option.
3716 10. Added mailfrom=xxxx to the callout verification option, for verify=
3719 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3720 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3722 12. Inserted some casts to stop certain compilers warning when using pointer
3723 differences as field lengths or precisions in printf-type calls (mostly
3724 affecting debugging statements).
3726 13. Added optional readline() support for -be (dynamically loaded).
3728 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3729 same clock tick as a message's arrival, so that its received time was the
3730 same as the "first fail" time on the retry record, and that message
3731 remained on the queue past the ultimate address timeout, every queue runner
3732 would try a delivery (because it was past the ultimate address timeout) but
3733 after another failure, the ultimate address timeout, which should have then
3734 bounced the address, did not kick in. This was a "< instead of <=" error;
3735 in most cases the first failure would have been in the next clock tick
3736 after the received time, and all would be well.
3738 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3739 being recognized when the domain list was tested by the match_domain
3740 condition in an expansion string.
3742 16. Added the ${str2b64: operator.
3744 17. Exim was always calling setrlimit() to set a large limit for the number of
3745 processes, without checking whether the existing limit was already
3746 adequate. (It did check for the limit on file descriptors.) Furthermore,
3747 errors from getrlimit() and setrlimit() were being ignored. Now they are
3748 logged to the main and panic logs, but Exim does carry on, to try to do its
3749 job under whatever limits there are.
3751 18. Imported PCRE 5.0.
3753 19. Trivial typo in log message " temporarily refused connection" (the leading
3756 20. If the log selector return_path_on_delivery was set and an address was
3757 redirected to /dev/null, the delivery process crashed because it assumed
3758 that a return path would always be set for a "successful" delivery. In this
3759 case, the whole delivery is bypassed as an optimization, and therefore no
3762 21. Internal re-arrangement: the function for sending a challenge and reading
3763 a response while authentication was assuming a zero-terminated challenge
3764 string. It's now changed to take a pointer and a length, to allow for
3765 binary data in such strings.
3767 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3769 23. Exim was not respecting finduser_retries when seeking the login of the
3770 uid under which it was called; it was always trying 10 times. (The default
3771 setting of finduser_retries is zero.) Also, it was sleeping after the final
3772 failure, which is pointless.
3774 24. Implemented tls_on_connect_ports.
3776 25. Implemented acl_smtp_predata.
3778 26. If the domain in control=submission is set empty, Exim assumes that the
3779 authenticated id is a complete email address when it generates From: or
3780 Sender: header lines.
3782 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3783 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3784 chown and chgrp in /bin and hostname in /usr/bin.
3786 28. Exim was keeping the "process log" file open after each use, just as it
3787 does for the main log. This opens the possibility of it remaining open for
3788 long periods when the USR1 signal hits a daemon. Occasional processlog
3789 errors were reported, that could have been caused by this. Anyway, it seems
3790 much more sensible not to leave this file open at all, so that is what now
3793 29. The long-running daemon process does not normally write to the log once it
3794 has entered its main loop, and it closes the log before doing so. This is
3795 so that log files can straightforwardly be renamed and moved. However,
3796 there are a couple of unusual error situations where the daemon does write
3797 log entries, and I had neglected to close the log afterwards.
3799 30. The text of an SMTP error response that was received during a remote
3800 delivery was being truncated at 512 bytes. This is too short for some of
3801 the long messages that one sometimes sees. I've increased the limit to
3804 31. It is now possible to make retry rules that apply only when a message has a
3805 specific sender, in particular, an empty sender.
3807 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3808 it possible to be selective about when SMTP synchronization is enforced.
3810 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3812 32. Implemented hosts_connection_nolog.
3814 33. Added an ACL for QUIT.
3816 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3819 35. Added mailbox_size and mailbox_filecount to appendfile.
3821 36. Added control = no_multiline_responses to ACLs.
3823 37. There was a bug in the logic of the code that waits for the clock to tick
3824 in the case where the clock went backwards by a substantial amount such
3825 that the microsecond fraction of "now" was more than the microsecond
3826 fraction of "then" (but the whole seconds number was less).
3828 38. Added support for the libradius Radius client library this is found on
3829 FreeBSD (previously only the radiusclient library was supported).
3835 1. When certain lookups returned multiple values in the form name=value, the
3836 quoting of the values was not always being done properly. Specifically:
3837 (a) If the value started with a double quote, but contained no whitespace,
3839 (b) If the value contained whitespace other than a space character (i.e.
3840 tabs or newlines or carriage returns) it was not quoted.
3841 This fix has been applied to the mysql and pgsql lookups by writing a
3842 separate quoting function and calling it from the lookup code. The fix
3843 should probably also be applied to nisplus, ibase and oracle lookups, but
3844 since I cannot test any of those, I have not disturbed their existing code.
3846 2. A hit in the callout cache for a specific address caused a log line with no
3847 reason for rejecting RCPT. Now it says "Previous (cached) callout
3848 verification failure".
3850 3. There was an off-by-one bug in the queryprogram router. An over-long
3851 return line was truncated at 256 instead of 255 characters, thereby
3852 overflowing its buffer with the terminating zero. As well as fixing this, I
3853 have increased the buffer size to 1024 (and made a note to document this).
3855 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3856 when Exim is waiting for an SMTP response from a remote server, Exim
3857 restarts its select() call on the socket, thereby resetting its timeout.
3858 This is not a problem when such interrupts are rare. Somebody set up a cron
3859 job to run exiwhat every 2 minutes, which is less than the normal select()
3860 timeout (5 or 10 minutes). This meant that the select() timeout never
3861 kicked in because it was always reset. I have fixed this by comparing the
3862 time when an interrupt arrives with the time at the start of the first call
3863 to select(). If more time than the timeout has elapsed, the interrupt is
3864 treated as a timeout.
3866 5. Some internal re-factoring in preparation for the addition of Sieve
3867 extensions (by MH). In particular, the "personal" test is moved to a
3868 separate function, and given an option for scanning Cc: and Bcc: (which is
3869 not set for Exim filters).
3871 6. When Exim created an email address using the login of the caller as the
3872 local part (e.g. when creating a From: or Sender: header line), it was not
3873 quoting the local part when it contained special characters such as @.
3875 7. Installed new OpenBSD configuration files.
3877 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3878 try to make them clearer.
3880 9. Callout options, other than the timeout value, were being ignored when
3881 verifying sender addresses in header lines. For example, when using
3883 verify = header_sender/callout=no_cache
3885 the cache was (incorrectly) being used.
3887 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3888 only the Cygwin environment.
3890 11. When return_path_on_delivery was set as a log selector, if different remote
3891 addresses in the same message used different return paths and parallel
3892 remote delivery occurred, the wrong values would sometimes be logged.
3893 (Whenever a remote delivery process finished, the return path value from
3894 the most recently started remote delivery process was logged.)
3896 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3897 lines when AUTH and/or TLS are in use. This is the "received protocol"
3898 field. Exim used to use "asmtp" for authenticated SMTP, without any
3899 indication (in the protocol name) for TLS use. Now it follows the RFC and
3900 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3901 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3902 names appear in log lines as well as in Received: header lines.
3904 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3905 extensions, and comparison tests, and to fix some bugs.
3907 14. Changes to the "personal" filter test:
3909 (1) The test was buggy in that it was just doing the equivalent of
3910 "contains" tests on header lines. For example, if a user's address was
3911 anne@some.where, the "personal" test would incorrectly be true for
3913 To: susanne@some.where
3915 This test is now done by extracting each address from the header in turn,
3916 and checking the entire address. Other tests that are part of "personal"
3917 are now done using regular expressions (for example, to check local parts
3918 of addresses in From: header lines).
3920 (2) The list of non-personal local parts in From: addresses has been
3921 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3922 taken from the Sieve specification recommendations.
3924 (3) If the message contains any header line starting with "List-" it is
3925 treated as non-personal.
3927 (4) The test for "circular" in the Subject: header line has been removed
3928 because it now seems ill-conceived.
3930 15. Minor typos in src/EDITME comments corrected.
3932 16. Installed latest exipick from John Jetmore.
3934 17. If headers_add on a router specified a text string that was too long for
3935 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3936 of string_sprintf() is now avoided.
3938 18. $message_body_size was not set (it was always zero) when running the DATA
3939 ACL and the local_scan() function.
3941 19. For the "mail" command in an Exim filter, no default was being set for
3942 the once_repeat time, causing a random time value to be used if "once" was
3943 specified. (If the value happened to be <= 0, no repeat happened.) The
3944 default is now 0s, meaning "never repeat". The "vacation" command was OK
3945 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3946 (I found it when inspecting the code).
3948 20. There is now an overall timeout for performing a callout verification. It
3949 defaults to 4 times the callout timeout, which applies to individual SMTP
3950 commands during the callout. The overall timeout applies when there is more
3951 than one host that can be tried. The timeout is checked before trying the
3952 next host. This prevents very long delays if there are a large number of
3953 hosts and all are timing out (e.g. when the network connections are timing
3954 out). The value of the overall timeout can be changed by specifying an
3955 additional sub-option for "callout", called "maxwait". For example:
3957 verify = sender/callout=5s,maxwait=20s
3959 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3960 to the end before writing, but this should make it even safer).
3962 22. Exim was forgetting that it had advertised PIPELINING for the second and
3963 subsequent messages on an SMTP connection. It was also not resetting its
3964 memory on STARTTLS and an internal HELO.
3966 23. When Exim logs an SMTP synchronization error within a session, it now
3967 records whether PIPELINING has been advertised or not.
3969 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3970 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3971 rather than long int.
3973 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3975 26. Added the never_mail option to autoreply.
3981 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3982 crash if the getsockname() call failed; this can happen if a connection is
3983 closed very soon after it is established. The problem was simply in the
3984 order in which certain operations were done, causing Exim to try to write
3985 to the SMTP stream before it had set up the file descriptor. The bug has
3986 been fixed by making things happen in the correct order.
3992 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3993 before the connection was closed, thus losing the rejection response.
3995 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3996 some early Solaris releases, but causes trouble in current releases where
3997 socklen_t is defined.
3999 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4002 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4005 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4006 that releasing the top of it at the end releases what was used for sub-
4007 expansions (unless the block got too big). However, discard this block if
4008 the first thing is a variable or header, so that we can use its block when
4009 it is dynamic (useful for very large $message_headers, for example).
4011 6. Lookups now cache *every* query, not just the most recent. A new, separate
4012 store pool is used for this. It can be recovered when all lookup caches are
4013 flushed. Lookups now release memory at the end of their result strings.
4014 This has involved some general refactoring of the lookup sources.
4016 7. Some code has been added to the store_xxx() functions to reduce the amount
4017 of flapping under certain conditions.
4019 8. log_incoming_interface used to affect only the <= reception log lines. Now
4020 it causes the local interface and port to be added to several more SMTP log
4021 lines, for example "SMTP connection from", and rejection lines.
4023 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4025 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4027 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4028 could be overwritten at the end of the current message (or the start of a
4029 new message if it was set in a HELO ACL). The value is now preserved for
4030 the duration of the SMTP connection.
4032 12. If a transport had a headers_rewrite setting, and a matching header line
4033 contained an unqualified address, that address was qualified, even if it
4034 did not match any rewriting rules. The underlying bug was that the values
4035 of the flags that permit the existence of unqualified sender and recipient
4036 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4037 non-local messages, and by -bnq for local messages) were not being
4038 preserved with the message after it was received.
4040 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4041 "next input=" as part of the text comprising the host identity instead of
4042 the correct text. The code was using the same buffer for two different
4043 strings. However, depending on which order the printing function evaluated
4044 its arguments, the bug did not always show up. Under Linux, for example, my
4045 test suite worked just fine.
4047 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4048 doesn't work with some older versions of Perl. It has been changed to "my",
4049 which in any case is probably the better facility to use.
4051 15. A really picky compiler found some instances of statements for creating
4052 error messages that either had too many or two few arguments for the format
4055 16. The size of the buffer for calls to the DNS resolver has been increased
4056 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4057 for addresses that have a lot of PTR records. This alleviates a problem; it
4058 does not fully solve it.
4060 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4061 buffer now truncates the list and logs the incident, which is the same
4062 action as happens when Exim is looking up a host name and its aliases.
4063 Previously in this situation something unpredictable would happen;
4064 sometimes it was "internal error: store_reset failed".
4066 18. If a server dropped the connection unexpectedly when an Exim client was
4067 using GnuTLS and trying to read a response, the client delivery process
4068 crashed while trying to generate an error log message.
4070 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4071 string, for example:
4073 warn message = H1: something\nH2: something
4075 the text was added as a single header line from Exim's point of view
4076 though it ended up OK in the delivered message. However, searching for the
4077 second and subsequent header lines using $h_h2: did not work. This has been
4078 fixed. Similarly, if a system filter added multiple headers in this way,
4079 the routers could not see them.
4081 20. Expanded the error message when iplsearch is called with an invalid key to
4082 suggest using net-iplsearch in a host list.
4084 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4085 ACLs are no longer actually imposed (and a message to that effect is
4088 22. If a "gecos" field in a passwd entry contained escaped characters, in
4089 particular, if it contained a \" sequence, Exim got it wrong when building
4090 a From: or a Sender: header from that name. A second bug also caused
4091 incorrect handling when an unquoted " was present following a character
4092 that needed quoting.
4094 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4095 was not being matched caselessly.
4097 24. Arranged for all hyphens in the exim.8 source to be escaped with
4100 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4101 properly. Recipient callout cache records were still being keyed to include
4102 the sender, even when use_sender was set false. This led to far more
4103 callouts that were necessary. The sender is no longer included in the key
4104 when use_sender is false.
4106 26. Added "control = submission" modifier to ACLs.
4108 27. Added the ${base62d: operator to decode base 62 numbers.
4110 28. dnsdb lookups can now access SRV records.
4112 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4113 the configuration file.
4115 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4116 (-v) mode. This makes the output for a verbose queue run more intelligible.
4118 31. Added a use_postmaster feature to recipient callouts.
4120 32. Added the $body_zerocount variable, containing the number of binary zero
4121 bytes in the message body.
4123 33. The time of last modification of the "new" subdirectory is now used as the
4124 "mailbox time last read" when there is a quota error for a maildir
4127 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4129 35. Added +ignore_unknown as a special item in host lists.
4131 36. Code for decoding IPv6 addresses in host lists is now included, even if
4132 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4133 address was recognized as an IP address, but was then not correctly decoded
4134 into binary, causing unexpected and incorrect effects when compared with
4141 1. Very minor rewording of debugging text in manualroute to say "list of
4142 hosts" instead of "hostlist".
4144 2. If verify=header_syntax was set, and a header line with an unqualified
4145 address (no domain) and a large number of spaces between the end of the
4146 name and the colon was received, the reception process suffered a buffer
4147 overflow, and (when I tested it) crashed. This was caused by some obsolete
4148 code that should have been removed. The fix is to remove it!
4150 3. When running in the test harness, delay a bit after writing a bounce
4151 message to get a bit more predictability in the log output.
4153 4. Added a call to search_tidyup() just before forking a reception process. In
4154 theory, someone could use a lookup in the expansion of smtp_accept_max_
4155 per_host which, without the tidyup, could leave open a database connection.
4157 5. Added the variables $recipient_data and $sender_data which get set from a
4158 lookup success in an ACL "recipients" or "senders" condition, or a router
4159 "senders" option, similar to $domain_data and $local_part_data.
4161 6. Moved the writing of debug_print from before to after the "senders" test
4164 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4165 problems for message scanning, either using a data ACL, or using
4166 local_scan() because the Received: header was not generated till after they
4167 were called (in order to set the time as the time of reception completion).
4168 I have revised the way this works. The header is now generated after the
4169 body is received, but before the ACL or local_scan() are called. After they
4170 are run, the timestamp in the header is updated.
4176 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4177 before starting a queue runner without re-exec. This happened only when
4178 deliver_drop_privilege was set or when the Exim user was set to root. The
4179 effect of the bug was that timeouts during subsequent deliveries caused
4180 crashes instead of being properly handled. The handler is now left at its
4181 default (and expected) setting.
4183 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4184 message, again when deliver_drop_privilege is set or Exim is run as root.
4185 The bug described in (1) was not present in this case, but the tidying up
4186 of the other signals was missing. I have made the two cases consistent.
4188 3. The ignore_target_hosts setting on a manualroute router was being ignored
4189 for hosts that were looked up using the /MX notation.
4191 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4194 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4195 operated on the sender address. After changing the $sender_address to <>
4196 for the sender address verify, Exim was re-instated it as the original
4197 (before rewriting) address, but remembering that it had rewritten it, so it
4198 wasn't rewriting it again. This bug also had the effect of breaking the
4199 sender address verification caching when the sender address was rewritten.
4201 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4202 This has been changed so that if the ip literal address matches
4203 ignore_target_hosts, the router declines.
4205 7. Added expansion conditions match_domain, match_address, and match_local_
4206 part (NOT match_host).
4208 8. The placeholder for the Received: header didn't have a length field set.
4210 9. Added code to Exim itself and to exim_lock to test for a specific race
4211 condition that could lead to file corruption when using MBX delivery. The
4212 issue is with the lockfile that is created in /tmp. If this file is removed
4213 after a process has opened it but before that process has acquired a lock,
4214 there is the potential for a second process to recreate the file and also
4215 acquire a lock. This could lead to two Exim processes writing to the file
4216 at the same time. The added code performs the same test as UW imapd; it
4217 checks after acquiring the lock that its file descriptor still refers to
4218 the same named file.
4220 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4221 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4224 11. Added the smtp_active_hostname option. If used, this will typically be made
4225 to depend on the incoming interface address. Because $interface_address is
4226 not set up until the daemon has forked a reception process, error responses
4227 that can happen earlier (such as "too many connections") no longer contain
4230 12. If an expansion in a condition on a "warn" statement fails because a lookup
4231 defers, the "warn" statement is abandoned, and the next ACL statement is
4232 processed. Previously this caused the whole ACL to be aborted.
4234 13. Added the iplsearch lookup type.
4236 14. Added ident_timeout as a log selector.
4238 15. Added tls_certificate_verified as a log selector.
4240 16. Added a global option tls_require_ciphers (compare the smtp transport
4241 option of the same name). This controls incoming TLS connections.
4243 17. I finally figured out how to make tls_require_ciphers do a similar thing
4244 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4245 before starting the TLS session.
4247 18. Tabs are now shown as \t in -bP output.
4249 19. If the log selector return_path_on_delivery was set, Exim crashed when
4250 bouncing a message because it had too many Received: header lines.
4252 20. If two routers both had headers_remove settings, and the first one included
4253 a superfluous trailing colon, the final name in the first list and the
4254 first name in the second list were incorrectly joined into one item (with a
4255 colon in the middle).
4261 1. Added -C and -D options to the exinext utility, mainly to make it easier
4262 to include in the automated testing, but these could be helpful when
4263 multiple configurations are in use.
4265 2. The exinext utility was not formatting the output nicely when there was
4266 an alternate port involved in the retry record key, nor when there was a
4267 message id as well (for retries that were specific to a specific message
4268 and a specific host). It was also confused by IPv6 addresses, because of
4269 the additional colons they contain. I have fixed the IPv4 problem, and
4270 patched it up to do a reasonable job for IPv6.
4272 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4273 delivery, the log line now contains "pipelined" if PIPELINING was used.
4275 4. An SMTP transport process used to panic and die if the bind() call to set
4276 an explicit outgoing interface failed. This has been changed; it is now
4277 treated in the same way as a connect() failure.
4279 5. A reference to $sender_host_name in the part of a conditional expansion
4280 that was being skipped was still causing a DNS lookup. This no longer
4283 6. The def: expansion condition was not recognizing references to header lines
4284 that used bh_ and bheader_.
4286 7. Added the _cache feature to named lists.
4288 8. The code for checking quota_filecount in the appendfile transport was
4289 allowing one more file than it should have been.
4291 9. For compatibility with Sendmail, the command line option
4299 and sets the incoming protocol and host name (for trusted callers). The
4300 host name and its colon can be omitted when only the protocol is to be set.
4301 Note the Exim already has two private options, -pd and -ps, that refer to
4302 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4303 "s", but I don't think that's a major issue.
4305 10. A number of refactoring changes to the code, none of which should affect
4308 (a) The number of logging options was getting close to filling up the
4309 32-bit word that was used as a bit map. I have split them into two classes:
4310 those that are passed in the argument to log_write(), and those that are
4311 only ever tested independently outside of that function. These are now in
4312 separate 32-bit words, so there is plenty of room for expansion again.
4313 There is no change in the user interface or the logging behaviour.
4315 (b) When building, for example, log lines, the code previously used a
4316 macro that called string_cat() twice, in order to add two strings. This is
4317 not really sufficiently general. Furthermore, there was one instance where
4318 it was actually wrong because one of the argument was used twice, and in
4319 one call a function was used. (As it happened, calling the function twice
4320 did not affect the overall behaviour.) The macro has been replaced by a
4321 function that can join an arbitrary number of extra strings onto a growing
4324 (c) The code for expansion conditions now uses a table and a binary chop
4325 instead of a serial search (which was left over from when there were very
4326 few conditions). Also, it now recognizes conditions like "pam" even when
4327 the relevant support is not compiled in: a suitably worded error message is
4328 given if an attempt is made to use such a condition.
4330 11. Added ${time_interval:xxxxx}.
4332 12. A bug was causing one of the ddress fields not to be passed back correctly
4333 from remote delivery subprocesses. The field in question was not being
4334 subsequently used, so this caused to problems in practice.
4336 13. Added new log selectors queue_time and deliver_time.
4338 14. Might have fixed a bug in maildirsizefile handling that threw up
4339 "unexpected character" debug warnings, and recalculated the data
4340 unnecessarily. In any case, I expanded the warning message to give more
4343 15. Added the message "Restricted characters in address" to the statements in
4344 the default ACL that block characters like @ and % in local parts.
4346 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4347 Three changes have been made:
4349 (a) There was a serious bug; a negative response to MAIL caused the whole
4350 recipient domain to be cached as invalid, thereby blocking all messages
4351 to all local parts at the same domain, from all senders. This bug has
4352 been fixed. The domain is no longer cached after a negative response to
4353 MAIL if the sender used is not empty.
4355 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4358 (c) A new callout option, "use_sender" has been added for people who want
4359 the modified behaviour.
4365 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4368 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4369 indeed breaks things for older releases.
4371 3. Added additional logging to the case where there is a problem reading data
4372 from a filter that is running in a subprocess using a pipe, in order to
4373 try to track down a specific problem.
4375 4. Testing facility fudge: when running in the test harness and attempting
4376 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4377 getting "No route to host". Convert this to a timeout.
4379 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4382 6. Some OS don't have socklen_t but use size_t instead. This affects the
4383 fifth argument of getsockopt() amongst other things. This is now
4384 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4385 can be set for individual OS. I have set it for SunOS5, OSF1, and
4386 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4387 some earlier ones do not.
4389 7. Change 4.30/15 was not doing the test caselessly.
4391 8. The standard form for an IPv6 address literal was being rejected by address
4392 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4393 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4394 this, as well as the form without the "IPv6" on the front (but only when
4395 address literals are enabled, of course).
4397 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4399 10. Exim crashed if a message with an empty sender address specified by -f
4400 encountered a router with an errors_to setting. This could be provoked only
4401 by a command such as
4405 where an empty string was supplied; "<>" did not hit this bug.
4407 11. Installed PCRE release 4.5.
4409 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4410 remained set. It is now erased.
4412 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4413 times from message ids (which are base 36 rather than the normal 62).
4415 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4416 were being counted as actual protocol errors, and logged if the log
4417 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4418 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4419 and DATA following a set of rejected RCPTs do not count as protocol errors.
4420 In other words, Exim assumes they were pipelined, though this may not
4421 actually be the case. Of course, in all cases the client gets an
4422 appropriate error code.
4424 15. If a lookup fails in an ACL condition, a message about the failure may
4425 be available; it is used if testing the ACL cannot continue, because most
4426 such messages specify what the cause of the deferral is. However, some
4427 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4428 that caused an old message to be retained and used if a later statement
4429 caused a defer, replacing the real cause of the deferral.
4431 16. If an IP address had so many PTR records that the DNS lookup buffer
4432 was not large enough to hold them, Exim could crash while trying to process
4433 the truncated data. It now detects and logs this case.
4435 17. Further to 4.21/58, another change has been made: if (and only if) the
4436 first line of a message (the first header line) ends with CRLF, a bare LF
4437 in a subsequent header line has a space inserted after it, so as not to
4438 terminate the header.
4440 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4441 unnecessarily, used atoi() instead of strtol(), and didn't check the
4442 termination when getting file sizes from file names by regex.
4444 19. Completely re-implemented the support for maildirsize files, in the light
4445 of a number of problems with the previous contributed implementation
4446 (4.30/29). In particular:
4448 . If the quota is zero, the maildirsize file is maintained, but no quota is
4451 . If the maildir directory does not exist, it is created before any attempt
4452 to write a maildirsize file.
4454 . The quota value in the file is just a cache; if the quota is changed in
4455 the transport, the new value overrides.
4457 . A regular expression is available for excluding directories from the
4460 20. The autoreply transport checks the characters in options that define the
4461 message's headers; it allows continued headers, but it was checking with
4462 isspace() after an embedded newline instead of explicitly looking for a
4465 21. If all the "regular" hosts to which an address was routed had passed their
4466 expiry times, and had not reached their retry times, the address was
4467 bounced, even if fallback hosts were defined. Now Exim should go on to try
4470 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4471 equivalent code in the SMTP transport. Some hosts send humungous responses
4472 to HELO/EHLO, more than 1024 it seems.
4474 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4475 gives compiler warnings in some environments. I've now done it "properly",
4478 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4479 (because of problems with the built-in one) was declared to return uschar *
4480 instead of char *, causing compiler failure.
4482 25. Fixed a file descriptor leak when processing alias/forward files.
4484 26. Fixed a minor format string issue in dbfn.c.
4486 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4488 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4489 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4492 29. When a sender address is verified, it is cached, to save repeating the test
4493 when there is more than one recipient in a message. However, when the
4494 verification involves a callout, it is possible for different callout
4495 options to be set for different recipients. It is too complicated to keep
4496 track of this in the cache, so now Exim always runs a verification when a
4497 callout is required, relying on the callout cache for the optimization.
4498 The overhead is duplication of the address routing, but this should not be
4501 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4502 to be verified with callout=postmaster, and the main callout worked but the
4503 postmaster check failed, the verification correctly failed. However, if a
4504 subsequent RCPT command asked for sender verification *without* the
4505 postmaster check, incorrect caching caused this verification also to fail,
4508 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4509 it was not caching the DNS options (qualify_single, search_parents) that
4510 were used when the lookup failed. A subsequent lookup with different
4511 options therefore always gave the same answer, though there were cases
4512 where it should not have. (Example: a "domains = !$mx_any" option on a
4513 dnslookup router: the "domains" option is always processed without any
4514 widening, but the router might have qualify_single set.) Now Exim uses the
4515 cached value only when the same options are set.
4517 32. Added John Jetmore's "exipick" utility to the distribution.
4519 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4520 than a timeout (e.g. a certificate is required, and is not provided), an
4521 Exim server now closes the connection immediately. Previously it waited for
4522 the client to close - but if the client is SSL, it seems that they each
4523 wait for each other, leading to a delay before one of them times out.
4525 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4526 maintained 0.8.x compatibility because I don't think many are using it, and
4527 it is clearly obsolete.
4529 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4532 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4533 client certificate was expired. A simple patch fixes this, though I don't
4534 understand the full logic of why the verify callback is called multiple
4537 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4538 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4539 which causes problems with some clients (such as the Certicom SSL Plus
4540 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4541 disables the coutermeasure allowing Eudora to connect."
4543 38. Exim was not checking that a write() to a log file succeeded. This could
4544 lead to Bad Things if a log got too big, in particular if it hit a file
4545 size limit. Exim now panics and dies if it cannot write to a log file, just
4546 as it does if it cannot open a log file.
4548 39. Modified OS/Makefile-Linux so that it now contains
4550 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4552 The two -D definitions ensure that Exim is compiled with large file
4553 support, which makes it possible to handle log files that are bigger than
4556 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4557 instance) a domain was checked against a named list that involved a lookup,
4558 causing $domain_data to be set, then another domain was checked against the
4559 same list, then the first domain was re-checked, the value of $domain_data
4560 after the final check could be wrong. In particular, if the second check
4561 failed, it could be set empty. This bug probably also applied to
4564 41. The strip_trailing_dot option was not being applied to the address given
4565 with the -f command-line option.
4567 42. The code for reading a message's header from the spool was incrementing
4568 $received_count, but never initializing it. This meant that the value was
4569 incorrect (doubled) while delivering a message in the same process in which
4570 it was received. In the most common configuration of Exim, this never
4571 happens - a fresh exec is done - but it can happen when
4572 deliver_drop_privilege is set.
4574 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4575 it now includes up to 150 characters of the unexpected data in the log
4578 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4579 and building data strings. The size of both of these buffers was 10 000
4580 bytes - far larger than anybody would *ever* want, thought I. Needless to
4581 say, somebody hit the limit. I have increased the maximum line length to
4582 20 000 and the maximum data length of concatenated lines to 100 000. I have
4583 also fixed two bugs, because there was no checking on these buffers. Tsk,
4584 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4585 buffer is too small.
4587 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4588 lsearch lookups. Now it does.
4590 46. When parsing a route_list item in a manualroute router, a fixed-length
4591 buffer was used for the list of hosts. I made this 1024 bytes long,
4592 thinking that nobody would ever have a list of hosts that long. Wrong.
4593 Somebody had a whole pile of complicated expansion conditions, and the
4594 string was silently truncated, leading to an expansion error. It turns out
4595 that it is easier to change to an unlimited length (owing to other changes
4596 that have happened since this code was originally written) than to build
4597 structure for giving a limitation error. The length of the item that
4598 expands into the list of hosts is now unlimited.
4600 47. The lsearch lookup could not handle data where the length of text line was
4601 more than 4095 characters. Such lines were truncated, leading to shortened
4602 data being returned. It should now handle lines of any length.
4604 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4605 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4608 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4609 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4610 no longer happens in comment lines. A list of replacements is now placed
4611 at the head of all of the source files, except those whose only change is
4612 to replace PERL_COMMAND in the very first #! line.
4614 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4615 messages on the queue, with a bottom-up merge sort, using code contributed
4616 by Michael Haardt. This should make operations like -bp somewhat faster on
4617 large queues. It won't affect queue runners, except when queue_run_in_order
4620 51. Installed eximstats 1.31 in the distribution.
4622 52. Added support for SRV lookups to the dnslookup router.
4624 53. If an ACL referred to $message_body or $message_body_end, the value was not
4625 reset for any messages that followed in the same SMTP session.
4627 54. The store-handling optimization for building very long strings was not
4628 differentiating between the different store pools. I don't think this
4629 actually made any difference in practice, but I've tidied it.
4631 55. While running the routers to verify a sender address, $sender_address
4632 was still set to the sender address. This is wrong, because when routing to
4633 send a bounce to the sender, it would be empty. Therefore, I have changed
4634 it so that, while verifying a sender address, $sender_address is set to <>.
4635 (There is no change to what happens when verifying a recipient address.)
4637 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4638 target A or AAAA records (if not already returned) without resetting the
4639 qualify_single or search_parents options of the DNS resolver. These are
4640 inappropriate in this case because the targets of MX and SRV records must
4641 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4642 target that, when qualified, matched something in the local domain. These
4643 two options are now turned off when doing these lookups.
4645 57. It seems that at least some releases of Reiserfs (which does not have the
4646 concept of a fixed number of inodes) returns zero and not -1 for the
4647 number of available inodes. This interacted badly with check_spool_inodes,
4648 which assumed that -1 was the "no such thing" setting. What I have done is
4649 to check that the total number of inodes is greater than zero before doing
4650 the test of how many are available.
4652 58. When a "warn" ACL statement has a log_message modifier, the message is
4653 remembered, and not repeated. This is to avoid a lot of repetition when a
4654 message has many recipients that cause the same warning to be written.
4655 Howewer, Exim was preserving the list of already written lines for an
4656 entire SMTP session, which doesn't seem right. The memory is now reset if a
4657 new message is started.
4659 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4660 debugging output unless log_rewrite was also set.
4662 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4663 of (int)(handle) when we know that handle contains (void *)(-1).
4665 61. The Exim daemon panic-logs an error return when it closes the incoming
4666 connection. However "connection reset by peer" seems to be common, and
4667 isn't really an error worthy of noting specially, so that particular error
4670 62. When Exim is trying to find all the local interfaces, it used to panic and
4671 die if the ioctl to get the interface flags failed. However, it seems that
4672 on at least one OS (Solaris 9) it is possible to have an interface that is
4673 included in the list of interfaces, but for which you get a failure error
4674 for this call. This happens when the interface is not "plumbed" into a
4675 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4676 failure of the "get flags" call assumes that the interface is down.
4678 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4679 makes life easier for people who are doing arithmetic on fields extracted
4680 from dates, where you often get leading zeros that should not be
4681 interpreted as octal.
4683 64. Added qualify_domain to the redirect router, to override the global
4686 65. If a pathologically long header line contained very many addresses (the
4687 report of this problem mentioned 10 000) and each of them was rewritten,
4688 Exim could use up a very large amount of memory. (It kept on making new
4689 copies of the header line as it rewrote, and never released the old ones.)
4690 At the expense of a bit more processing, the header rewriting function has
4691 been changed so that it no longer eats memory in this way.
4693 66. The generation of the Received: header has been moved from the time that a
4694 message starts to be received, to the time that it finishes. The timestamp
4695 in the Received: header should now be very close to that of the <= log
4696 line. There are two side-effects of this change:
4698 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4699 logged header lines no longer include the local Received: line, because
4700 it has not yet been created. The same applies to a copy of the message
4701 that is returned to a non-SMTP sender when a message is rejected.
4703 (b) When a filter file is tested using -bf, no additional Received: header
4704 is added to the test message. After some thought, I decided that this
4707 This change does not affect the value of $received_for. It is still set
4708 after address rewriting, but before local_scan() is called.
4710 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4712 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4713 gave an unhelpful panic error message, and a defer error. I have managed to
4714 change this behaviour so that it now rejects any supplied certificate,
4715 which seems right, as the list of acceptable certificates is empty.
4717 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4718 gave an unhelpful defer error. I have not managed to make this reject any
4719 supplied certificates, but the error message it gives is "no certificate
4720 supplied", which is not helpful.
4722 70. exigrep's output now also includes lines that are not associated with any
4723 message, but which match the given pattern. Implemented by a patch from
4724 Martin Sluka, which also tidied up the Perl a bit.
4726 71. Recipient callout verification, like sender verification, was using <> in
4727 the MAIL FROM command. This isn't really the right thing, since the actual
4728 sender may affect whether the remote host accepts the recipient or not. I
4729 have changed it to use the actual sender in the callout; this means that
4730 the cache record is now keyed on a recipient/sender pair, not just the
4731 recipient address. There doesn't seem to be a real danger of callout loops,
4732 since a callout by the remote host to check the sender would use <>.
4733 [SEE ABOVE: changed after hitting problems.]
4735 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4736 temporary errors. However, in the case of such a code being given after
4737 the end of a data transmission (i.e. after ".") Exim was failing to write
4738 a retry record for the message. (Yes, there was some broken host that was
4739 actually sending 8xx at this point.)
4741 73. An unknown lookup type in a host list could cause Exim to panic-die when
4742 the list was checked. (An example that provoked this was putting <; in the
4743 middle of a list instead of at the start.) If this happened during a DATA
4744 ACL check, a -D file could be left lying around. This kind of configuration
4745 error no longer causes Exim to die; instead it causes a defer errror. The
4746 incident is still logged to the main and panic logs.
4748 74. Buglet left over from Exim 3 conversion. The message "too many messages
4749 in one connection" was written to the rejectlog but not the mainlog, except
4750 when address rewriting (yes!) was being logged.
4752 75. Added write_rejectlog option.
4754 76. When a system filter was run not as root (that is, when system_filter_user
4755 was set), the values of the $n variables were not being returned to the
4756 main process; thus, they were not subsequently available in the $sn
4759 77. Added +return_path_on_delivery log selector.
4761 78. A connection timeout was being treated differently from recipients deferred
4762 when testing hosts_max_try with a message that was older than the host's
4763 retry timeout. (The host should not be counted, thus allowing all hosts to
4764 be tried at least once before bouncing.) This may have been the cause of an
4765 occasionally reported bug whereby a message would remain on the queue
4766 longer than the retry timeout, but would be bounced if a delivery was
4767 forced. I say "may" because I never totally pinned down the problem;
4768 setting up timeout/retry tests is difficult. See also the next item.
4770 79. The ultimate address timeout was not being applied to errors that involved
4771 a combination of host plus message (for example, a timeout on a MAIL
4772 command). When an address resolved to a number of possible hosts, and they
4773 were not all tried for each delivery (e.g. because of hosts_max_try), a
4774 message could remain on the queue longer than the retry timeout.
4776 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4779 81. Fixed an obscure SMTP outgoing bug which required at least the following
4780 conditions: (a) there was another message waiting for the same server;
4781 (b) the server returned 5xx to all RCPT commands in the first message so
4782 that the message was not completed; (c) the server dropped the connection
4783 or gave a negative response to the RSET that Exim sends to abort the
4784 transaction. The observed case was a dropped connection after DATA that had
4785 been sent in pipelining mode. That is, the server had advertised PIPELINING
4786 but was not implementing it correctly. The effect of the bug was incorrect
4787 behaviour, such as trying another host, and this could lead to a crash.
4793 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4794 and daemon.c were passed as pointers to ints; they should have been
4795 pointers to socklen_t variables (which are typically unsigned ints).
4797 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4800 3. Fixed a really odd bug that affected only the testing scheme; patching a
4801 certain fixed string in the binary changed the value of another string that
4802 happened to be identical to the end of the original first string.
4804 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4805 name", it returns that address as the IP address. On some operating
4806 systems (e.g. Solaris), it also passes back the IP address string as the
4807 "host name". However, on others (e.g. Linux), it passes back an empty
4808 string. Exim wasn't checking for this, and was changing the host name to an
4809 empty string, assuming it had been canonicized.
4811 5. Although rare, it is permitted to have more than one PTR record for a given
4812 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4813 all the names associated with an address, because they do in Solaris.
4814 However, it seems that they do not in Linux for data that comes from the
4815 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4816 I found this out when I moved to a new Linux workstation and tried to run
4817 the Exim test suite.
4819 To get round this problem I have changed the code so that it now does its
4820 own call to the DNS to look up PTR records when searching for a host name.
4821 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4822 addresses that are only in /etc/hosts are still found.
4824 This behaviour is, however, controlled by an option called host_lookup_
4825 order, which defaults to "bydns:byaddr". If people want to use the other
4826 order, or indeed, just use one or the other means of lookup, they can
4827 specify it in this variable.
4829 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4830 some operating systems, this comes back from gethostbyaddr() as an empty
4831 string, and this is what Exim used to test for. However, it seems that in
4832 other systems, "." is yielded. Exim now tests for this case too.
4834 7. The values of check_spool_space and check_log_space are now held internally
4835 as a number of kilobytes instead of an absolute number of bytes. If a
4836 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4837 kilobyte. This means that much larger values can be stored.
4839 8. Exim monitor: an attempt to get the action menu when not actually pointing
4840 at a message produces an empty menu entitled "No message selected". This
4841 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4842 no entries in it ("Shell widget menu has zero width and/or height"). So I
4843 have added a single, blank menu entry in this case.
4845 9. Added ${quote_local_part.
4847 10. MIME decoding is now applied to the contents of Subject: header lines when
4850 11. Now that a reference to $sender_host_address automatically causes a reverse
4851 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4852 host lookup before query-style lookups in lists that might use this
4853 variable. This has therefore been abolished, and the "net-" prefix is no
4854 longer necessary for query-style lookups.
4856 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4857 have been a typo for LFLAGS, so it has been changed.
4859 13. The install script calls Exim with "-C /dev/null" in order to find the
4860 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4861 to be output. Howeve, since Exim outputs its version number before the
4862 error, it didn't break the script. It just looked ugly. I fixed this by
4863 always allowing "-C /dev/null" if the caller is root.
4865 14. Ignore overlarge ACL variable number when reading spool file - insurance
4866 against a later release with more variables having written the file.
4868 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4869 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4870 this, as well as the form without the "IPv6" on the front.
4872 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4873 OS/Makefile-Darwin file.
4875 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4876 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4878 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4879 variable. However, some OS use ints for inodes. I've added cast to long int
4880 to get rid of the compiler warning.
4882 19. I had forgotten to lock out "/../" in configuration file names when
4883 ALT_CONFIG_PREFIX was set.
4885 20. Routers used for verification do not need to specify transports. However,
4886 if such a router generated a host list, and callout was configured, Exim
4887 crashed, because it could not find a port number from the (non-existent)
4888 transport. It now assumes port 25 in this circumstance.
4890 21. Added the -t option to exigrep.
4892 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4893 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4894 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4895 them is set, LOOKUP_LSEARCH is forced.
4897 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4899 24. Added sender and host information to the "rejected by local_scan()" log
4900 line; previously there was no indication of these.
4902 25. Added .include_if_exists.
4904 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4905 Linux. It turns out that not all file systems support this. Apparently some
4906 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4907 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4908 not supported on the file descriptor, is now ignored when Exim is trying to
4909 sync a directory. This applies only to Linux.
4911 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4913 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4914 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4917 29. Added support for maildirsize files from supplied patch (modified a bit).
4919 30. The use of :fail: followed by an empty string could lead Exim to respond to
4920 sender verification failures with (e.g.):
4922 550 Verification failed for <xxx>
4923 550 Sender verify failed
4925 where the first response line was missing the '-' that indicates it is not
4926 the final line of the response.
4928 31. The loop for finding the name of the user that called Exim had a hardwired
4929 limit of 10; it now uses the value of finduser_retries, which is used for
4930 all other user lookups.
4932 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4935 33. Exim was neglecting to zero errno before one call of strtol() when
4936 expanding a string and expecting an integer value. On some systems this
4937 resulted in spurious "integer overflow" errors. Also, it was casting the
4938 result into an int without checking.
4940 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4941 did not work. The code looks as if it has *never* worked, though it appears
4942 to have been documented since at least releast 1.62. I have made it work.
4944 35. The "timeout_DNS" error in retry rules, also documented since at least
4945 1.62, also never worked. As it isn't clear exactly what this means, and
4946 clearly it isn't a major issue, I have abolished the feature by treating it
4947 as "timeout", and writing a warning to the main and panic logs.
4949 36. The display of retry rules for -brt wasn't always showing the error code
4952 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4953 timeout_connect_A, timeout_connect_MX.
4955 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4956 to the empty sender.
4958 39. The daemon was not analysing the content of -oX till after it had closed
4959 stderr and disconnected from the controlling terminal. This meant that any
4960 syntax errors were only noted on the panic log, and the return code from
4961 the command was 0. By re-arranging the code a little, I've made the
4962 decoding happen first, so such errors now appear on stderr, and the return
4963 code is 1. However, the actual setting up of the sockets still happens in
4964 the disconnected process, so errors there are still only recorded on the
4967 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4968 connections (as happens on some IP stacks) was logged at start up time as
4969 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4970 it from "IPv6 and IPv4", which means that two separate sockets are being
4973 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4974 says whether AF_INET or AF_INET6 was passed as an argument.
4976 42. Exiwhat output was messed up when time zones were included in log
4979 43. Exiwhat now gives more information about the daemon's listening ports,
4980 and whether -tls-on-connect was used.
4982 44. The "port" option of the smtp transport is now expanded.
4984 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4985 silently ignored. Now an error message is written to the main and panic
4988 46. There's a new ACL modifier called "logwrite" which writes to a log file
4989 as soon as it is encountered.
4991 47. Added $local_user_uid and $local_user_gid at routing time.
4993 48. Exim crashed when trying to verify a sender address that was being
4996 49. Exim was recognizing only a space character after ".include". It now also
4997 recognizes a tab character.
4999 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5000 extracting the relevant information from the specification. The man page no
5001 longer contains scrambled data for the -d option, and I've added a section
5002 at the front about calling Exim under different names.
5004 51. Added "extra_headers" argument to the "mail" command in filter files.
5006 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5009 53. Installed eximstats 1.29.
5011 54. Added transport_filter_timeout as a generic transport option.
5013 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5014 Cc: header lines. This was required by RFC 822, but it not required by RFC
5017 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5018 incoming messages that did not have them. Now it does so only if the
5019 message originates locally, that is, if there is no associated remote host
5020 address. When Resent- header lines are present, this applies to the Resent-
5021 lines rather than the non-Resent- lines.
5023 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5024 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5026 58. Messages for configuration errors now include the name of the main
5027 configuration file - useful now that there may be more than one file in a
5028 list (.included file names were always shown).
5030 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5031 for those rare installations that do not start the daemon as root or run it
5032 setuid root. I've cut out the call to initgroups() if the daemon is not
5035 60. The Exim user and group can now be bound into the binary as text strings
5036 that are looked up at the start of Exim's processing.
5038 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5040 62. Added $mailstore_basename variable.
5042 63. Installed patch to sieve.c from Michael Haardt.
5044 64. When Exim failed to open the panic log after failing to open the main log,
5045 the original message it was trying to log was written to stderr and debug
5046 output, but if they were not available (the usual case in production), it
5047 was lost. Now it is written to syslog before the two lines that record the
5048 failures to open the logs.
5050 65. Users' Exim filters run in subprocesses under the user's uid. It is
5051 possible for a "deliver" command or an alias in a "personal" command to
5052 provoke an address rewrite. If logging of address rewriting is configured,
5053 this fails because the process is not running as root or exim. There may be
5054 a better way of dealing with this, but for the moment (because 4.30 needs
5055 to be released), I have disabled address rewrite logging when running a
5056 filter in a non-root, non-exim process.
5062 1. The buildconfig auxiliary program wasn't quoting the value set for
5063 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5064 not defined. This bug was masked in 4.22 by the effect that was fixed in
5067 2. Some messages that were rejected after a message id was allocated were
5068 shown as "incomplete" by exigrep. It no longer does this for messages that
5069 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5071 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5072 have allow_domain_literals set, the ID did not get logged in the <= line.
5073 Domain literals are now always recognized in Message-ID: header lines.
5075 4. The first argument for a ${extract expansion item is the key name or field
5076 number. Leading and trailing spaces in this item were not being ignored,
5077 causing some misleading effects.
5079 5. When deliver_drop_privilege was set, single queue runner processes started
5080 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5081 same command in the process it spins off) were not dropping privilege.
5083 6. When the daemon running as "exim" started a queue runner, it always
5084 re-executed Exim in the spun-off process. This is a waste of effort when
5085 deliver_drop_privilege is set. The new process now just calls the
5086 queue-runner function directly.
5092 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5095 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5096 ignored. Though the use of -oP was forcing the writing of a pid file, it
5097 was always written to the default place.
5099 3. If the message "no IP address found for host xxxx" is generated during
5100 incoming verification, it is now followed by identification of the incoming
5101 connection (so you can more easily find what provoked it).
5103 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5105 5. Added some features to "harden" Exim a bit more against certain attacks:
5107 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5108 be put in Local/Makefile. This is like the never_users runtime option,
5109 but it cannot be overridden. The default setting is "root".
5111 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5112 prefix string with which any file named in a -C command line option
5115 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5116 is retained for -C and -D only if the caller of Exim is root. Without
5117 it, the exim user may also use -C and -D and retain privilege.
5119 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5120 command line option is disabled.
5122 6. Macro names set by the -D option must start with an upper case letter, just
5123 like macro names defined in the configuration file.
5125 7. Added "dereference=" facility to LDAP.
5127 8. Two instances of the typo "uknown" in the source files are fixed.
5129 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5130 the Configure-Makefile script screwed up while processing it.
5132 10. Incorporated PCRE 4.4.
5134 11. The SMTP synchronization check was not operating right at the start of an
5135 SMTP session. For example, it could not catch a HELO sent before the client
5136 waited for the greeting. There is now a check for outstanding input at the
5137 point when the greeting is written. Because of the duplex, asynchronous
5138 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5139 way, but not yet received, when the check is performed.
5141 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5142 on TCP/IP sockets, because this apparently causes some broken clients to
5145 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5146 unchanged) from the Cygwin maintainer.
5148 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5149 was supported instead of testing for mbx. Effectively a typo.
5151 15. The spa authenticator server code was not checking that the input it
5152 received was valid base64.
5154 16. The debug output line for the "set" modifier in ACLs was not showing the
5155 name of the variable that was being set.
5157 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5159 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5160 lookup on its own. Something else was needed to trigger the lookup. For
5161 example, a match in host_lookup or the need for a host name in a host list.
5162 Now, if $sender_host_name is referenced and the host name has not yet been
5163 looked up, a lookup is performed. If the lookup fails, the variable remains
5164 empty, and $host_lookup_failed is set to "1".
5166 19. Added "eqi" as a case-independent comparison operator.
5168 20. The saslauthd authentication condition could segfault if neither service
5169 nor realm was specified.
5171 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5172 error message that was logged was misleading, and incoming SMTP
5173 connections were dropped. The message is now more accurate, and temporary
5174 errors are given to SMTP connections.
5176 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5177 (except RSET) with a 421 error, until QUIT is received. However, it was
5178 failing to send a response to QUIT.
5180 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5181 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5182 ACL is now run after the helo_try_verify_hosts code.
5184 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5185 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5186 case-independent, so other case variants are also recognized). Apparently
5187 some systems use these upper case variants.
5189 25. If more than two messages were waiting for the same host, and a transport
5190 filter was specified for the transport, Exim sent two messages over the
5191 same TCP/IP connection, and then failed with "socket operation on non-
5192 socket" when it tried to send the third.
5194 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5196 27. The extern definition of crypt16() in expand.c was not being excluded when
5197 the OS had its own crypt16() function.
5199 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5200 as a preferred synonym for return_size_limit, both as an option and as an
5203 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5205 30. Changed the default configuration ACL to relax the local part checking rule
5206 for addresses that are not in any local domains. For these addresses,
5207 slashes and pipe symbols are allowed within local parts, but the sequence
5208 /../ is explicitly forbidden.
5210 31. SPA server authentication was not clearing the challenge buffer before
5213 32. log_message in a "warn" ACL statement was writing to the reject log as
5214 well as to the main log, which contradicts the documentation and doesn't
5215 seem right (because no rejection is happening). So I have stopped it.
5217 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5218 However, I am unable to do any testing of this.
5220 34. Fixed an infelicity in the appendfile transport. When checking directories
5221 for a mailbox, to see if any needed to be created, it was accidentally
5222 using path names with one or more superfluous leading slashes; tracing
5223 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5225 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5226 contents are added to the log line that is written for every discarded
5227 recipient. (Previously a log_message setting was ignored.)
5229 36. The ${quote: operator now quotes the string if it is empty.
5231 37. The install script runs exim in order to find its version number. If for
5232 some reason other than non-existence or emptiness, which it checks, it
5233 could not run './exim', it was installing it with an empty version number,
5234 i.e. as "exim-". This error state is now caught, and the installation is
5237 38. An argument was missing from the function that creates an error message
5238 when Exim fails to connect to the socket for saslauthd authentication.
5239 This could cause Exim to crash, or give a corrupted message.
5241 39. Added isip, isip4, and isip6 to ${if conditions.
5243 40. The ACL variables $acl_xx are now saved with the message, and can be
5244 accessed later in routers, transports, and filters.
5246 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5247 strings in the file are not string-expanded.
5249 42. If a MAIL command specified a SIZE value that was too large to fit into an
5250 int variable, the check against message_size_limit failed. Such values are
5251 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5252 day this will have to be increased, but I don't think I want to be around
5253 when emails are that large.
5260 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5261 iconv() is not standard in FreeBSD.
5263 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5264 IPv6 enabled. The observed symptom was a segmentation fault on return from
5265 the function os_common_find_running_interfaces() in src/os.c.
5267 3. In the check_special_case() function in daemon.c I had used "errno" as an
5268 argument name, which causes warnings on some systems. This was basically a
5269 typo, since it was named "eno" in the comments!
5271 4. The code that waits for the clock to tick (at a resolution of some fraction
5272 of a second) so as to ensure message-id uniqueness was always waiting for
5273 at least one whole tick, when it could have waited for less. [This is
5274 almost certainly not relevant at current processor speeds, where it is
5275 unlikely to ever wait at all. But we try to future-proof.]
5277 5. The function that sleeps for a time interval that includes fractions of a
5278 second contained a race. It did not block SIGALRM between setting the
5279 timer, and suspending (a couple of lines later). If the interval was short
5280 and the sigsuspend() was delayed until after it had expired, the suspension
5281 never ended. On busy systems this could lead to processes getting stuck for
5284 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5285 process, before it forks any delivery processes. The open lookup caching
5286 mechanism meant that the open file or database connection was passed into
5287 the delivery process. The problem was that delivery processes always tidy
5288 up cached lookup data. This could cause a problem for the next delivery
5289 process started by the queue runner, because the external queue runner
5290 process does not know about the closure. So the next delivery process
5291 still has data in the lookup cache. In the case of a file lookup, there was
5292 no problem because closing a file descriptor in a subprocess doesn't affect
5293 the parent. However, if the lookup was caching a connection to a database,
5294 the connection was closed, and the second delivery process was likely to
5295 see errors such as "PGSQL: query failed: server closed the connection
5296 unexpectedly". The problem has been fixed by closing all cached lookups
5297 in a queue runner before running a delivery process.
5299 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5300 seem to have the "const" qualifier which it has on other OS. I've
5303 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5304 *of the same type* (client or server) with the same public name that an
5305 error should be diagnosed.
5307 9. When Exim looked up a host name for an IP address, but failed to find the
5308 original IP address when looking up the host name (a safety check), it
5309 output the message "<ip address> does not match any IP for NULL", which was
5310 confusing, to say the least. The bug was that the host name should have
5311 appeared instead of "NULL".
5313 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5314 user that is built into the binary, and the -C or -D options is used, root
5315 privilege is dropped before the configuration file is read. In addition,
5316 logging is switched to stderr instead of the normal log files. If the
5317 configuration then re-defines the Exim user, the unprivileged environment
5318 is probably not what is expected, so Exim logs a panic warning message (but
5321 However, if deliver_drop_privilege is set, the unprivileged state may well
5322 be exactly what is intended, so the warning has been cut out in that case,
5323 and Exim is allowed to try to write to its normal log files.
5329 1. smtp_return_error_details was not giving details for temporary sender
5330 or receiver verification errors.
5332 2. Diagnose a configuration error if two authenticators have the same public
5335 3. Exim used not to create the message log file for a message until the first
5336 delivery attempt. This could be confusing when incoming messages were held
5337 for policy or load reasons. The message log file is now created at the time
5338 the message is received, and an initial "Received" line is written to it.
5340 4. The automatically generated man page for command line options had a minor
5341 bug that caused no ill effects; however, a more serious problem was that
5342 the procedure for building the man page automatically didn't always
5343 operate. Consequently, release 4.20 contains an out-of-date version. This
5344 shouldn't happen again.
5346 5. When building Exim with embedded Perl support, the script that builds the
5347 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5348 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5350 6. The freeze_tell option was not being used for messages that were frozen on
5351 arrival, either by an ACL or by local_scan().
5353 7. Added the smtp_incomplete_transaction log selector.
5355 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5356 was accepting AUTH without a new EHLO.
5358 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5359 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5362 10. Logging of TCP/IP connections (when configured) now happens in the main
5363 daemon process instead of the child process, so that the TCP/IP connection
5364 count is more accurate (but it can never be perfect).
5366 11. The use of "drop" in a nested ACL was not being handled correctly in the
5367 outer ACL. Now, if condition failure induced by the nested "drop" causes
5368 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5369 or "require"), the connection is dropped.
5371 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5372 that yield "discard" can now be used with an "accept" or a "discard" verb,
5373 but an error is generated for any others (because I can't see a useful way
5374 to define what should happen).
5376 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5377 are now processed in the same way as lines in the Exim configuration file.
5378 In particular, continuation lines are supported.
5380 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5382 15. Added -ti meaning -t -i.
5384 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5385 domains, and warn by logging if others are found.
5387 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5388 returned from GIFCONF when Exim is trying to find the list of interfaces on
5389 a host. The code in os.c has been modified to copy each ifreq to an aligned
5390 structure in all cases.
5392 Also, in some cases, the returned ifreq's were being copied to a 'struct
5393 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5394 means the last couple of bytes of an IPv6 address could be chopped if the
5395 ifreq contained only a normal sockaddr (14 bytes storage).
5397 18. Named domain lists were not supported in the hosts_treat_as_local option.
5398 An entry such as +xxxx was not recognized, and was treated as a literal
5401 19. Ensure that header lines added by a DATA ACL are included in the reject log
5402 if the ACL subsequently rejects the message.
5404 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5405 MD5 (which is deprecated).
5407 21. When testing a filter file using -bf, Exim was writing a message when it
5408 took the sender from a "From " line in the message, but it was not doing so
5409 when it took $return_path from a Return-Path: header line. It now does.
5411 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5412 with a valid header line field name (a series of printing characters
5413 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5415 23. Changed "disc" in the source to "disk" to conform to the documentation and
5416 the book and for uniformity.
5418 24. Ignore Sendmail's -Ooption=value command line item.
5420 25. When execve() failed while trying to run a command in a pipe transport,
5421 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5422 could be confused with a return value of 69 from the command itself. This
5423 has been changed to 127, the value the shell returns if it is asked to run
5424 a non-existent command. The wording for the related log line suggests a
5425 non-existent command as the problem.
5427 26. If received_header_text expands to an empty string, do not add a Received:
5428 header line to the message. (Well, it adds a token one on the spool, but
5429 marks it "old" so that it doesn't get used or transmitted.)
5431 27. Installed eximstats 1.28 (addition of -nt option).
5433 28. There was no check for failure on the call to getsockname() in the daemon
5434 code. This can fail if there is a shortage of resources on the system, with
5435 ENOMEM, for example. A temporary error is now given on failure.
5437 29. Contrary to the C standard, it seems that in some environments, the
5438 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5439 program. Exim now does this explicitly; it affects the formatting of
5440 timestamps using strftime().
5442 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5443 complaints. I've now initialized all the variables, to avoid this.
5445 32. Header lines added by a system filter were not being "seen" during
5446 transport-time rewrites.
5448 33. The info_callback() function passed to OpenSSL is set up with type void
5449 (*)(SSL *, int, int), as described somewhere. However, when calling the
5450 function (actually a macro) that sets it up, the type void(*)() is
5451 expected. I've put in a cast to prevent warnings from picky compilers.
5453 34. If a DNS black list lookup found a CNAME record, but there were no A
5454 records associated with the domain it pointed at, Exim crashed.
5456 35. If a DNS black list lookup returned more than one A record, Exim ignored
5457 all but the first. It now scans all returned addresses if a particular IP
5458 value is being sought. In this situation, the contents of the
5459 $dnslist_value variable are a list of all the addresses, separated by a
5462 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5463 to accept a host name and all its aliases if the forward lookup for any of
5464 them yielded the IP address of the incoming connection. Now it accepts only
5465 those names whose forward lookup yields the correct IP address. Any other
5466 names are discarded. This closes a loophole whereby a rogue DNS
5467 administrator could create reverse DNS records to break through a
5468 wildcarded host restriction in an ACL.
5470 37. If a user filter or a system filter that ran in a subprocess used any of
5471 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5472 the wrong values were passed to the pipe command ($thisaddress had the
5473 value of $0, $0 had the value of $1, etc). This bug was introduced by
5474 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5476 38. Improved the line breaking for long SMTP error messages from ACLs.
5477 Previously, if there was no break point between 40 and 75 characters, Exim
5478 left the rest of the message alone. Two changes have been made: (a) I've
5479 reduced the minimum length to 35 characters; (b) if it can't find a break
5480 point between 35 and 75 characters, it looks ahead and uses the first one
5481 that it finds. This may give the occasional overlong line, but at least the
5482 remaining text gets split now.
5484 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5485 file descriptors might be low, and that setting 1000 would always raise it.
5486 It turns out that in some environments, the limit is already over 1000 and
5487 that lowering it causes trouble. So now Exim takes care not to decrease it.
5489 40. When delivering a message, the value of $return_path is set to $sender_
5490 address at the start of routing (routers may change the value). By an
5491 oversight, this default was not being set up when an address was tested by
5492 -bt or -bv, which affected the outcome if any router or filter referred to
5495 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5496 log only when "message" or "log_message" are set. However, if one of the
5497 conditions was an address verification, or a call to a nested ACL, the
5498 messages generated by the underlying test were being passed through. This
5499 no longer happens. The underlying message is available in $acl_verify_
5500 message for both "message" and "log_message" expansions, so it can be
5501 passed through if needed.
5503 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5504 new expansion $bh_ to give the encoded byte string without charset
5505 translation. Translation happens only if iconv() is available; HAVE_ICONV
5506 indicates this at build time. HEADERS_CHARSET gives the charset to
5507 translate to; headers_charset can change it in the configuration, and
5508 "headers charset" can change it in an individual filter file.
5510 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5511 that creates RFC 2047 encoded "words" labels them as that charset instead
5512 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5513 expansion operator; (ii) when Exim creates a From: line for a local
5514 message; (iii) when a header line is rewritten to include a "phrase" part.
5516 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5517 buggy, causing it to skip the first lines of messages whose message ID
5518 ended in 'D'. This would not have bitten before Exim release 4.14, because
5519 message IDs were unlikely to end in 'D' before then. The effect was to have
5520 incorrect size information for certain domains.
5522 45. #include "config.h" was missing at the start of the crypt16.c module. This
5523 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5526 46. If there was a timeout during a "random" callout check, Exim treated it as
5527 a failure of the random address, and carried on sending RSET and the real
5528 address. If the delay was just some slowness somewhere, the response to the
5529 original RCPT would be taken as a response to RSET and so on, causing
5530 mayhem of various kinds.
5532 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5533 when I implemented it. It didn't allow for the fact that some option values
5534 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5535 the right test for positive values.
5537 48. Domain names in DNS records are case-independent. Exim always looks them up
5538 in lower case. Some resolvers return domain names in exactly the case they
5539 appear in the zone file, that is, they may contain uppercase letters. Not
5540 all resolvers do this - some return always lower case. Exim was treating a
5541 change of case by a resolver as a change of domain, similar to a widening
5542 of a domain abbreviation. This triggered its re-routing code and so it was
5543 trying to route what was effectively the same domain again. This normally
5544 caused routing to fail (because the router wouldn't handle the domain
5545 twice). Now Exim checks for this case specially, and just changes the
5546 casing of the domain that it ultimately uses when it transmits the message
5549 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5552 50. If a filter generated a file delivery with a non-absolute name (possible if
5553 no home directory exists for the router), the forbid_file option was not
5556 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5557 the existing equality matching.
5559 52. Exim was using ints instead of ino_t variables in some places where it was
5560 dealing with inode numbers.
5562 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5563 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5564 called TMPDIR, and if it finds it is different, it changes its value.
5566 54. The smtp_printf() function is now made available to local_scan() so
5567 additional output lines can be written before returning. There is also an
5568 smtp_fflush() function to enable the detection of a dropped connection.
5569 The variables smtp_input and smtp_batched_input are exported to
5572 55. Changed the default runtime configuration: the message "Unknown user"
5573 has been removed from the ACL, and instead placed on the localuser router,
5574 using the cannot_route_message feature. This means that any verification
5575 failures that generate their own messages won't get overridden. Similarly,
5576 the "Unrouteable address" message that was in the ACL for unverifiable
5577 relay addresses has also been removed.
5579 56. Added hosts_avoid_esmtp to the smtp transport.
5581 57. The exicyclog script was not checking for the esoteric option
5582 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5583 will work only if exicyclog is run under the appropriate euid.
5585 58. Following a discussion on the list, the rules by which Exim recognises line
5586 endings on incoming messages have been changed. The -dropcr and drop_cr
5587 options are now no-ops, retained only for backwards compatibility. The
5588 following line terminators are recognized: LF CRLF CR. However, special
5589 processing applies to CR:
5591 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5592 nor a local message in the state where . is a terminator.
5594 (ii) If a bare CR is encountered in a header line, an extra space is added
5595 after the line terminator so as not to end the header. The reasoning
5596 behind this is that bare CRs in header lines are most likely either
5597 to be mistakes, or people trying to play silly games.
5599 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5600 was being incorrectly given as 18 bytes larger than it should have been.
5601 This is a VOB (very old bug).
5603 60. This may never have affected anything current, but just in case it has:
5604 When the local host is found other than at the start of a list of hosts,
5605 the local host, those with the same MX, and any that follow, are discarded.
5606 When the list in question was part of a longer list of hosts, the following
5607 hosts (not currently being processed) were also being discarded. This no
5608 longer happens. I'm not sure if this situation could ever has previously
5611 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5614 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5615 header. This is something that is recommended in a new Internet Draft, and
5616 is something that is documented as being done by Sendmail. There are two
5617 possible values. For messages generated by the autoreply transport, Exim
5620 Auto-Submitted: auto-replied
5622 whereas for all other generated messages (e.g. bounces) it adds
5624 Auto-Submitted: auto-generated
5626 63. The "personal" condition in filters now includes a test for the
5627 Auto-Submitted: header. If it contains the string "auto-" the message it
5628 not considered personal.
5630 64. Added rcpt_include_affixes as a generic transport option.
5632 65. Added queue_only_override (default true).
5634 66. Added the syslog_duplication option.
5636 67. If what should have been the first header line of a message consisted of
5637 a space followed by a colon, Exim was mis-interpreting it as a header line.
5638 It isn't of course - it is syntactically invalid and should therefore be
5639 treated as the start of the message body. The misbehaviour could have
5640 caused a number of strange effects, including loss of data in subsequent
5641 header lines, and spool format errors.
5643 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5644 client host had authenticated. This control can now be exercised by an ACL
5645 for more flexibility.
5647 69. By default, callouts do not happen when testing with -bh. There is now a
5648 variant, -bhc, which does actually run the callout code, including
5649 consulting and updating the callout cache.
5651 70. Added support for saslauthd authentication, courtesy of Alexander
5654 71. If statvfs() failed on the spool or log directories while checking their
5655 size for availability, Exim confusingly gave the error "space shortage".
5656 Furthermore, in debugging mode it crashed with a floating point exception.
5657 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5658 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5659 serious problem, Exim now writes to the main and panic logs when this
5660 happens, with details of the failure. It then refuses to accept the
5661 incoming message, giving the message "spool directory problem" or "log
5662 directory problem" with a 421 code for SMTP messages.
5664 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5665 0, 1, and 2 exist, because some OS complain for execs without them (see
5666 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5667 descriptors. However, the code omitted to check that the open succeeded,
5668 causing mysterious errors if for some reason the permissions on /dev/null
5669 got screwed. Now Exim writes a message to the main and panic logs, and
5670 bombs out if it can't open /dev/null.
5672 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5673 interact so that it is all more flexible. It is supposed to remain
5674 backwards compatible. Also added extra_local_interfaces.
5676 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5677 to bomb out with an assertion failure - to the client this appears as a
5678 connection drop. This problem occurs in the part of the code that was taken
5679 from the Samba project. Fortunately, the assertion is in a very simple
5680 function, so I have fixed this by reproducing the function inline in the
5681 one place where it is called, and arranging for authentication to fail
5682 instead of killing the process with assert().
5684 75. The SPA client code was not working when the server requested OEM rather
5685 than Unicode encoding.
5687 76. Added code to make require_files with a specific uid setting more usable in
5688 the case where statting the file as root fails - usually a non-root-mounted
5689 NFS file system. When this happens and the failure is EACCES, Exim now
5690 forks a subprocess and does the per-uid checking as the relevant uid.
5692 77. Added process_log_path.
5694 78. If log_file_path was not explicitly set, a setting of check_log_space or
5695 check_log_inodes was ignored.
5697 79. If a space check for the spool or log partitions fails, the incident is now
5698 logged. Of course, in the latter case the data may get lost...
5700 80. Added the %p formatting code to string_format() so that it can be used to
5701 print addresses in debug_print(). Adjusted all the address printing in the
5702 debugging in store.c to use %p rather than %d.
5704 81. There was a concern that a line of code in smtp_in.c could overflow a
5705 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5706 initially expressed, the concern was not well-founded, because trailing
5707 spaces are removed early. However, if the trailing spaces were followed by
5708 a NULL, they did not get removed, so the overflow was possible. Two fixes
5711 (a) I re-wrote the offending code in a cleaner fashion.
5712 (b) If an incoming SMTP command contains a NULL character, it is rejected
5715 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5716 runs initgroups(), so that if the Exim user is in any additional groups,
5717 they will be used during message reception.
5723 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob