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.
78 PP/01 Handle short writes when writing local log-files.
79 In practice, only affects FreeBSD (8 onwards).
80 Bugzilla 1053, with thanks to Dmitry Isaikin.
82 NM/01 Bugzilla 949 - Documentation tweak
84 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
87 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
89 PP/02 Implemented gsasl authenticator.
91 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
93 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
94 `pkg-config foo` for cflags/libs.
96 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
97 with rest of GSASL and with heimdal_gssapi.
99 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
100 `pkg-config foo` for cflags/libs for the TLS implementation.
102 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
103 properties get this fed in as external SSF. A number of robustness
104 and debugging improvements to the cyrus_sasl authenticator.
106 PP/08 cyrus_sasl server now expands the server_realm option.
108 PP/09 Bugzilla 1214 - Log authentication information in reject log.
109 Patch by Jeremy Harris.
111 PP/10 Added dbmjz lookup type.
113 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
115 PP/12 MAIL args handles TAB as well as SP, for better interop with
116 non-compliant senders.
117 Analysis and variant patch by Todd Lyons.
119 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
120 Bug report from Lars Müller <lars@samba.org> (via SUSE),
121 Patch from Dirk Mueller <dmueller@suse.com>
123 PP/13 tls_peerdn now print-escaped for spool files.
124 Observed some $tls_peerdn in wild which contained \n, which resulted
125 in spool file corruption.
127 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
128 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
129 or write after TLS renegotiation, which otherwise led to messages
132 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
133 as a tracking header (ie: a signed header comes before the signature).
134 Patch from Wolfgang Breyha.
136 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
137 comma-sep list; embedded commas doubled.
139 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
141 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
143 Report and patch from Dmitry Banschikov.
145 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
146 Removed SSL_clear() after SSL_new() which led to protocol negotiation
147 failures. We appear to now support TLS1.1+ with Exim.
149 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
150 lets Exim select keys and certificates based upon TLS SNI from client.
151 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
152 before an outbound SMTP session. New log_selector, +tls_sni.
154 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
155 NULL dereference. Report and patch from Alun Jones.
157 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
158 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
159 Not seeing resolver debug output on NetBSD, but suspect this is a
160 resolver implementation change.
162 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
163 Left warnings. Added "eximon gdb" invocation mode.
165 PP/21 Defaulting "accept_8bitmime" to true, not false.
167 PP/22 Added -bw for inetd wait mode support.
169 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
170 locate the relevant includes and libraries. Made this the default.
172 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
173 Bugzilla 1246, report and most of solution from Tomasz Kusy.
175 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
176 This may cause build issues on older platforms.
178 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
179 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
180 gnutls_require_mac & gnutls_require_protocols (no longer supported).
181 Added SNI support via GnuTLS too.
182 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
184 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
186 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
187 Applied second patch from Janne, implementing suggestion to default
188 multiple-strings-in-record handling to match SPF spec.
190 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
192 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
193 read-only, out of scope).
194 Patch from Wolfgang Breyha, report from Stuart Northfield.
196 PP/29 Fix three issues highlighted by clang analyser static analysis.
197 Only crash-plausible issue would require the Cambridge-specific
198 iplookup router and a misconfiguration.
199 Report from Marcin Mirosław.
201 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
203 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
204 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
205 As part of this, removing so much warning spew let me fix some minor
206 real issues in debug logging.
208 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
209 assignment on my part. Fixed.
211 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
212 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
213 Janne Snabb (who went above and beyond: thank you).
215 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
216 string otherwise requires a connection and a bunch more work and it's
217 relatively easy to get wrong. Should also expose TLS library linkage
220 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
221 64-bit ${eval} (JH/03).
223 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
224 GNU libc to support some of the 64-bit stuff, should not lead to
225 conflicts. Defined before os.h is pulled in, so if a given platform
226 needs to override this, it can.
228 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
229 protection layer was required, which is not implemented.
230 Bugzilla 1254, patch from Wolfgang Breyha.
232 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
233 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
234 tls_dhparam take prime identifiers. Also unbreak combination of
235 OpenSSL+DH_params+TLSSNI.
237 PP/39 Disable SSLv2 by default in OpenSSL support.
243 PP/01 Solaris build fix for Oracle's LDAP libraries.
244 Bugzilla 1109, patch from Stephen Usher.
246 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
248 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
251 TF/02 Fix a couple more cases where we did not log the error message
252 when unlink() failed. See also change 4.74-TF/03.
254 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
255 lock up or crash if it happened to be inside a call to libc when it
256 got a SIGUSR1 from exiwhat.
258 The SIGUSR1 handler appends the current process status to the process
259 log which is later printed by exiwhat. It used to use the general
260 purpose logging code to do this, but several functions it calls are
261 not safe for signals.
263 The new output code in the SIGUSR1 handler is specific to the process
264 log, and simple enough that it's easy to inspect for signal safety.
265 Removing some special cases also simplifies the general logging code.
266 Removing the spurious timestamps from the process log simplifies
269 TF/04 Improved ratelimit ACL condition.
271 The /noupdate option has been deprecated in favour of /readonly which
272 has clearer semantics. The /leaky, /strict, and /readonly update modes
273 are mutually exclusive. The update mode is no longer included in the
274 database key; it just determines when the database is updated. (This
275 means that when you upgrde Exim will forget old rate measurements.)
277 Exim now checks that the per_* options are used with an update mode that
278 makes sense for the current ACL. For example, when Exim is processing a
279 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
280 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
281 must specify per_mail/readonly. If you omit the update mode it defaults to
282 /leaky where that makes sense (as before) or /readonly where required.
284 The /noupdate option is now undocumented but still supported for
285 backwards compatibility. It is equivalent to /readonly except that in
286 ACLs where /readonly is required you may specify /leaky/noupdate or
287 /strict/noupdate which are treated the same as /readonly.
289 A useful new feature is the /count= option. This is a generalization
290 of the per_byte option, so that you can measure the throughput of other
291 aggregate values. For example, the per_byte option is now equivalent
292 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
294 The per_rcpt option has been generalized using the /count= mechanism
295 (though it's more complicated than the per_byte equivalence). When it is
296 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
297 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
298 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
299 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
300 Note that using per_rcpt with a non-readonly update mode in more than
301 one ACL will cause the recipients to be double-counted. (The per_mail
302 and per_byte options don't have this problem.)
304 The handling of very low rates has changed slightly. If the computed rate
305 is less than the event's count (usually one) then this event is the first
306 after a long gap. In this case the rate is set to the same as this event's
307 count, so that the first message of a spam run is counted properly.
309 The major new feature is a mechanism for counting the rate of unique
310 events. The new per_addr option counts the number of different
311 recipients that someone has sent messages to in the last time period. It
312 behaves like per_rcpt if all the recipient addresses are different, but
313 duplicate recipient addresses do not increase the measured rate. Like
314 the /count= option this is a general mechanism, so the per_addr option
315 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
316 example, measure the rate that a client uses different sender addresses
317 with the options per_mail/unique=$sender_address. There are further
318 details in the main documentation.
320 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
322 TF/06 Removed a few PCRE remnants.
324 TF/07 Automatically extract Exim's version number from tags in the git
325 repository when doing development or release builds.
327 PP/02 Raise smtp_cmd_buffer_size to 16kB.
328 Bugzilla 879. Patch from Paul Fisher.
330 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
331 Heavily based on revision 40f9a89a from Simon Arlott's tree.
334 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
336 PP/05 Variable $av_failed, true if the AV scanner deferred.
337 Bugzilla 1078. Patch from John Horne.
339 PP/06 Stop make process more reliably on build failure.
340 Bugzilla 1087. Patch from Heiko Schlittermann.
342 PP/07 Make maildir_use_size_file an _expandable_ boolean.
343 Bugzilla 1089. Patch from Heiko Schlittermann.
345 PP/08 Handle ${run} returning more data than OS pipe buffer size.
346 Bugzilla 1131. Patch from Holger Weiß.
348 PP/09 Handle IPv6 addresses with SPF.
349 Bugzilla 860. Patch from Wolfgang Breyha.
351 PP/10 GnuTLS: support TLS 1.2 & 1.1.
353 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
356 PP/11 match_* no longer expand right-hand-side by default.
357 New compile-time build option, EXPAND_LISTMATCH_RHS.
358 New expansion conditions, "inlist", "inlisti".
360 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
362 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
364 PP/14 fix log_write() format string regression from TF/03.
365 Bugzilla 1152. Patch from Dmitry Isaikin.
371 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
373 PP/02 Harmonised TLS library version reporting; only show if debugging.
374 Layout now matches that introduced for other libraries in 4.74 PP/03.
376 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
378 PP/04 New "dns_use_edns0" global option.
380 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
383 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
384 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
386 TK/01 Updated PolarSSL code to 0.14.2.
387 Bugzilla 1097. Patch from Andreas Metzler.
389 PP/07 Catch divide-by-zero in ${eval:...}.
392 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
395 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
396 format-string attack -- SECURITY: remote arbitrary code execution.
398 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
399 time unintentionally subject to list matching rules, letting the header
400 cause arbitrary Exim lookups (of items which can occur in lists, *not*
401 arbitrary string expansion). This allowed for information disclosure.
403 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
404 INT_MIN/-1 -- value coerced to INT_MAX.
410 NM/01 Workround for PCRE version dependency in version reporting
413 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
414 This fixes portability to compilers other than gcc, notably
415 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
417 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
418 makefiles for portability to HP-UX and POSIX correctness.
420 PP/01 Permit LOOKUP_foo enabling on the make command-line.
421 Also via indented variable definition in the Makefile.
422 (Debugging by Oliver Heesakkers).
424 PP/02 Restore caching of spamd results with expanded spamd_address.
425 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
427 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
428 Improves build reliability. Fix from: Frank Elsner
430 NM/02 Fix wide character breakage in the rfc2047 coding
431 Fixes bug 1064. Patch from Andrey N. Oktyabrski
433 NM/03 Allow underscore in dnslist lookups
434 Fixes bug 1026. Patch from Graeme Fowler
436 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
437 Code patches from Adam Ciarcinski of NetBSD.
439 NM/04 Fixed exiqgrep to cope with mailq missing size issue
442 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
443 is logged, to avoid truncation. Patch from John Horne.
445 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
446 Patch from Jakob Hirsch.
448 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
449 SQL string expansion failure details.
450 Patch from Andrey Oktyabrski.
452 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
453 Patch from Simon Arlott.
455 PP/09 New lookups functionality failed to compile on old gcc which rejects
456 extern declarations in function scope.
457 Patch from Oliver Fleischmann
459 PP/10 Use sig_atomic_t for flags set from signal handlers.
460 Check getgroups() return and improve debugging.
461 Fixed developed for diagnosis in bug 927 (which turned out to be
464 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
465 Patch from Mark Zealey.
467 PP/12 Bugzilla 1056: Improved spamd server selection.
468 Patch from Mark Zealey.
470 PP/13 Bugzilla 1086: Deal with maildir quota file races.
471 Based on patch from Heiko Schlittermann.
473 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
474 Patch from Uwe Doering, sign-off by Michael Haardt.
476 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
477 variable declaration deep within a block. Bug and patch from
480 PP/15 lookups-Makefile IRIX compatibilty coercion.
482 PP/16 Make DISABLE_DKIM build knob functional.
484 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
485 Patch by Simon Arlott
487 TF/03 Fix valgrind.h portability to C89 compilers that do not support
488 variable argument macros. Our copy now differs from upstream.
494 TF/01 Failure to get a lock on a hints database can have serious
495 consequences so log it to the panic log.
497 TF/02 Log LMTP confirmation messages in the same way as SMTP,
498 controlled using the smtp_confirmation log selector.
500 TF/03 Include the error message when we fail to unlink a spool file.
502 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
503 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
504 for maintaining out-of-tree patches for some time.
506 PP/01 Bugzilla 139: Documentation and portability issues.
507 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
508 Handle per-OS dynamic-module compilation flags.
510 PP/02 Let /dev/null have normal permissions.
511 The 4.73 fixes were a little too stringent and complained about the
512 permissions on /dev/null. Exempt it from some checks.
513 Reported by Andreas M. Kirchwitz.
515 PP/03 Report version information for many libraries, including
516 Exim version information for dynamically loaded libraries. Created
517 version.h, now support a version extension string for distributors
518 who patch heavily. Dynamic module ABI change.
520 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
521 privilege escalation vulnerability whereby the Exim run-time user
522 can cause root to append content of the attacker's choosing to
525 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
528 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
529 If dropping privileges for untrusted macros, we disabled normal logging
530 on the basis that it would fail; for the Exim run-time user, this is not
531 the case, and it resulted in successful deliveries going unlogged.
532 Fixed. Reported by Andreas Metzler.
538 PP/01 Date: & Message-Id: revert to normally being appended to a message,
539 only prepend for the Resent-* case. Fixes regression introduced in
540 Exim 4.70 by NM/22 for Bugzilla 607.
542 PP/02 Include check_rfc2047_length in configure.default because we're seeing
543 increasing numbers of administrators be bitten by this.
545 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
547 PP/03 Bugzilla 994: added openssl_options main configuration option.
549 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
551 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
553 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
555 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
556 without a peer certificate, leading to a segfault because of an
557 assumption that peers always have certificates. Be a little more
558 paranoid. Problem reported by Martin Tscholak.
560 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
561 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
562 NB: ClamAV planning to remove STREAM in "middle of 2010".
563 CL also introduces -bmalware, various -d+acl logging additions and
564 more caution in buffer sizes.
566 PP/09 Implemented reverse_ip expansion operator.
568 PP/10 Bugzilla 937: provide a "debug" ACL control.
570 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
572 PP/12 Bugzilla 973: Implement --version.
574 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
576 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
578 PP/15 Bugzilla 816: support multiple condition rules on Routers.
580 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
581 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
582 ignore trailing whitespace.
584 JJ/02 prevent non-panic DKIM error from being sent to paniclog
586 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
589 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
590 Notification from Dr Andrew Aitchison.
592 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
593 ExtendedDetectionInfo response format.
594 Notification from John Horne.
596 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
599 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
600 XSL and documented dependency on system catalogs, with examples of how
603 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
606 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
607 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
608 configuration file which is writeable by the Exim user or group.
610 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
611 of configuration files to cover files specified with the -C option if
612 they are going to be used with root privileges, not just the default
615 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
616 option (effectively making it always true).
618 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
619 files to be used while preserving root privileges.
621 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
622 that rogue child processes cannot use them.
624 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
625 run-time user, instead of root.
627 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
628 Exim run-time user without dropping privileges.
630 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
631 result string, instead of calling string_vformat() twice with the same
634 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
635 for other users. Others should always drop root privileges if they use
636 -C on the command line, even for a whitelisted configure file.
638 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
640 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
646 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
647 $data_path, and $header_path variables; fixed documentation bugs and
650 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
651 exipick to access non-standard spools, including the "frozen" queue
654 NM/01 Bugzilla 965: Support mysql stored procedures.
655 Patch from Alain Williams
657 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
659 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
660 Patch from Andreas Metzler
662 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
663 Patch from Kirill Miazine
665 NM/05 Bugzilla 671: Added umask to procmail example.
667 JJ/03 installed exipick 20100323.0, fixing doc bug
669 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
670 directory. Notification and patch from Dan Rosenberg.
672 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
674 TK/02 Improve log output when DKIM signing operation fails.
676 MH/01 Treat the transport option dkim_domain as a colon separated
677 list, not as a single string, and sign the message with each element,
678 omitting multiple occurences of the same signer.
680 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
681 Bugzilla 985, 986. Patch by Simon Arlott
683 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
684 Patch by Simon Arlott
686 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
687 MBX locking. Notification from Dan Rosenberg.
693 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
695 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
697 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
699 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
701 NM/04 Bugzilla 915: Flag broken perl installation during build.
707 TK/01 Added patch by Johannes Berg that expands the main option
708 "spamd_address" if it starts with a dollar sign.
710 TK/02 Write list of recipients to X-Envelope-Sender header when building
711 the mbox-format spool file for content scanning (suggested by Jakob
714 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
715 (http://www.dcc-servers.net/) support via dccifd. Activated by
716 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
718 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
719 by Mark Daniel Reidel <mr@df.eu>.
721 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
722 When building exim an external PCRE library is now needed -
723 PCRE is a system library on the majority of modern systems.
724 See entry on PCRE_LIBS in EDITME file.
726 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
727 conversation. Added nologin parameter to request.
728 Patch contributed by Kirill Miazine.
730 TF/01 Do not log submission mode rewrites if they do not change the address.
732 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
734 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
735 log files in place. Contributed by Roberto Lima.
737 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
739 TF/03 Bugzilla 615: When checking the local_parts router precondition
740 after a local_part_suffix or local_part_prefix option, Exim now
741 does not use the address's named list lookup cache, since this
742 contains cached lookups for the whole local part.
744 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
745 Robert Millan. Documentation is in experimental-spec.txt.
747 TF/04 Bugzilla 668: Fix parallel build (make -j).
749 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
751 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
752 Patch provided by Jan Srzednicki.
754 TF/05 Leading white space used to be stripped from $spam_report which
755 wrecked the formatting. Now it is preserved.
757 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
758 that they are available at delivery time.
760 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
762 TF/08 TLS error reporting now respects the incoming_interface and
763 incoming_port log selectors.
765 TF/09 Produce a more useful error message if an SMTP transport's hosts
766 setting expands to an empty string.
768 NM/06 Bugzilla 744: EXPN did not work under TLS.
769 Patch provided by Phil Pennock.
771 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
772 Patch provided by Richard Godbee.
774 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
775 acl_smtp_notquit, added index entry.
777 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
778 Patch provided by Eugene Bujak.
780 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
781 accept(). Patch provided by Maxim Dounin.
783 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
784 Patch provided by Phil Pennock.
786 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
788 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
789 Patch provided by Brad "anomie" Jorsch.
791 NM/14 Bugzilla 622: Added timeout setting to transport filter.
792 Patch provided by Dean Brooks.
794 TK/05 Add native DKIM support (does not depend on external libraries).
796 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
797 Patch provided by Graeme Fowler.
799 NM/16 Bugzilla 851: Documentation example syntax fix.
801 NM/17 Changed NOTICE file to remove references to embedded PCRE.
803 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
806 NM/19 Bugzilla 745: TLS version reporting.
807 Patch provided by Phil Pennock.
809 NM/20 Bugzilla 167: bool: condition support.
810 Patch provided by Phil Pennock.
812 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
813 clients. Patch provided by Phil Pennock.
815 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
816 Patch provided by Brad "anomie" Jorsch.
818 NM/23 Bugzilla 687: Fix misparses in eximstats.
819 Patch provided by Heiko Schlittermann.
821 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
822 Patch provided by Heiko Schlittermann.
824 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
825 plus update to original patch.
827 NM/26 Bugzilla 799: Documentation correction for ratelimit.
829 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
830 Patch provided by David Brownlee.
832 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
834 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
836 NM/30 Bugzilla 888: TLS documentation bugfixes.
838 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
840 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
841 Unlike the original bugzilla I have changed all shell scripts in src tree.
843 NM/33 Bugzilla 898: Transport filter timeout fix.
844 Patch by Todd Rinaldo.
846 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
847 Patch by Serge Demonchaux.
849 NM/35 Bugzilla 39: Base64 decode bug fixes.
850 Patch by Jakob Hirsch.
852 NM/36 Bugzilla 909: Correct connect() call in dcc code.
854 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
856 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
858 NM/39 Bugzilla 911: Fixed MakeLinks build script.
864 TK/01 Add preliminary DKIM support. Currently requires a forked version of
865 ALT-N's libdkim that I have put here:
866 http://duncanthrax.net/exim-experimental/
868 Note to Michael Haardt: I had to rename some vars in sieve.c. They
869 were called 'true' and it seems that C99 defines that as a reserved
870 keyword to be used with 'bool' variable types. That means you could
871 not include C99-style headers which use bools without triggering
872 build errors in sieve.c.
874 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
875 as mailq or other aliases. Changed the --help handling significantly
876 to do whats expected. exim_usage() emits usage/help information.
878 SC/01 Added the -bylocaldomain option to eximstats.
880 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
882 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
884 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
890 PH/01 Another patch from the Sieve maintainer.
892 PH/02 When an IPv6 address is converted to a string for single-key lookup
893 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
894 dots are used instead of colons so that keys in lsearch files need not
895 contain colons. This was done some time before quoting was made available
896 in lsearch files. However, iplsearch files do require colons in IPv6 keys
897 (notated using the quote facility) so as to distinguish them from IPv4
898 keys. This meant that lookups for IP addresses in host lists did not work
899 for iplsearch lookups.
901 This has been fixed by arranging for IPv6 addresses to be expressed with
902 colons if the lookup type is iplsearch. This is not incompatible, because
903 previously such lookups could never work.
905 The situation is now rather anomolous, since one *can* have colons in
906 ordinary lsearch keys. However, making the change in all cases is
907 incompatible and would probably break a number of configurations.
909 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
912 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
913 conversion specification without a maximum field width, thereby enabling
914 a rogue spamd server to cause a buffer overflow. While nobody in their
915 right mind would setup Exim to query an untrusted spamd server, an
916 attacker that gains access to a server running spamd could potentially
917 exploit this vulnerability to run arbitrary code as the Exim user.
919 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
920 $primary_hostname instead of what libspf2 thinks the hosts name is.
922 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
923 a directory entry by the name of the lookup key. Previously, if a
924 symlink pointed to a non-existing file or a file in a directory that
925 Exim lacked permissions to read, a lookup for a key matching that
926 symlink would fail. Now it is enough that a matching directory entry
927 exists, symlink or not. (Bugzilla 503.)
929 PH/03 The body_linecount and body_zerocount variables are now exported in the
932 PH/04 Added the $dnslist_matched variable.
934 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
935 This means they are set thereafter only if the connection becomes
938 PH/06 Added the client_condition to authenticators so that some can be skipped
939 by clients under certain conditions.
941 PH/07 The error message for a badly-placed control=no_multiline_responses left
942 "_responses" off the end of the name.
944 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
946 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
947 (without spaces) instead of just copying the configuration text.
949 PH/10 Added the /noupdate option to the ratelimit ACL condition.
951 PH/11 Added $max_received_linelength.
953 PH/12 Added +ignore_defer and +include_defer to host lists.
955 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
956 way in which PCRE > 7.0 is built.
958 PH/14 Implemented queue_only_load_latch.
960 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
961 MAIL command. The effect was to mangle the value on 64-bit systems.
963 PH/16 Another patch from the Sieve maintainer.
965 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
967 PH/18 If a system quota error occurred while trying to create the file for
968 a maildir delivery, the message "Mailbox is full" was not appended to the
969 bounce if the delivery eventually timed out. Change 4.67/27 below applied
970 only to a quota excession during the actual writing of the file.
972 PH/19 It seems that peer DN values may contain newlines (and other non-printing
973 characters?) which causes problems in log lines. The DN values are now
974 passed through string_printing() before being added to log lines.
976 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
977 and InterBase are left for another time.)
979 PH/21 Added message_body_newlines option.
981 PH/22 Guard against possible overflow in moan_check_errorcopy().
983 PH/23 POSIX allows open() to be a macro; guard against that.
985 PH/24 If the recipient of an error message contained an @ in the local part
986 (suitably quoted, of course), incorrect values were put in $domain and
987 $local_part during the evaluation of errors_copy.
993 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
994 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
997 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
998 issue a MAIL command.
1000 PH/02 In an ACL statement such as
1002 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1004 if a client was not listed at all, or was listed with a value other than
1005 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1006 the condition was not true (as it should be), so access was not denied.
1007 The bug was that the ! inversion was incorrectly passed on to the second
1008 item. This has been fixed.
1010 PH/03 Added additional dnslists conditions == and =& which are different from
1011 = and & when the dns lookup returns more than one IP address.
1013 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1014 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1016 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1017 FSYNC, which compiles an option called disable_fsync that allows for
1018 bypassing fsync(). The documentation is heavily laced with warnings.
1020 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1022 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1023 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1024 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1025 including adding "make clean"; (3) Added -fPIC when compiling the test
1026 dynamically loaded module, to get rid of a warning.
1028 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1029 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1030 = 0s. The bug is otherwise harmless.
1032 PH/07 There was a bug in the dovecot authenticator such that the value of
1033 $auth1 could be overwritten, and so not correctly preserved, after a
1034 successful authentication. This usually meant that the value preserved by
1035 the server_setid option was incorrect.
1037 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1039 PH/09 Installed PCRE release 7.0.
1041 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1042 run for batched SMTP input. It is now run at the start of every message
1043 in the batch. While fixing this I discovered that the process information
1044 (output by running exiwhat) was not always getting set for -bs and -bS
1045 input. This is fixed, and it now also says "batched" for BSMTP.
1047 PH/11 Added control=no_pipelining.
1049 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1050 patch, slightly modified), and move the expansion of helo_data till after
1051 the connection is made in the smtp transport (so it can use these
1054 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1056 PH/14 Added log_selector = +pid.
1058 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1060 PH/16 Add ${if forany and ${if forall.
1062 PH/17 Added dsn_from option to vary the From: line in DSNs.
1064 PH/18 Flush SMTP output before performing a callout, unless control =
1065 no_callout_flush is set.
1067 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1068 was true (the default) a successful delivery failed to delete the retry
1069 item, thus causing premature timeout of the address. The bug is now
1072 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1074 PH/21 Long custom messages for fakedefer and fakereject are now split up
1075 into multiline reponses in the same way that messages for "deny" and
1076 other ACL rejections are.
1078 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1079 with slight modification.
1081 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1082 draft, changing the syntax and factoring some duplicate code".
1084 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1085 for deliveries of the second and subsequent messages over the same SMTP
1088 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1089 ${reduce, with only minor "tidies".
1091 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1093 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1094 expansion side effects.
1096 PH/27 When a message times out after an over-quota error from an Exim-imposed
1097 quota, the bounce message says "mailbox is full". This message was not
1098 being given when it was a system quota that was exceeded. It now should
1101 MH/03 Made $recipients available in local_scan(). local_scan() already has
1102 better access to the recipient list through recipients_list[], but
1103 $recipients can be useful in postmaster-provided expansion strings.
1105 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1106 in the case of a MAIL command with additional options following the
1107 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1108 were accidentally chopped off.
1110 PH/29 SMTP synchronization checks are implemented when a command is read -
1111 there is a check that no more input is waiting when there shouldn't be
1112 any. However, for some commands, a delay in an ACL can mean that it is
1113 some time before the response is written. In this time, more input might
1114 arrive, invalidly. So now there are extra checks after an ACL has run for
1115 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1116 pipelining has not been advertised.
1118 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1120 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1121 returned in the SMTP response when smtp_return_error_details was false.
1122 This has been fixed.
1124 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1125 instead of the C I/O that was originally supplied, because problems were
1126 reported on Solaris.
1128 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1129 Exim which did not show up earlier: it was assuming that a call to
1130 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1131 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1132 was a macro that became an assignment, so it seemed to work. This has
1133 changed to a proper function call with a void return, hence the compile
1134 error. Exim's code has been fixed.
1136 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1139 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1141 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1143 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1144 because it turns out to be too big or there is a timeout) there is no
1145 "Completed" line in the log. When some messages of this type were
1146 selected by exigrep, they were listed as "not completed". Others were
1147 picked up by some special patterns. I have improved the selection
1148 criteria to be more general.
1150 PH/38 The host_find_failed option in the manualroute router can now be set
1151 to "ignore", to completely ignore a host whose IP address cannot be
1152 found. If all hosts are ignored, the behaviour is controlled by the new
1153 host_all_ignored option.
1155 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1156 homing or because of multiple MX records with /mx) generated more than
1157 one IP address, and the following item turned out to be the local host,
1158 all the secondary addresses of the first item were incorrectly removed
1159 from the list, along with the local host and any following hosts (which
1160 is what is supposed to happen).
1162 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1163 whoever called Exim into the -H file. In the case of the daemon it was
1164 behaving confusingly. When first started, it used values for whoever
1165 started the daemon, but after a SIGHUP it used the Exim user (because it
1166 calls itself on a restart). I have changed the code so that it now always
1169 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1170 message are rejected with the same error (e.g. no authentication or bad
1171 sender address), and a DATA command is nevertheless sent (as can happen
1172 with PIPELINING or a stupid MUA), the error message that was given to the
1173 RCPT commands is included in the rejection of the DATA command. This is
1174 intended to be helpful for MUAs that show only the final error to their
1177 PH/42 Another patch from the Sieve maintainer.
1179 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1180 Eximstats - Fixed some broken HTML links and added missing column headers
1182 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1183 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1185 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1187 PH/43 Yet another patch from the Sieve maintainer.
1189 PH/44 I found a way to check for a TCP/IP connection going away before sending
1190 the response to the final '.' that terminates a message, but only in the
1191 case where the client has not sent further data following the '.'
1192 (unfortunately, this is allowed). However, in many cases there won't be
1193 any further data because there won't be any more messages to send. A call
1194 to select() can be used: if it shows that the input is "ready", there is
1195 either input waiting, or the socket has been closed. An attempt to read
1196 the next input character can distinguish the two cases. Previously, Exim
1197 would have sent an OK response which the client would never have see.
1198 This could lead to message repetition. This fix should cure that, at
1199 least in a lot of common cases.
1201 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1202 advertised in response to EHLO.
1208 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1209 fixed by 4.65/MH/01 (is this a record?) are fixed:
1211 (i) An empty string was always treated as zero by the numeric comparison
1212 operators. This behaviour has been restored.
1214 (ii) It is documented that the numeric comparison operators always treat
1215 their arguments as decimal numbers. This was broken in that numbers
1216 starting with 0 were being interpreted as octal.
1218 While fixing these problems I realized that there was another issue that
1219 hadn't been noticed. Values of message_size_limit (both the global option
1220 and the transport option) were treated as octal if they started with 0.
1221 The documentation was vague. These values are now always treated as
1222 decimal, and I will make that clear in the documentation.
1228 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1229 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1232 MH/01 Don't check that the operands of numeric comparison operators are
1233 integers when their expansion is in "skipping" mode (fixes bug
1234 introduced by 4.64-PH/07).
1236 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1237 child addresses, Exim now panics and dies. Previously, because the count
1238 is held in a short int, deliveries were likely to be lost. As such a
1239 large number of recipients for a single message is ridiculous
1240 (performance will be very, very poor), I have chosen to impose a limit
1241 rather than extend the field.
1247 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1248 leftover -K file (the existence of which was triggered by #402).
1249 While we were at it, introduced process PID as part of the -K
1250 filename. This should rule out race conditions when creating
1253 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1254 processing considerably. Previous code took too long for large mails,
1255 triggering a timeout which in turn triggers #401.
1257 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1258 in the DK code in transports.c. sendfile() is not really portable,
1259 hence the _LINUX specificness.
1261 TF/01 In the add_headers option to the mail command in an Exim filter,
1262 there was a bug that Exim would claim a syntax error in any
1263 header after the first one which had an odd number of characters
1266 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1267 callout verification, Exim cached a "reject" for the entire domain. This
1268 is correct for most verifications, but it is not correct for a recipient
1269 verification with use_sender or use_postmaster set, because in that case
1270 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1271 case of MAIL FROM:<> rejection from other early rejections (e.g.
1272 rejection of HELO). When verifying a recipient using a non-null MAIL
1273 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1274 Whatever the result of the callout, the value of the domain cache is
1275 left unchanged (for any other kind of callout, getting as far as trying
1276 RCPT means that the domain itself is ok).
1278 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1281 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1282 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1283 session, the daemon ignores ECONNECT errors and logs others; it now
1284 ignores EPIPE as well.
1286 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1287 (quoted-printable decoding).
1289 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1290 later the small subsequent patch to fix an introduced bug.
1292 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1294 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1296 PH/08 An error is now given if message_size_limit is specified negative.
1298 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1299 to be given (somewhat) arbitrary names.
1301 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1304 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1305 miscellaneous code fixes
1307 PH/10 Added the log_reject_target ACL modifier to specify where to log
1310 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1311 hostname. This is wrong, because it relates to the incoming message (and
1312 probably the interface on which it is arriving) and not to the outgoing
1313 callout (which could be using a different interface). This has been
1314 changed to use the value of the helo_data option from the smtp transport
1315 instead - this is what is used when a message is actually being sent. If
1316 there is no remote transport (possible with a router that sets up host
1317 addresses), $smtp_active_hostname is used.
1319 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1320 tweaks were necessary in order to get it to work (see also 21 below):
1321 (a) The code assumed that strncpy() returns a negative number on buffer
1322 overflow, which isn't the case. Replaced with Exim's string_format()
1324 (b) There were several signed/unsigned issues. I just did the minimum
1325 hacking in of casts. There is scope for a larger refactoring.
1326 (c) The code used strcasecmp() which is not a standard C function.
1327 Replaced with Exim's strcmpic() function.
1328 (d) The code set only $1; it now sets $auth1 as well.
1329 (e) A simple test gave the error "authentication client didn't specify
1330 service in request". It would seem that Dovecot has changed its
1331 interface. Fortunately there's a specification; I followed it and
1332 changed what the client sends and it appears to be working now.
1334 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1337 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1338 address A is aliased to B and C, where B exists and C does not. Without
1339 -v the output is "A verified" because verification stops after a
1340 successful redirection if more than one address is generated. However,
1341 with -v the child addresses are also verified. Exim was outputting "A
1342 failed to verify" and then showing the successful verification for C,
1343 with its parentage. It now outputs "B failed to verify", showing B's
1344 parentage before showing the successful verification of C.
1346 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1347 look up a TXT record in a specific list after matching in a combined
1350 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1351 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1352 they consult the DNS. I had assumed they would set it the way they
1353 wanted; and indeed my experiments on Linux seem to show that in some
1354 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1355 To be on the safe side, however, I have now made the interface to
1356 host_find_byname() similar to host_find_bydns(), with an argument
1357 containing the DNS resolver options. The host_find_byname() function now
1358 sets these options at its start, just as host_find_bydns() does. The smtp
1359 transport options dns_qualify_single and dns_search_parents are passed to
1360 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1361 of host_find_byname() use the default settings of RES_DEFNAMES
1362 (qualify_single) but not RES_DNSRCH (search_parents).
1364 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1365 spool_read_header() do less string testing, by means of a preliminary
1366 switch on the second character of optional "-foo" lines. (This is
1367 overdue, caused by the large number of possibilities that now exist.
1368 Originally there were few.) While I was there, I also converted the
1369 str(n)cmp tests so they don't re-test the leading "-" and the first
1370 character, in the hope this might squeeze out yet more improvement.
1372 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1373 flag allowing group syntax was set by the header_syntax check but not
1374 turned off, possible causing trouble later; (2) The flag was not being
1375 set at all for the header_verify test, causing "group"-style headers to
1376 be rejected. I have now set it in this case, and also caused header_
1377 verify to ignore an empty address taken from a group. While doing this, I
1378 came across some other cases where the code for allowing group syntax
1379 while scanning a header line wasn't quite right (mostly, not resetting
1380 the flag correctly in the right place). These bugs could have caused
1381 trouble for malformed header lines. I hope it is now all correct.
1383 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1384 with the "reply" argument non-NULL. The code, however (which originally
1385 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1386 but it didn't always do it. This confused somebody who was copying the
1387 code for some other use. I have removed all the tests.
1389 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1390 feature that was used to support insecure browsers during the U.S. crypto
1391 embargo. It requires special client support, and Exim is probably the
1392 only MTA that supported it -- and would never use it because real RSA is
1393 always available. This code has been removed, because it had the bad
1394 effect of slowing Exim down by computing (never used) parameters for the
1395 RSA_EXPORT functionality.
1397 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1398 authenticator to fail if there's a tab character in the incoming data
1399 (there should never be unless someone is messing about, as it's supposed
1400 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1401 if the connection is using TLS or if the remote IP is the same as the
1402 local IP, and the "valid-client-cert option" if a client certificate has
1405 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1406 authenticators. This can be used for authorization after authentication
1407 succeeds. (In the case of plaintext, it servers for both authentication
1410 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1411 if any retry times were supplied.
1413 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1414 connection, where there is no client IP address to check. In this
1415 situation, the verify now always succeeds.
1417 PH/25 Applied John Jetmore's -Mset patch.
1419 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1421 PH/27 In a string expansion for a processed (not raw) header when multiple
1422 headers of the same name were present, leading whitespace was being
1423 removed from all of them, but trailing whitespace was being removed only
1424 from the last one. Now trailing whitespace is removed from each header
1425 before concatenation. Completely empty headers in a concatenation (as
1426 before) are ignored.
1428 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1429 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1431 PH/29 [Removed. This was a change that I later backed out, and forgot to
1432 correct the ChangeLog entry (that I had efficiently created) before
1433 committing the later change.]
1435 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1436 address errors (4xx response to RCPT) over the same connection as other
1437 messages routed to the same hosts. Such deliveries are always "forced",
1438 so retry times are not inspected. This resulted in far too many retries
1439 for the affected addresses. The effect occurred only when there were more
1440 hosts than the hosts_max_try setting in the smtp transport when it had
1441 the 4xx errors. Those hosts that it had tried were not added to the list
1442 of hosts for which the message was waiting, so if all were tried, there
1443 was no problem. Two fixes have been applied:
1445 (i) If there are any address or message errors in an SMTP delivery, none
1446 of the hosts (tried or untried) are now added to the list of hosts
1447 for which the message is waiting, so the message should not be a
1448 candidate for sending over the same connection that was used for a
1449 successful delivery of some other message. This seems entirely
1450 reasonable: after all the message is NOT "waiting for some host".
1451 This is so "obvious" that I'm not sure why it wasn't done
1452 previously. Hope I haven't missed anything, but it can't do any
1453 harm, as the worst effect is to miss an optimization.
1455 (ii) If, despite (i), such a delivery is accidentally attempted, the
1456 routing retry time is respected, so at least it doesn't keep
1457 hammering the server.
1459 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1460 in ${readsocket because some servers need this prod.
1462 PH/32 Added some extra debug output when updating a wait-xxx database.
1464 PH/33 The hint "could be header name not terminated by colon", which has been
1465 given for certain expansion errors for a long time, was not being given
1466 for the ${if def:h_colon_omitted{... case.
1468 PH/34 The spec says: "With one important exception, whenever a domain list is
1469 being scanned, $domain contains the subject domain." There was at least
1470 one case where this was not true.
1472 PH/35 The error "getsockname() failed: connection reset by peer" was being
1473 written to the panic log as well as the main log, but it isn't really
1474 panic-worthy as it just means the connection died rather early on. I have
1475 removed the panic log writing for the ECONNRESET error when getsockname()
1478 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1479 runs only) independently of the message's sender address. This meant
1480 that, if the 4xx error was in fact related to the sender, a different
1481 message to the same recipient with a different sender could confuse
1482 things. In particualar, this can happen when sending to a greylisting
1483 server, but other circumstances could also provoke similar problems.
1484 I have changed the default so that the retry time for these errors is now
1485 based a combination of the sender and recipient addresses. This change
1486 can be overridden by setting address_retry_include_sender=false in the
1489 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1490 remote server are returned as part of bounce messages. This was not
1491 happening for LMTP over a pipe (the lmtp transport), but now it is the
1492 same for both kinds of LMTP.
1494 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1495 in header lines that were in fact CNAMEs. This is no longer the case.
1497 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1498 and queue runs started by the daemon processed all messages. This has
1499 been fixed so that -R and -S can now usefully be given with -q<time>.
1501 PH/40 Import PCRE release 6.7 (fixes some bugs).
1503 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1505 PH/42 Give an error if -q is specified more than once.
1507 PH/43 Renamed the variables $interface_address and $interface_port as
1508 $received_ip_address and $received_port, to make it clear that these
1509 values apply to message reception, and not to the outgoing interface when
1510 a message is delivered. (The old names remain recognized, of course.)
1512 PH/44 There was no timeout on the connect() call when using a Unix domain
1513 socket in the ${readsocket expansion. There now is.
1515 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1516 be meaningful with "accept".
1518 SC/01 Eximstats V1.43
1519 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1521 SC/02 Eximstats V1.44
1522 Use a glob alias rather than an array ref in the generated
1523 parser. This improves both readability and performance.
1525 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1526 Collect SpamAssassin and rejection statistics.
1527 Don't display local sender or destination tables unless
1528 there is data to show.
1529 Added average volumes into the top table text output.
1531 SC/04 Eximstats V1.46
1532 Collect data on the number of addresses (recipients)
1533 as well as the number of messages.
1535 SC/05 Eximstats V1.47
1536 Added 'Message too big' to the list of mail rejection
1537 reasons (thanks to Marco Gaiarin).
1539 SC/06 Eximstats V1.48
1540 Mainlog lines which have GMT offsets and are too short to
1541 have a flag are now skipped.
1543 SC/07 Eximstats V1.49 (Alain Williams)
1544 Added the -emptyok flag.
1546 SC/08 Eximstats V1.50
1547 Fixes for obtaining the IP address from reject messages.
1549 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1550 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1551 whitesspace changes from 4.64-PH/27
1553 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1556 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1557 are found, allow negative numbers in numeric criteria)
1559 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1561 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1562 to match changes made in 4.64-PH/43
1564 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1566 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1567 "domain" is actually an IP address, and force a failure. This locks out
1568 those revolvers/nameservers that support "A-for-A" lookups, in
1569 contravention of the specifications.
1571 PH/48 When a host name was looked up from an IP address, and the subsequent
1572 forward lookup of the name timed out, the host name was left in
1573 $sender_host_name, contrary to the specification.
1575 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1576 restricted to single-key lookups, Exim was not diagnosing an error if
1577 * or *@ was used with a query-style lookup.
1579 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1581 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1582 long ago, but noone interested enough thought of it. Let's just say that
1583 the "1.1" means that there are some new functions that weren't there at
1584 some point in the past.
1586 PH/51 Error processing for expansion failure of helo_data from an smtp
1587 transport during callout processing was broken.
1589 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1590 tested/used via the -bh/-bhc/-bs options.
1592 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1593 bug, fixed in subsequent PCRE releases).
1595 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1596 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1598 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1604 SC/01 Use a glob alias rather than an array ref in eximstats generated
1605 parser. This improves both readability and performance.
1607 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1608 Don't display local sender or destination tables in eximstats unless
1609 there is data to show.
1610 Added average volumes into the eximstats top table text output.
1612 SC/03 Collect data on the number of addresses (recipients) as well
1613 as the number of messages in eximstats.
1615 TF/01 Correct an error in the documentation for the redirect router. Exim
1616 does (usually) call initgroups() when daemonizing.
1618 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1619 with consistent privilege compared to when running as a daemon.
1621 TF/03 Note in the spec that $authenticated_id is not set for local
1622 submissions from trusted users.
1624 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1625 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1627 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1628 by adding some example configuration directives to the default
1629 configuration file. A little bit of work is required to uncomment the
1630 directives and define how usernames and passwords are checked, but
1631 there is now a framework to start from.
1633 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1634 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1635 without this. I don't know how relevant this is to other LDAP libraries.
1637 PH/02 Add the verb name to the "unknown ACL verb" error.
1639 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1641 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1643 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1644 directory not expanded when it should be if an expanded home directory
1645 was set for the address (which is overridden by the transport).
1647 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1650 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1651 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1652 because it is too late at that time, and has no effect.
1654 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1655 security issue with \' (bugzilla #107). I could not use the
1656 PQescapeStringConn() function, because it needs a PGconn value as one of
1659 PH/08 When testing addresses using -bt, indicate those final addresses that
1660 are duplicates that would not cause an additional delivery. At least one
1661 person was confused, thinking that -bt output corresponded to deliveries.
1662 (Suppressing duplicates isn't a good idea as you lose the information
1663 about possibly different redirections that led to the duplicates.)
1665 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1666 systems where poll() doesn't work, in particular OS X.
1668 PH/10 Added more information to debugging output for retry time not reached.
1670 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1671 operations in malware.c.
1673 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1676 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1677 syslog_duplication set false, log lines that would normally be written
1678 both the the main log and to the reject log were not written to syslog at
1681 PH/14 In the default configuration, change the use of "message" in ACL warn
1682 statements to "add_header".
1684 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1685 not followed by a command (e.g. "seen endif").
1687 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1688 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1691 PH/17 Added extra conditions to the default value of delay_warning_condition
1695 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1696 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1697 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1700 The Auto-Submitted: and various List- headers are standardised, whereas I
1701 don't think Precedence: ever was.
1703 PH/18 Refactored debugging code in route_finduser() to show more information,
1704 in particular, the error code if getpwnam() issues one.
1706 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1707 This is apparently needed in addition to the PH/07 change above to avoid
1708 any possible encoding problems.
1710 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1711 but not after initializing Perl.
1713 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1714 output them only if debugging. By default they are written stderr,
1715 apparently, which is not desirable.
1717 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1720 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1723 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1725 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1726 authenticated or an ident call has been made. Suppress the default
1727 values for $authenticated_id and $authenticated_sender (but permit -oMai
1728 and -oMas) when testing with -bh.
1730 PH/24 Re-jigged the order of the tests in the default configuration so that the
1731 tests for valid domains and recipients precede the DNS black list and CSA
1732 tests, on the grounds that those ones are more expensive.
1734 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1735 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1736 This bug exists in every version of Exim that I still have, right back to
1739 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1740 However, an attempt to turn on case-sensitivity in a regex key by
1741 including (?-i) didn't work because the subject string was already
1742 lowercased, and the effects were non-intuitive. It turns out that a
1743 one-line patch can be used to allow (?-i) to work as expected.
1749 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1750 other effects) broke the use of negated acl sub-conditions.
1752 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1755 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1756 "Deny" causes Exim to reject the incoming connection with a 554 error.
1757 Unfortunately, if there is a major crisis, such as a disk failure,
1758 tcp-wrappers gives "deny", whereas what one would like would be some
1759 kind of temporary error. A kludge has been added to help with this.
1760 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1761 554 error is used if errno is still zero or contains ENOENT (which occurs
1762 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1765 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1767 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1768 errors. Otherwise a message that provokes a temporary error (when other
1769 messages do not) can cause a whole host to time out.
1771 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1772 addresses were routed directly to files or pipes from a redirect router.
1773 File deliveries just didn't batch; pipe deliveries might have suffered
1776 PH/06 A failure to get a lock for a hints database would erroneously always say
1777 "Failed to get write lock", even when it was really a read lock.
1779 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1780 of 0600. This has been changed to use the value of the lockfile_mode
1781 option (which defaults to 0600).
1783 PH/08 Applied small patch from the Sieve maintainer.
1785 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1786 folder from quota calculations, a direct delivery into this folder messed
1787 up the contents of the maildirsize file. This was because the regex was
1788 used only to exclude .Trash (or whatever) when the size of the mailbox
1789 was calculated. There was no check that a delivery was happening into an
1790 excluded directory. This bug has been fixed by ignoring all quota
1791 processing for deliveries into excluded directories.
1793 PH/10 Added the maildirfolder_create_regex option to appendfile.
1799 PH/01 The code for finding all the local interface addresses on a FreeBSD
1800 system running IPv6 was broken. This may well have applied to all BSD
1801 systems, as well as to others that have similar system calls. The broken
1802 code found IPv4 interfaces correctly, but gave incorrect values for the
1803 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1804 that it would not match correctly against @[] and not recognize the IPv6
1807 PH/02 The ipliteral router was not recognizing addresses of the form user@
1808 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1810 PH/03 Added disable_ipv6.
1812 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1813 decoded form, because it is most often used to construct To: headers
1814 lines in autoreplies, and the decoded form may well be syntactically
1815 invalid. However, $reply_address has leading white space removed, and all
1816 newlines turned into spaces so that the autoreply transport does not
1819 PH/05 If group was specified without a user on a router, and no group or user
1820 was specified on a transport, the group from the router was ignored.
1822 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1823 for visible compile-time settings that can be used to change these
1824 numbers, for those that want even more. Backwards compatibility with old
1825 spool files has been maintained. However, going back to a previous Exim
1826 release will lost any variables that are in spool files.
1828 PH/07 Two small changes when running in the test harness: increase delay when
1829 passing a TCP/IP connection to a new process, in case the original
1830 process has to generate a bounce, and remove special handling of
1831 127.0.0.2 (sic), which is no longer necessary.
1833 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1834 be the same on different OS.
1836 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1839 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1840 whether --show-vars was specified or not
1842 JJ/02 exipick: Added support for new ACL variable spool format introduced
1845 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1846 syntactically invalid From: or Reply-to: line, and a filter used this to
1847 generate an autoreply, and therefore failed to obtain an address for the
1848 autoreply, Exim could try to deliver to a non-existent relative file
1849 name, causing unrelated and misleading errors. What now happens is that
1850 it logs this as a hard delivery error, but does not attempt to create a
1853 PH/11 The exinext utility has a -C option for testing purposes, but although
1854 the given file was scanned by exinext itself; it wasn't being passed on
1855 when Exim was called.
1857 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1858 an end-of-file indication when reading a command response.
1860 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1861 compiled. In many other places in Exim, IPv6 addresses are always
1862 recognized, so I have changed this. It also means that IPv4 domain
1863 literals of the form [IPV4:n.n.n.n] are now always recognized.
1865 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1866 used if the router is not running as root, for example, when verifying at
1867 ACL time, or when using -bh. The debugging output from this situation was
1868 non-existent - all you got was a failure to exec. I have made two
1871 (a) Failures to set uid/gid, the current directory, or a process leader
1872 in a subprocess such as that created by queryprogram now generate
1873 suitable debugging ouput when -d is set.
1875 (b) The queryprogram router detects when it is not running as root,
1876 outputs suitable debugging information if -d is set, and then runs
1877 the subprocess without attempting to change uid/gid.
1879 PH/15 Minor change to Makefile for building test_host (undocumented testing
1882 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1883 additional section of a DNS packet that returns MX or SRV records.
1884 Instead, it always explicitly searches for A/AAAA records. This avoids
1885 major problems that occur when a DNS server includes only records of one
1886 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1887 fixed another bug: if SRV records were looked up and the corresponding
1888 address records were *not* found in the additional section, the port
1889 values from the SRV records were lost.
1891 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1892 using the correct key (the original address) when searching the retry
1893 rules in order to find which one to use for generating the retry hint.
1895 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1896 adding the default one. Similarly, if it contains a Reply-To: header, the
1897 errors_reply_to option, if set, is not used.
1899 PH/19 When calculating a retry time, Exim used to measure the "time since
1900 failure" by looking at the "first failed" field in the retry record. Now
1901 it does not use this if it is later than than the arrival time of the
1902 message. Instead it uses the arrival time. This makes for better
1903 behaviour in cases where some deliveries succeed, thus re-setting the
1904 "first failed" field. An example is a quota failure for a huge message
1905 when small messages continue to be delivered. Without this change, the
1906 "time since failure" will always be short, possible causing more frequent
1907 delivery attempts for the huge message than are intended.
1908 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1910 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1911 $1, $2, $3) because the numerical variables can be reset during some
1912 expansion items (e.g. "match"), thereby losing the authentication data.
1914 PH/21 Make -bV show the size of off_t variables so that the test suite can
1915 decide whether to run tests for quotas > 2G.
1917 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1918 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1919 filecount value is greater than 2G or if a quota value is greater than 2G
1920 on a system where the size of off_t is not greater than 4, a panic error
1923 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1924 never match. The debug and -bh output now contains an explicit error
1925 message indicating a malformed IPv4 address or mask.
1927 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1928 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1929 PH/23 above applies.
1931 PH/25 Do not write to syslog when running in the test harness. The only
1932 occasion when this arises is a failure to open the main or panic logs
1933 (for which there is an explicit test).
1935 PH/26 Added the /no_tell option to "control=freeze".
1937 PH/27 If a host name lookup failed very early in a connection, for example, if
1938 the IP address matched host_lookup and the reverse lookup yielded a name
1939 that did not have a forward lookup, an error message of the form "no IP
1940 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1941 could be logged. Now it outputs the IP address instead of "NULL".
1943 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1944 allows the sender and the authenticated sender to be set when
1945 submitting a message from within Exim. Since child_open_exim() is
1946 documented for local_scan(), the new function should be too.
1948 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1949 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1950 results in an empty string is now treated as unset.
1952 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1954 PH/31 Added sender_verify_fail logging option.
1956 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1957 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1958 tidied the source and removed it altogether.
1960 PH/33 When a queue run was abandoned because the load average was too high, a
1961 log line was always written; now it is written only if the queue_run log
1962 selector is set. In addition, the log line for abandonment now contains
1963 information about the queue run such as the pid. This is always present
1964 in "start" and "stop" lines but was omitted from the "abandon" line.
1966 PH/34 Omit spaces between a header name and the colon in the error message that
1967 is given when verify = headers_syntax fails (if there are lots of them,
1968 the message gets confusing).
1970 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1971 names, as there are now some PTR records that contain slashes. This check
1972 is only to protect against broken name servers that fall over on strange
1973 characters, so the fact that it applies to all lookups doesn't matter.
1975 PH/36 Now that the new test suite is complete, we can remove some of the
1976 special code in Exim that was needed for the old test suite. For example,
1977 sorting DNS records because real resolvers return them in an arbitrary
1978 order. The new test suite's fake resolver always returns records in the
1981 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1982 bounces) except when queue_only is set, to avoid logging races between
1983 the different processes.
1985 PH/38 Panic-die if .include specifies a non-absolute path.
1987 PH/39 A tweak to the "H" retry rule from its user.
1989 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1990 a label. They prevented compilation on older perls.
1992 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1993 a warning to be raised on newish perls.
1995 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1996 on queue. Changes to match documented behaviour of showing count of
1997 messages matching specified criteria.
1999 PH/40 Changed the default ident timeout from 30s to 5s.
2001 PH/41 Added support for the use of login_cap features, on those BSD systems
2002 that have them, for controlling the resources used by pipe deliveries.
2004 PH/42 The content-scanning code uses fopen() to create files in which to put
2005 message data. Previously it was not paying any attention to the mode of
2006 the files. Exim runs with umask(0) because the rest of the code creates
2007 files with open(), and sets the required mode explicitly. Thus, these
2008 files were ending up world-writeable. This was not a big issue, because,
2009 being within the spool directory, they were not world-accessible. I have
2010 created a function called modefopen, which takes an additional mode
2011 argument. It sets umask(777), creates the file, chmods it to the required
2012 mode, then resets the umask. All the relevant calls to fopen() in the
2013 content scanning code have been changed to use this function.
2015 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2016 to 24 hours. This avoids potential overflow problems when processing G
2017 and H retry rules. I suspect nobody ever tinkers with this value.
2019 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2021 PH/45 When the plaintext authenticator is running as a client, the server's
2022 challenges are checked to ensure they are valid base64 strings. By
2023 default, the authentication attempt is cancelled if an invalid string is
2024 received. Setting client_ignore_invalid_base64 true ignores these errors.
2025 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2026 they are received. Thus, the responses can be made to depend on the
2027 challenges. If an invalid string is ignored, an empty string is placed in
2030 PH/46 Messages that are created by the autoreply transport now contains a
2031 References: header, in accordance with RFCs 2822 and 3834.
2033 PH/47 Added authenticated_sender_force to the smtp transport.
2035 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2037 PH/49 Installed latest patch from the Sieve maintainer.
2039 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2040 was also set, the appendfile transport was unnecessarily scanning a
2041 directory of message files (e.g. for maildir delivery) to find the count
2042 of files (along with the size), even though it did not need this
2043 information. It now does the scan only if it needs to find either the
2044 size of the count of files.
2046 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2048 PH/52 Two bugs concerned with error handling when the smtp transport is
2051 (i) Exim was not creating retry information for temporary errors given
2052 for individual recipients after the DATA command when the smtp transport
2053 was used in LMTP mode. This meant that they could be retried too
2054 frequently, and not timed out correctly.
2056 (ii) Exim was setting the flag that allows error details to be returned
2057 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2058 recipients that were returned after the DATA command.
2060 PH/53 This is related to PH/52, but is more general: for any failing address,
2061 when detailed error information was permitted to be returned to the
2062 sender, but the error was temporary, then after the final timeout, only
2063 "retry timeout exceeded" was returned. Now it returns the full error as
2064 well as "retry timeout exceeded".
2066 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2067 do this, and (what is worse) MTAs that accept it.
2069 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2070 will now be deprecated.
2072 PH/56 New os.c-cygwin from the Cygwin maintainer.
2074 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2075 formats (previously only available in exim formats via -bpr, -bpru,
2076 and -bpra. Now also available in native and exiqgrep formats)
2078 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2079 with very large, slow to parse queues
2081 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2083 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2085 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2086 responses to authentication challenges, though it was showing the
2087 challenges; (ii) I've removed the CR characters from the debug output for
2090 PH/58 Allow for the insertion of a newline as well as a space when a string
2091 is turned into more than one encoded-word during RFC 2047 encoding. The
2092 Sieve code now uses this.
2094 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2095 data_4xx, lost_connection, tls_required.
2097 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2098 message was being sent as an SMTP response.
2100 PH/61 Add -l and -k options to exicyclog.
2102 PH/62 When verifying, if an address was redirected to one new address, so that
2103 verification continued, and the new address failed or deferred after
2104 having set something in $address_data, the value of $address_data was not
2105 passed back to the ACL. This was different to the case when no
2106 redirection occurred. The value is now passed back in both cases.
2108 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2109 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2110 use login_cap.h, so on its own it isn't the distinguishing feature. The
2111 new name refers directly to the setclassresources() function.
2113 PH/65 Added configuration files for NetBSD3.
2115 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2117 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2118 is preferred over IPv4.
2120 PH/68 The bounce_return_message and bounce_return_body options were not being
2121 honoured for bounces generated during the reception of non-SMTP messages.
2122 In particular, this applied to messages rejected by the ACL. This bug has
2123 been fixed. However, if bounce_return_message is true and bounce_return_
2124 body is false, the headers that are returned for a non-SMTP message
2125 include only those that have been read before the error was detected.
2126 (In the case of an ACL rejection, they have all been read.)
2128 PH/69 The HTML version of the specification is now built in a directory called
2129 spec_html instead of spec.html, because the latter looks like a path with
2130 a MIME-type, and this confuses some software.
2132 PH/70 Catch two compiler warnings in sieve.c.
2134 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2135 function verify_get_ident() calls ip_connect() to connect a socket, but
2136 if the "connect()" function timed out, ip_connect() used to close the
2137 socket. However, verify_get_ident() also closes the socket later, and in
2138 between Exim writes to the log, which may get opened at this point. When
2139 the socket was closed in ip_connect(), the log could get the same file
2140 descriptor number as the socket. This naturally causes chaos. The fix is
2141 not to close the socket in ip_connect(); the socket should be closed by
2142 the function that creates it. There was only one place in the code where
2143 this was missing, in the iplookup router, which I don't think anybody now
2144 uses, but I've fixed it anyway.
2146 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2147 well as to direct DNS lookups. Otherwise the handling of names in host
2148 lists is inconsistent and therefore confusing.
2154 PH/01 Two changes to the default runtime configuration:
2156 (1) Move the checks for relay_from_hosts and authenticated clients from
2157 after to before the (commented out) DNS black list checks.
2159 (2) Add control=submission to the relay_from_hosts and authenticated
2160 clients checks, on the grounds that messages accepted by these
2161 statements are most likely to be submissions.
2163 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2165 (1) Generate an error if the third argument for the ${prvs expansion is
2168 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2171 (3) Reset the variables that are obtained from the first argument of
2172 ${prvscheck and used in the second argument before leaving the code,
2173 because their memory is reclaimed, so using them afterwards may do
2176 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2177 one (it's much easier than Tom thought :-).
2179 (5) Because of (4), we can now allow for the use of $prvscheck_result
2180 inside the third argument.
2182 PH/03 For some reason, the default setting of PATH when running a command from
2183 a pipe transport was just "/usr/bin". I have changed it to
2186 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2187 anything to be listed in the output from -bV.
2189 PH/05 When a filter generated an autoreply, the entire To: header line was
2190 quoted in the delivery log line, like this:
2192 => >A.N.Other <ano@some.domain> <original@ddress> ...
2194 This has been changed so that it extracts the operative address. There
2195 may be more than one such address. If so, they are comma-separated, like
2198 => >ano@some.domain,ona@other.domain <original@ddress> ...
2200 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2201 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2202 not being looked up in the rDNS to get a host name, Exim was showing the
2203 IP address twice in Received: lines, even though the IP addresses were
2204 identical. For example:
2206 Received: from [1.2.3.4] (helo=[1.2.3.4])
2208 However, if the real host name was known, it was omitting the HELO data
2209 if it matched the actual IP address. This has been tidied up so that it
2210 doesn't show the same IP address twice.
2212 PH/07 When both +timestamp and +memory debugging was on, the value given by
2213 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2214 called by the debug printing, thereby overwriting the timestamp buffer.
2215 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2217 PH/08 When the original message was included in an autoreply transport, it
2218 always said "this is a copy of the message, including all the headers",
2219 even if body_only or headers_only was set. It now gives an appropriate
2222 PH/09 Applied a patch from the Sieve maintainer which:
2224 o fixes some comments
2225 o adds the (disabled) notify extension core
2226 o adds some debug output for the result of if/elsif tests
2227 o points to the current vacation draft in the documentation
2228 and documents the missing references header update
2232 o fixes a bug in processing the envelope test (when testing
2233 multiple envelope elements, the last element determinted the
2236 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2237 Electronic Mail") by including:
2239 Auto-submitted: auto-generated
2241 in the messages that it generates (bounce messages and others, such as
2242 warnings). In the case of bounce messages for non-SMTP mesages, there was
2243 also a typo: it was using "Auto_submitted" (underscore instead of
2244 hyphen). Since every message generated by Exim is necessarily in response
2245 to another message, thes have all been changed to:
2247 Auto-Submitted: auto-replied
2249 in accordance with these statements in the RFC:
2251 The auto-replied keyword:
2253 - SHOULD be used on messages sent in direct response to another
2254 message by an automatic process,
2256 - MUST NOT be used on manually-generated messages,
2258 - MAY be used on Delivery Status Notifications (DSNs) and Message
2259 Disposition Notifications (MDNs),
2261 - MUST NOT be used on messages generated by automatic or periodic
2262 processes, except for messages which are automatic responses to
2265 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2266 to the default Received: header definition.
2268 PH/12 Added log selector acl_warn_skipped (default on).
2270 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2271 variables because (a) they are in the wrong storage pool and (b) even if
2272 they were copied, it wouldn't work properly because of the caching.
2274 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2275 checking when decoding. Apparently there are clients that generate
2276 overlong encoded strings. Why am I not surprised?
2278 PH/15 If the first argument of "${if match_address" was not empty, but did not
2279 contain an "@" character, Exim crashed. Now it writes a panic log message
2280 and treats the condition as false.
2282 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2284 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2285 extension "envelope-auth". The code is finished and in agreement with
2286 other implementations, but there is no documentation so far and in fact,
2287 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2288 not changing the active code.
2290 Print executed "if" and "elsif" statements when debugging is used. This
2291 helps a great deal to understand what a filter does.
2293 Document more things not specified clearly in RFC3028. I had all this
2294 sorted out, when out of a sudden new issues came to my mind. Oops."
2296 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2299 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2300 canonical form (as documented). However, after a host name lookup from
2301 the IP address, check_host() was doing a simple string comparison with
2302 addresses acquired from the DNS when checking that the found name did
2303 have the original IP as one of its addresses. Since any found IPv6
2304 addresses are likely to be in abbreviated form, the comparison could
2305 fail. Luckily, there already exists a function for doing the comparison
2306 by converting both addresses to binary, so now that is used instead of
2307 the text comparison.
2309 PH/20 There was another similar case to PH/19, when a complete host name was
2310 given in a host list; looking up its IP address could give an abbreviated
2311 form, whereas the current host's name might or might not be abbreviated.
2312 The same fix has been applied.
2318 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2319 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2322 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2323 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2325 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2327 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2328 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2329 Also, for "Auto-Submitted", treat anything other than "no" as
2330 non-personal, in accordance with RFC 3834. (Previously it treated
2331 anything starting "auto-" as non-personal.)
2333 TF/01 The control=submission/name=... option had a problem with syntax
2334 errors if the name included a slash character. The /name= option
2335 now slurps the rest of the string, so it can include any characters
2336 but it must come last in the list of options (after /sender_retain
2339 PH/05 Some modifications to the interface to the fake nameserver for the new
2347 TK/01 Added the "success_on_redirect" address verification option. See
2348 NewStuff for rationale and an example.
2350 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2352 PH/02 Patch to exigrep to allow it to work on syslog lines.
2354 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2355 fread() to skip over the body file's header line, because in Cygwin the
2356 header line is locked and is inaccessible.
2358 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2359 co-exist for some time) to make it clear that it is the Exim ID that is
2360 referenced, not the Message-ID: header line.
2362 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2363 string_format() function, because snprintf() does not exist on all
2366 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2367 ${stat: expansion item.
2369 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2370 protocol synchronization error", to keep the pedants happy.
2372 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2373 well as for IRIX systems, when gcc is being used. See the host.c source
2376 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2378 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2381 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2382 transport and to the smtp transport in LMTP mode.
2384 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2386 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2387 run a filter in a subprocess. This could lead to confusion in subsequent
2388 lookups in the parent process. There should also be a search_tidyup() at
2389 the end of the subprocess.
2391 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2392 only if the host matched helo_try_verify_hosts, which caused the
2393 verification to occur when the EHLO/HELO command was issued. The ACL just
2394 tested the remembered result. Now, if a previous verification attempt has
2395 not happened, "verify = helo" does it there and then.
2397 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2399 TK/03 Fix log output including CR from clamd.
2401 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2402 exist provoked a memory error which could cause a segfault.
2404 PH/15 Installed PCRE 6.2
2406 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2408 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2409 of the problem. Specifically, suggested +O2 rather than +O1 for the
2412 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2414 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2415 record, the port was not correctly logged when the outgoing_port log
2416 selector was set (it logged the transort's default port).
2418 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2419 fallback_hosts, and "hosts" in the smtp transport.
2421 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2422 host errors such as "Connection refused".
2424 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2425 authentication with radiusclient 0.4.9:
2427 - Error returned from rc_read_config was caught wrongly
2428 - Username/password not passed on to radius server due to wrong length.
2430 The presumption is that some radiusclient API changes for 4.51/PH/17
2431 were not taken care of correctly. The code is still untested by me (my
2432 Linux distribution still has 0.3.2 of radiusclient), but it was
2433 contributed by a Radius user.
2435 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2436 expanding the "port" option of the smtp transport.
2438 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2439 while reading a MIME header. Thanks to Tom Hughes for a patch.
2441 PH/24 Include config.h inside local_scan.h so that configuration settings are
2444 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2445 that in an ACL for RCPT (for example), you can examine exactly what was
2448 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2449 commands, but it was not correctly comparing the address with the actual
2450 client host address. Thus, it would show the EHLO address in Received:
2451 header lines when this was not necessary.
2453 PH/27 Added the % operator to ${eval:}.
2455 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2456 it should be ignoring failures (because with -C, for example, it has lost
2457 privilege). It wasn't ignoring creation failures other than "already
2460 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2463 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2464 because an input line was too long, either on its own, or by virtue of
2465 too many continuations, the temporary file was not being removed, and the
2466 return code was incorrect.
2468 PH/31 Missing "BOOL" in function definition in filtertest.c.
2470 PH/32 Applied Sieve patches from the maintainer.
2472 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2474 PH/33 Added "verify = not_blind".
2476 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2477 Local/Makefile (with some defaults set). These are used in built scripts
2478 such as exicyclog, but they have never been used in the exim_install
2479 script (though there are many overriding facilities there). I have
2480 arranged that the exim_install script now takes note of these two
2483 PH/35 Installed configuration files for Dragonfly.
2485 PH/36 When a locally submitted message by a trusted user did not contain a
2486 From: header, and the sender address was obtained from -f or from an SMTP
2487 MAIL command, and the trusted user did not use -F to supply a sender
2488 name, $originator_name was incorrectly used when constructing a From:
2489 header. Furthermore, $originator_name was used for submission mode
2490 messages from external hosts without From: headers in a similar way,
2491 which is clearly wrong.
2493 PH/37 Added control=suppress_local_fixups.
2495 PH/38 When log_selector = +received_sender was set, and the addition of the
2496 sender made the log line's construction buffer exactly full, or one byte
2497 less than full, an overflow happened when the terminating "\n" was
2500 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2501 when the result of a list match is failure because a DNS lookup failed.
2503 PH/40 RM_COMMAND is now used in the building process.
2505 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2506 the "build-* directories that it finds.
2508 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2509 address in a domain literal was a prefix of an interface address.
2511 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2512 when verifying a sender address, unless rewrite_headers is false.
2514 PH/44 Wrote a long comment about why errors_to addresses are verified as
2515 recipients, not senders.
2517 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2518 the ratelimit ACL was added.
2520 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2522 PH/46 Added extra information about PostgreSQL errors to the error string.
2524 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2525 avoiding the need to install special zones in a real server. This is
2526 backwards compatible; if it can't find the fake resolver, it drops back.
2527 Thus, both old and new test suites can be run.
2529 TF/02 Added util/ratelimit.pl
2531 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2532 clock is set back in time.
2534 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2535 Candler <B.Candler@pobox.com>.
2537 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2538 OK for addresses that are the result of redirections.
2540 PH/48 A number of further additions for the benefit of the new test suite,
2541 including a fake gethostbyname() that interfaces to the fake DNS resolver
2544 TF/06 The fix for widen_domains has also been applied to qualify_single and
2545 search_parents which are the other dnslookup options that can cause
2548 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2551 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2553 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2554 service" if it encountered an MX record with an empty target hostname.
2555 The message is now "an MX or SRV record indicated no SMTP service".
2557 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2558 if the DNS of the sending site is misconfigured. This is quite a
2559 common situation. This change restores the behaviour of treating a
2560 helo verification defer as a failure.
2562 PH/51 If self=fail was set on a router, the bounce message did not include the
2563 actual error message.
2569 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2571 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2572 command itself ended in error, the underlying message about the transport
2573 filter timeout was being overwritten with the pipe command error. Now the
2574 underlying error message should be appended to the second error message.
2576 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2578 PH/02 Reduce the amount of output that "make" produces by default. Full output
2579 can still be requested.
2581 PH/03 The warning log line about a condition test deferring for a "warn" verb
2582 was being output only once per connection, rather than after each
2583 occurrence (because it was using the same function as for successful
2584 "warn" verbs). This seems wrong, so I have changed it.
2586 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2587 it should not have, which might have caused a crash in the right
2588 circumstances, but probably never did.
2590 PH/04 Installed a modified version of Tony Finch's patch to make submission
2591 mode fix the return path as well as the Sender: header line, and to
2592 add a /name= option so that you can make the user's friendly name appear
2595 TF/03 Added the control = fakedefer ACL modifier.
2597 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2598 Mark Lowes for thorough testing.
2600 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2602 TK/03 Merged latest SRS patch from Miles Wilton.
2604 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2605 with the definition in sysexits.h (which is #included earlier).
2606 Fortunately, Exim does not actually use EX_OK. The code used to try to
2607 preserve the sysexits.h value, by assumimg that macro definitions were
2608 scanned for macro replacements. I have been disabused of this notion,
2609 so now the code just undefines EX_OK before #including unistd.h.
2611 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2612 in the smtp transport. When a block could not be written in a single
2613 write() function, the timeout was being re-applied to each part-write.
2614 This seems wrong - if the receiver was accepting one byte at a time it
2615 would take for ever. The timeout is now adjusted when this happens. It
2616 doesn't have to be particularly precise.
2618 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2619 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2621 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2622 without a domain if the check to <postmaster@domain> fails.
2624 SC/01 Eximstats: added -xls and the ability to specify output files
2625 (patch written by Frank Heydlauf).
2627 SC/02 Eximstats: use FileHandles for outputing results.
2629 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2631 SC/04 Eximstats: fixed display of large numbers with -nvr option
2633 SC/05 Eximstats: fixed merging of reports with empty tables.
2635 SC/06 Eximstats: added the -include_original_destination flag
2637 SC/07 Eximstats: removed tabs and trailing whitespace.
2639 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2641 TK/06 MBOX spool code: Add real "From " MBOX separator line
2642 so the .eml file is really in mbox format (even though
2643 most programs do not really care). Patch from Alex Miller.
2645 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2646 The latter is generated from $received_to and is only set if the
2647 message has one envelope recipient. SA can use these headers,
2648 obviously out-of-the-box. Patch from Alex Miller.
2650 PH/08 The ${def test on a variable was returning false if the variable's
2651 value was "0", contrary to what the specification has always said!
2652 The result should be true unless the variable is empty.
2654 PH/09 The syntax error of a character other than { following "${if
2655 def:variable_name" (after optional whitespace) was not being diagnosed.
2656 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2657 accidental colon was present, for example, could give incorrect results.
2659 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2660 result is used (not including appendfile, where other changes are about
2663 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2664 This involved changing a lot of size variables from int to off_t. It
2665 should work with maildirs and everything.
2667 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2668 spamd dying while we are connected to it.
2670 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2673 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2674 function for BDB 4.3.
2676 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2678 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2679 the sources for PCRE have been re-arranged and more files are now
2682 PH/15 The code I had for printing potentially long long variables in PH/11
2683 above was not the best (it lost precision). The length of off_t variables
2684 is now inspected at build time, and an appropriate printing format (%ld
2685 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2686 to be "long long int" or "long int". This is needed for the internal
2687 formatting function string_vformat().
2689 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2690 the configuration file to be ":syslog", then the script "guesses" where
2691 the logs files are, rather than using the compiled in default. In our
2692 case the guess is not the same as the compiled default, so the script
2693 suddenly stopped working when I started to use syslog. The patch checks
2694 to see if log_file_path is "". If so, it attempts to read it from exim
2695 with no configuration file to get the compiled in version, before it
2696 falls back to the previous guessing code."
2698 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2699 implementing BATV in an Exim configuration. See NewStuff for the gory
2702 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2703 Makefile that are specific to HP-UX.
2705 PH/18 If the "use_postmaster" option was set for a recipient callout together
2706 with the "random" option, the postmaster address was used as the MAIL
2707 FROM address for the random test, but not for the subsequent recipient
2708 test. It is now used for both.
2710 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2711 patch removes a few documentation additions to RFC 3028, because the
2712 latest draft now contains them. It adds the new en;ascii-case comparator
2713 and a new error check for 8bit text in MIME parts. Comparator and
2714 require names are now matched exactly. I enabled the subaddress
2715 extension, but it is not well tested yet (read: it works for me)."
2717 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2718 rework some of the code of TK/09 above to avoid the hardwired use of
2719 "%lld" and "long long". Replaced the call to snprintf() with a call to
2722 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2723 records point to non-existent hosts", "retry timeout exceeded", and
2724 "retry time not reached for any host after a long failure period".
2726 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2727 experimental DomainKeys support:
2729 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2730 (2) On an error such as an illegally used "control", the wrong name for
2731 the control was given.
2733 These problems did NOT occur unless DomainKeys support was compiled.
2735 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2737 PH/24 Added ${if match_ip condition.
2739 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2740 databases so that it will be absolutely obvious if a crash occurs in the
2741 DB library. This is a regular occurrence (often caused by mis-matched
2744 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2745 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2746 picked up on a user's system that detects such things. There doesn't seem
2747 to be a gcc warning option for this - only an attribute that has to be
2748 put on the function's prototype. It seems that in Fedora Core 4 they have
2749 set this on a number of new functions. No doubt there will be more in due
2752 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2753 specify a transport. However, if an address that was verified by such a
2754 router was the subject of a callout, Exim crashed because it tried to
2755 read the rcpt_include_affixes from the non-existent transport. Now it
2756 just assumes that the setting of that option is false. This bug was
2757 introduced by 4.51/PH/31.
2759 PH/28 Changed -d+all to exclude +memory, because that information is very
2760 rarely of interest, but it makes the output a lot bigger. People tend to
2761 do -d+all out of habit.
2763 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2764 code in os-type was giving problems when libc.so lives in lib64, like on
2767 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2768 aren't the modern standard, and it seems that some systems' include files
2769 don't always have them. Exim was already checking for some of the newer
2770 ones like T_AAAA, and defining it itself. I've added checks for all the
2771 record types that Exim uses.
2773 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2774 not automatically generating a new one, as it is supposed to. This
2775 prevented TLS from working. If the file did exist, but contained invalid
2776 data, a new version was generated, as expected. It was only the case of a
2777 non-existent file that was broken.
2779 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2780 with a change in libdomainkeys > 0.64.
2782 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2783 from DNS. If the selector record carries the flag, it now has
2784 precedence over the domain-wide flag.
2786 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2788 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2789 the use of a port name that isn't defined in /etc/services) occurred, the
2790 message was deferred as in a normal delivery, and thus remained on the
2791 spool, instead of being failed because of the mua_wrapper setting. This
2792 is now fixed, and I tidied up some of the mua_wrapper messages at the
2795 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2796 the messages in a hash of arrays rather than using individual hashes.
2797 This is a bit cleaner and results in dramatic memory savings, albeit
2798 at a slight CPU cost.
2800 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2801 as requested by Marc Sherman.
2803 SC/10 Eximstats: added histograms for user specified patterns as requested
2806 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2808 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2809 fopen() in the content-scanning modules that did not already have it.
2815 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2816 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2818 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2820 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2822 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2823 to test Sieve filters that use "vacation".
2825 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2826 that changes the way the GnuTLS parameters are stored in the cache file.
2827 The new format can be generated externally. For backward compatibility,
2828 if the data in the cache doesn't make sense, Exim assumes it has read an
2829 old-format file, and it generates new data and writes a new file. This
2830 means that you can't go back to an older release without removing the
2833 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2834 work if there are any delivery delays because "one_time" forces the
2835 parent to be marked "delivered", so its unseen clone is never tried
2836 again. For this reason, Exim now forbids the simultaneous setting of
2839 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2840 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2841 entry if you want to know the details. The fix, however, neglected to
2842 consider the case when local delivery batching is involved. The test for
2843 "previously delivered" was not happening when checking to see if an
2844 address could be batched with a previous (undelivered) one; under
2845 certain circumstances this could lead to multiple deliveries to the same
2848 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2849 in its include files, and this causes problems building Exim.
2851 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2852 header_syntax) but Exim was just ignoring anything given after a slash.
2853 In particular, this caused confusion with an attempt to use "verify =
2854 reverse_host_lookup/defer_ok". An error is now given when options are
2855 supplied for verify items that do not have them. (Maybe reverse_host_
2856 lookup should have a defer_ok option, but that's a different point.)
2858 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2859 defined by RFC 821) to 2048, because there were problems with some AUTH
2860 commands, and RFC 1869 says the size should be increased for extended
2861 SMTP commands that take arguments.
2863 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2866 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2867 "unknown" error; now it says that the functionality isn't in the binary.
2869 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2870 an address' error message when a string expansion fails (syntax or
2871 whatever). Otherwise the password may appear in the log. Following change
2872 PH/42 below, there is no longer a chance of it appearing in a bounce
2875 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2877 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2878 of its addresses was ever tried. (Bugzilla bug #2.)
2880 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2881 the result incorrectly in the debug output. (It correctly added a newline
2882 to what was transported.)
2884 TF/01 Added $received_time.
2886 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2887 commented out examples of how to interface to a virus scanner and to
2888 SpamAssassin. Also added commented examples of av_scanner and
2889 spamd_address settings.
2891 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2892 and controls are allowed in which ACLs. There were a couple of minor
2893 errors. Some of the entries in the conditions table (which is a table of
2894 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2895 negation of where the condition IS allowed.
2897 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2899 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2900 header file does not have a version number, so I've had to invent a new
2901 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2902 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2903 radiusclient), but it was contributed by a Radius user.
2905 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2906 files or directories, for OpenSSL.
2908 PH/19 When an Exim process that is running as root has to create an Exim log
2909 file, it does so in a subprocess that runs as exim:exim so as to get the
2910 ownership right at creation (otherwise, other Exim processes might see
2911 the file with the wrong ownership). There was no test for failure of this
2912 fork() call, which would lead to the process getting stuck as it waited
2913 for a non-existent subprocess. Forks do occasionally fail when resources
2914 run out. I reviewed all the other calls to fork(); they all seem to check
2917 PH/20 When checking for unexpected SMTP input at connect time (before writing
2918 the banner), Exim was not dealing correctly with a non-positive return
2919 from the read() function. If the client had disconnected by this time,
2920 the result was a log entry for a synchronization error with an empty
2921 string after "input=" when read() returned zero. If read() returned -1
2922 (an event I could not check), uninitialized data bytes were printed.
2923 There were reports of junk text (parts of files, etc) appearing after
2926 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2928 PH/22 Added support for macro redefinition, and (re)definition in between
2929 driver and ACL definitions.
2931 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2932 forgetting to use the resulting value; it was using the unexpanded value.
2934 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2935 hadn't been configured. The fix is from Juergen Kreileder, who
2936 understands it better than I do:
2938 "Here's what I see happening with three configured cyrus_sasl
2939 authenticators configured (plain, login, cram-md5):
2941 On startup auth_cyrus_sasl_init() gets called for each of these.
2942 This means three calls to sasl_listmech() without a specified mech_list.
2943 => SASL tests which mechs of all available mechs actually work
2944 => three warnings about OTP not working
2945 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2947 With the patch, sasl_listmech() also gets called three times. But now
2948 SASL's mech_list option is set to the server_mech specified in the the
2949 authenticator. Or in other words, the answer from sasl_listmech()
2950 gets limited to just the mech you're testing for (which is different
2952 => the return list contains just 'plain' or 'login', 'cram-md5' or
2953 nothing depending on the value of ob->server_mech.
2955 I've just tested the patch: Authentication still works fine,
2956 unavailable mechs specified in the exim configuration are still
2957 caught, and the auth.log warnings about OTP are gone."
2959 PH/25 When debugging is enabled, the contents of the command line are added
2960 to the debugging output, even when log_selector=+arguments is not
2963 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2964 answer is "GNU", and only if the return is "GNU/something" is the answer
2967 PH/27 $acl_verify_message is now set immediately after the failure of a
2968 verification in an ACL, and so is available in subsequent modifiers. In
2969 particular, the message can be preserved by coding like this:
2971 warn !verify = sender
2972 set acl_m0 = $acl_verify_message
2974 Previously, $acl_verify_message was set only while expanding "message"
2975 and "log_message" when a very denied access.
2977 PH/28 Modified OS/os.c-Linux with
2979 -#ifndef OS_LOAD_AVERAGE
2980 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2982 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2983 nomenclature these days.)
2985 PH/29 Installed patch from the Sieve maintainer that adds the options
2986 sieve_useraddress and sieve_subaddress to the redirect router.
2988 PH/30 In these circumstances:
2989 . Two addresses routed to the same list of hosts;
2990 . First host does not offer TLS;
2991 . First host accepts first address;
2992 . First host gives temporary error to second address;
2993 . Second host offers TLS and a TLS session is established;
2994 . Second host accepts second address.
2995 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2996 and peerdn, if requested) that were in fact used only for the second
2999 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3000 attention to any local part prefix or suffix that was matched by the
3001 router that accepted the address. It now behaves in the same way as it
3002 does for delivery: the affixes are removed from the local part unless
3003 rcpt_include_affixes is set on the transport.
3005 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3006 timeout during the DATA phase of an incoming message.
3008 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3009 applied a patch sanctioned by the Sieve maintainer.
3011 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3012 the uid or gid is negative. A case of a negative gid caused this to be
3013 noticed. The fix allows for either to be negative.
3015 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3016 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3019 PH/36 The change PH/12 above was broken. Fixed it.
3021 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3022 the grounds that routing the same address twice would always produce the
3023 same answer. This might have been true once, but it is certainly no
3024 longer true now. Routing a child address may depend on the previous
3025 routing that produced that child. Some complicated redirection strategies
3026 went wrong when messages had multiple recipients, and made Exim's
3027 behaviour dependent on the order in which the addresses were given.
3029 I have moved the duplicate checking until after the routing is complete.
3030 Exim scans the addresses that are assigned to local and remote
3031 transports, and removes any duplicates. This means that more work will be
3032 done, as duplicates will always all be routed, but duplicates are
3033 presumably rare, so I don't expect this is of any significance.
3035 For deliveries to pipes, files, and autoreplies, the duplicate checking
3036 still happens during the routing process, since they are not going to be
3039 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3040 It corrects a timeout issue with spamd. This is Ian's comment: "The
3041 background is that sometimes spamd either never reads data from a
3042 connection it has accepted, or it never writes response data. The exiscan
3043 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3044 blindly assumes that writes won't block so it may never time out."
3046 PH/39 Allow G after quota size as well as K and M.
3048 PH/40 The value set for $authenticated_id in an authenticator may not contain
3049 binary zeroes or newlines because the value is written to log lines and
3050 to spool files. There was no check on this. Now the value is run through
3051 the string_printing() function so that such characters are converted to
3052 printable escape sequences.
3054 PH/41 $message_linecount is a new variable that contains the total number of
3055 lines in the message. Compare $body_linecount, which is the count for the
3058 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3059 bounce and delay warning messages, except in certain special cases, which
3062 (a) An SMTP error message from a remote host;
3063 (b) A message specified in a :fail: redirection;
3064 (c) A message specified in a "fail" command in a system filter;
3065 (d) A message specified in a FAIL return from the queryprogram router;
3066 (e) A message specified by the cannot_route_message router option.
3068 In these cases only, Exim does include the error details in bounce and
3069 warning messages. There are also a few cases where bland messages such
3070 as "unrouteable address" or "local delivery error" are given.
3072 PH/43 $value is now also set for the "else" part of a ${run expansion.
3074 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3075 being worked on, but at least Exim now implements the latest version to
3078 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3079 process to complete was treated as a delivery failure, a timeout while
3080 writing the message to the pipe was logged, but erroneously treated as a
3081 successful delivery. Such timeouts include transport filter timeouts. For
3082 consistency with the overall process timeout, these timeouts are now
3083 treated as errors, giving rise to delivery failures by default. However,
3084 there is now a new Boolean option for the pipe transport called
3085 timeout_defer, which, if set TRUE, converts the failures into defers for
3086 both kinds of timeout. A transport filter timeout is now identified in
3089 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3090 systems where "make" and "gmake" are different, calling "gmake" at top
3091 level broke things. I've arranged for the value of $(MAKE) to be passed
3092 from the Makefile to this script so that it can call the same version of
3096 A note about Exim versions 4.44 and 4.50
3097 ----------------------------------------
3099 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3100 changes of various kinds. As a consequence, a big documentation update was
3101 needed. This delayed the release for rather longer than seemed good, especially
3102 in the light of a couple of (minor) security issues. Therefore, the changes
3103 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3104 release. So 4.44 and 4.50 are in effect two different branches that both start
3107 I have left the 4.50 change log unchanged; it contains all the changes since
3108 4.43. The change log for 4.44 is below; many of its items are identical to
3109 those for 4.50. This seems to be the most sensible way to preserve the
3110 historical information.
3116 1. Minor wording change to the doc/README.SIEVE file.
3118 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3119 computation of the current number of files was incorrect.
3121 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3122 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3125 4. Give more explanation in the error message when the command for a transport
3126 filter fails to execute.
3128 5. There are several places where Exim runs a non-Exim command in a
3129 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3130 was being done only for the command run by the queryprogram router. It is
3131 now done for all such subprocesses. The other cases are: ${run, transport
3132 filters, and the commands run by the lmtp and pipe transports.
3134 6. Added CONFIGURE_GROUP build-time option.
3136 7. Some older OS have a limit of 256 on the maximum number of file
3137 descriptors. Exim was using setrlimit() to set 1000 as a large value
3138 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3139 systems. I've change it so that if it can't get 1000, it tries for 256.
3141 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3142 was an oversight, and furthermore, ever since the addition of extra
3143 controls (e.g. 4.43/32), the checks on when to allow different forms of
3144 "control" were broken. There should now be diagnostics for all cases when a
3145 control that does not make sense is encountered.
3147 9. Added the /retain_sender option to "control=submission".
3149 10. $recipients is now available in the predata ACL (oversight).
3151 11. Tidy the search cache before the fork to do a delivery from a message
3152 received from the command line. Otherwise the child will trigger a lookup
3153 failure and thereby defer the delivery if it tries to use (for example) a
3154 cached ldap connection that the parent has called unbind on.
3156 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3157 of $address_data from the recipient verification was clobbered by the
3158 sender verification.
3160 13. The value of address_data from a sender verification is now available in
3161 $sender_address_data in subsequent conditions in the ACL statement.
3163 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3165 15. Added a new option "connect=<time>" to callout options, to set a different
3168 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3169 was its contents. (It was OK if the option was not defined at all.)
3171 17. A "Completed" log line is now written for messages that are removed from
3172 the spool by the -Mrm option.
3174 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3175 information about exactly what failed.
3177 19. Added -dd to debug only the daemon process.
3179 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3180 handles timeouts, both on the server side and network timeouts. Renamed the
3181 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3183 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3184 It is now set to "smtps".
3186 22. $host_address is now set to the target address during the checking of
3187 ignore_target_hosts.
3189 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3190 being passed; this would have caused $sender_host_name to have been used if
3191 matching the list had actually called for a host name (not very likely,
3192 since this list is usually IP addresses). A host name is now passed as
3195 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3196 code that specifies a non-restarting handler (typically sigaction() in
3197 modern systems) in an attempt to fix a rare and obscure crash bug.
3199 25. Narrowed the window for a race in the daemon that could cause it to ignore
3200 SIGCHLD signals. This is not a major problem, because they are used only to
3201 wake it up if nothing else does.
3203 26. A malformed maildirsize file could cause Exim to calculate negative values
3204 for the mailbox size or file count. Odd effects could occur as a result.
3205 The maildirsize information is now recalculated if the size or filecount
3208 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3209 support for a long time. Removed HAVE_SYS_VFS_H.
3211 28. Installed the latest version of exipick from John Jetmore.
3213 29. In an address list, if the pattern was not a regular expression, an empty
3214 subject address (from a bounce message) matched only if the pattern was an
3215 empty string. Non-empty patterns were not even tested. This was the wrong
3216 because it is perfectly reasonable to use an empty address as part of a
3217 database query. An empty address is now tested by patterns that are
3218 lookups. However, all the other forms of pattern expect the subject to
3219 contain a local part and a domain, and therefore, for them, an empty
3220 address still always fails if the pattern is not itself empty.
3222 30. Exim went into a mad DNS loop when attempting to do a callout where the
3223 host was specified on an smtp transport, and looking it up yielded more
3224 than one IP address.
3226 31. Re-factored the code for checking spool and log partition space into a
3227 function that finds that data and another that does the check. The former
3228 is then used to implement four new variables: $spool_space, $log_space,
3229 $spool_inodes, and $log_inodes.
3231 32. The RFC2047 encoding function was originally intended for short strings
3232 such as real names; it was not keeping to the 75-character limit for
3233 encoded words that the RFC imposes. It now respects the limit, and
3234 generates multiple encoded words if necessary. To be on the safe side, I
3235 have increased the buffer size for the ${rfc2047: expansion operator from
3238 33. It is now permitted to omit both strings after an "if" condition; if the
3239 condition is true, the result is "true". As before, when the second string
3240 is omitted, a false condition yields an empty string. This makes it less
3241 cumbersome to write custom ACL and router conditions.
3243 34. Failure to deliver a bounce message always caused it to be frozen, even if
3244 there was an errors_to setting on the router. The errors_to setting is now
3247 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3248 canonical form (fully expanded) before being placed in
3249 $sender_host_address.
3251 36. The table in the code that translates DNS record types into text (T_A to
3252 "A" for instance) was missing entries for NS and CNAME. It is just possible
3253 that this could have caused confusion if both these types were looked up
3254 for the same domain, because the text type is used as part of Exim's
3255 per-process caching. But the chance of anyone hitting this buglet seems
3258 37. The dnsdb lookup has been extended in a number of ways.
3260 (1) There is a new type, "zns", which walks up the domain tree until it
3261 finds some nameserver records. It should be used with care.
3263 (2) There is a new type, "mxh", which is like "mx" except that it returns
3264 just the host names, not the priorities.
3266 (3) It is now possible to give a list of domains (or IP addresses) to be
3267 looked up. The behaviour when one of the lookups defers can be
3268 controlled by a keyword.
3270 (4) It is now possible to specify the separator character for use when
3271 multiple records are returned.
3273 38. The dnslists ACL condition has been extended: it is now possible to supply
3274 a list of IP addresses and/or domains to be looked up in a particular DNS
3277 39. Added log_selector=+queue_time_overall.
3279 40. When running the queue in the test harness, wait just a tad after forking a
3280 delivery process, to get repeatability of debugging output.
3282 41. Include certificate and key file names in error message when GnuTLS fails
3283 to set them up, because the GnuTLS error message doesn't include the name
3284 of the failing file when there is a problem reading it.
3286 42. Allow both -bf and -bF in the same test run.
3288 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3290 44. The "Exiscan patch" is now merged into the mainline Exim source.
3292 45. Sometimes the final signoff response after QUIT could fail to get
3293 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3294 < 0 before doing a fflush(). This bug looks as though it goes back to the
3295 introduction of TLS in release 3.20, but "sometimes" must have been rare
3296 because the tests only now provoked it.
3298 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3299 (this can affect the format of dates).
3301 47. exim_tidydb, when checking for the continued existence of a message for
3302 which it has found a message-specific retry record, was not finding
3303 messages that were in split spool directories. Consequently, it was
3304 deleting retry records that should have stayed in existence.
3306 48. Steve fixed some bugs in eximstats.
3308 49. The SPA authentication driver was not abandoning authentication and moving
3309 on to the next authenticator when an expansion was forced to fail,
3310 contradicting the general specification for all authenticators. Instead it
3311 was generating a temporary error. It now behaves as specified.
3313 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3314 (the order specifies the preference for clients). The order is now AES256,
3315 AES128, 3DES, ARCFOUR128.
3317 51. Small patch to Sieve code - explicitly set From: when generating an
3320 52. Exim crashed if a remote delivery caused a very long error message to be
3321 recorded - for instance if somebody sent an entire SpamAssassin report back
3322 as a large number of 550 error lines. This bug was coincidentally fixed by
3323 increasing the size of one of Exim's internal buffers (big_buffer) that
3324 happened as part of the Exiscan merge. However, to be on the safe side, I
3325 have made the code more robust (and fixed the comments that describe what
3328 53. Now that there can be additional text after "Completed" in log lines (if
3329 the queue_time_overall log selector is set), a one-byte patch to exigrep
3330 was needed to allow it to recognize "Completed" as not the last thing in
3333 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3334 patch that reportedly fixes this has been added. I am not expert enough to
3335 create a test for it. This is what the patch creator wrote:
3337 "I found a little strange behaviour of ldap code when working with
3338 Windows 2003 AD Domain, where users was placed in more than one
3339 Organization Units. When I tried to give exim partial DN, the exit code
3340 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3341 But simultaneously result of request was absolutely normal ldap result,
3342 so I produce this patch..."
3344 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3345 REFERENCE, so I have modified the code to exclude the patch when that macro
3348 55. Some experimental protocols are using DNS PTR records for new purposes. The
3349 keys for these records are domain names, not reversed IP addresses. The
3350 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3351 leaves it alone. Component reversal etc. now happens only for IP addresses.
3354 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3356 57. Double the size of the debug message buffer (to 2048) so that more of very
3357 long debug lines gets shown.
3359 58. The exicyclog utility now does better if the number of log files to keep
3360 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3362 59. Two changes related to the smtp_active_hostname option:
3364 (1) $smtp_active_hostname is now available as a variable.
3365 (2) The default for smtp_banner uses $smtp_active_hostname instead
3366 of $primary_hostname.
3368 60. The host_aton() function is supposed to be passed a string that is known
3369 to be a valid IP address. However, in the case of IPv6 addresses, it was
3370 not checking this. This is a hostage to fortune. Exim now panics and dies
3371 if the condition is not met. A case was found where this could be provoked
3372 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3373 components; fortuitously, this particular loophole had already been fixed
3374 by change 4.50/55 above.
3376 If there are any other similar loopholes, the new check in host_aton()
3377 itself should stop them being exploited. The report I received stated that
3378 data on the command line could provoke the exploit when Exim was running as
3379 exim, but did not say which command line option was involved. All I could
3380 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3381 running as the user.
3384 61. There was a buffer overflow vulnerability in the SPA authentication code
3385 (which came originally from the Samba project). I have added a test to the
3386 spa_base64_to_bits() function which I hope fixes it.
3389 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3390 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3392 63. The daemon start-up calls getloadavg() while still root for those OS that
3393 need the first call to be done as root, but it missed one case: when
3394 deliver_queue_load_max is set with deliver_drop_privilege. This is
3395 necessary for the benefit of the queue runner, because there is no re-exec
3396 when deliver_drop_privilege is set.
3398 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3399 This has been fixed.
3401 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3402 was in use, was not putting the data itself into the right store pool;
3403 consequently, it could be overwritten for a subsequent message in the same
3404 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3407 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3409 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3410 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3411 the function were treating the return as a boolean value, which happened to
3412 work because 0=false and not-0=true, but is not correct code.
3414 68. The host_aton() function was not handling scoped IPv6 addresses (those
3415 with, for example, "%eth0" on the end) correctly.
3417 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3418 negated items (that is, ~something) in unsigned ints. Some compilers
3419 apparently mutter when there is no cast.
3421 70. If an address verification called from an ACL failed, and did not produce a
3422 user-specific message (i.e. there was only a "system" message), nothing was
3423 put in $acl_verify_message. In this situation, it now puts the system
3426 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3427 session; change 4.31/43 added the unwanted input to the log line - except
3428 that it did not do this in the start of session case. It now does.
3430 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3431 This is wrong and can cause the other end to generate a synchronization
3432 error if it is another Exim or anything else that does the synchronization
3433 check. A QUIT command is no longer sent after a timeout.
3435 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3436 during host lookups.
3438 74. The defer_ok option of callout verification was not working if it was used
3439 when verifying addresses in header lines, that is, for this case:
3441 verify = header_sender/callout=defer_ok
3443 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3444 those file descriptors could be used for SMTP connections. If anything
3445 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3446 could be sent to the SMTP client, causing chaos. The daemon now opens
3447 stdin, stdout, and stderr to /dev/null when it puts itself into the
3450 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3451 log by default. The user can override this with suitable Perl magic.
3453 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3454 the log message when discard triggers, was not working for the DATA ACL or
3455 for the non-SMTP ACL.
3457 78. Error message wording change in sieve.c.
3459 79. If smtp_accept_max_per_host was set, the number of connections could be
3460 restricted to fewer than expected, because the daemon was trying to set up
3461 a new connection before checking whether the processes handling previous
3462 connections had finished. The check for completed processes is now done
3463 earlier. On busy systems, this bug wouldn't be noticed because something
3464 else would have woken the daemon, and it would have reaped the completed
3467 80. If a message was submitted locally by a user whose login name contained one
3468 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3469 It caused a spool format error. I have fixed the spool reading code. A
3470 related problem was that the "from" clause in the Received: line became
3471 illegal because of the space(s). It is now covered by ${quote_local_part.
3473 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3476 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3478 83. Patch from Sieve maintainer to fix unterminated string problem in
3479 "vacation" handling.
3481 84. Some minor changes to the Linux configuration files to help with other
3482 OS variants using glibc.
3484 85. One more patch for Sieve to update vacation handling to latest spec.
3487 ----------------------------------------------------
3488 See the note above about the 4.44 and 4.50 releases.
3489 ----------------------------------------------------
3495 1. Change 4.43/35 introduced a bug that caused file counts to be
3496 incorrectly computed when quota_filecount was set in an appendfile
3499 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3500 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3503 3. Give more explanation in the error message when the command for a transport
3504 filter fails to execute.
3506 4. There are several places where Exim runs a non-Exim command in a
3507 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3508 was being done only for the command run by the queryprogram router. It is
3509 now done for all such subprocesses. The other cases are: ${run, transport
3510 filters, and the commands run by the lmtp and pipe transports.
3512 5. Some older OS have a limit of 256 on the maximum number of file
3513 descriptors. Exim was using setrlimit() to set 1000 as a large value
3514 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3515 systems. I've change it so that if it can't get 1000, it tries for 256.
3517 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3518 was an oversight, and furthermore, ever since the addition of extra
3519 controls (e.g. 4.43/32), the checks on when to allow different forms of
3520 "control" were broken. There should now be diagnostics for all cases when a
3521 control that does not make sense is encountered.
3523 7. $recipients is now available in the predata ACL (oversight).
3525 8. Tidy the search cache before the fork to do a delivery from a message
3526 received from the command line. Otherwise the child will trigger a lookup
3527 failure and thereby defer the delivery if it tries to use (for example) a
3528 cached ldap connection that the parent has called unbind on.
3530 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3531 of $address_data from the recipient verification was clobbered by the
3532 sender verification.
3534 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3535 was its contents. (It was OK if the option was not defined at all.)
3537 11. A "Completed" log line is now written for messages that are removed from
3538 the spool by the -Mrm option.
3540 12. $host_address is now set to the target address during the checking of
3541 ignore_target_hosts.
3543 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3544 being passed; this would have caused $sender_host_name to have been used if
3545 matching the list had actually called for a host name (not very likely,
3546 since this list is usually IP addresses). A host name is now passed as
3549 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3550 code that specifies a non-restarting handler (typically sigaction() in
3551 modern systems) in an attempt to fix a rare and obscure crash bug.
3553 15. Narrowed the window for a race in the daemon that could cause it to ignore
3554 SIGCHLD signals. This is not a major problem, because they are used only to
3555 wake it up if nothing else does.
3557 16. A malformed maildirsize file could cause Exim to calculate negative values
3558 for the mailbox size or file count. Odd effects could occur as a result.
3559 The maildirsize information is now recalculated if the size or filecount
3562 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3563 support for a long time. Removed HAVE_SYS_VFS_H.
3565 18. Updated exipick to current release from John Jetmore.
3567 19. Allow an empty sender to be matched against a lookup in an address list.
3568 Previously the only cases considered were a regular expression, or an
3571 20. Exim went into a mad DNS lookup loop when doing a callout where the
3572 host was specified on the transport, if the DNS lookup yielded more than
3575 21. The RFC2047 encoding function was originally intended for short strings
3576 such as real names; it was not keeping to the 75-character limit for
3577 encoded words that the RFC imposes. It now respects the limit, and
3578 generates multiple encoded words if necessary. To be on the safe side, I
3579 have increased the buffer size for the ${rfc2047: expansion operator from
3582 22. Failure to deliver a bounce message always caused it to be frozen, even if
3583 there was an errors_to setting on the router. The errors_to setting is now
3586 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3587 canonical form (fully expanded) before being placed in
3588 $sender_host_address.
3590 24. Updated eximstats to version 1.33
3592 25. Include certificate and key file names in error message when GnuTLS fails
3593 to set them up, because the GnuTLS error message doesn't include the name
3594 of the failing file when there is a problem reading it.
3596 26. Expand error message when OpenSSL has problems setting up cert/key files.
3599 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3600 (this can affect the format of dates).
3602 28. exim_tidydb, when checking for the continued existence of a message for
3603 which it has found a message-specific retry record, was not finding
3604 messages that were in split spool directories. Consequently, it was
3605 deleting retry records that should have stayed in existence.
3607 29. eximstats updated to version 1.35
3608 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3609 1.35 - bugfix such that pie charts by volume are generated correctly
3611 30. The SPA authentication driver was not abandoning authentication and moving
3612 on to the next authenticator when an expansion was forced to fail,
3613 contradicting the general specification for all authenticators. Instead it
3614 was generating a temporary error. It now behaves as specified.
3616 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3617 (the order specifies the preference for clients). The order is now AES256,
3618 AES128, 3DES, ARCFOUR128.
3620 31. Small patch to Sieve code - explicitly set From: when generating an
3623 32. Exim crashed if a remote delivery caused a very long error message to be
3624 recorded - for instance if somebody sent an entire SpamAssassin report back
3625 as a large number of 550 error lines. This bug was coincidentally fixed by
3626 increasing the size of one of Exim's internal buffers (big_buffer) that
3627 happened as part of the Exiscan merge. However, to be on the safe side, I
3628 have made the code more robust (and fixed the comments that describe what
3631 33. Some experimental protocols are using DNS PTR records for new purposes. The
3632 keys for these records are domain names, not reversed IP addresses. The
3633 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3634 leaves it alone. Component reversal etc. now happens only for IP addresses.
3637 34. The host_aton() function is supposed to be passed a string that is known
3638 to be a valid IP address. However, in the case of IPv6 addresses, it was
3639 not checking this. This is a hostage to fortune. Exim now panics and dies
3640 if the condition is not met. A case was found where this could be provoked
3641 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3642 components; fortuitously, this particular loophole had already been fixed
3643 by change 4.50/55 or 4.44/33 above.
3645 If there are any other similar loopholes, the new check in host_aton()
3646 itself should stop them being exploited. The report I received stated that
3647 data on the command line could provoke the exploit when Exim was running as
3648 exim, but did not say which command line option was involved. All I could
3649 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3650 running as the user.
3653 35. There was a buffer overflow vulnerability in the SPA authentication code
3654 (which came originally from the Samba project). I have added a test to the
3655 spa_base64_to_bits() function which I hope fixes it.
3658 36. The daemon start-up calls getloadavg() while still root for those OS that
3659 need the first call to be done as root, but it missed one case: when
3660 deliver_queue_load_max is set with deliver_drop_privilege. This is
3661 necessary for the benefit of the queue runner, because there is no re-exec
3662 when deliver_drop_privilege is set.
3664 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3665 was in use, was not putting the data itself into the right store pool;
3666 consequently, it could be overwritten for a subsequent message in the same
3667 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3670 38. Sometimes the final signoff response after QUIT could fail to get
3671 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3672 < 0 before doing a fflush(). This bug looks as though it goes back to the
3673 introduction of TLS in release 3.20, but "sometimes" must have been rare
3674 because the tests only now provoked it.
3680 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3681 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3682 Now it returns an integer. A number of calls were still expecting a T/F
3683 return. Fortuitously, in all cases, the tests worked in OK situations,
3684 which is the norm. However, things would have gone wrong on any write
3685 failures on the smtp file descriptor. This function is used when sending
3686 messages over SMTP and also when doing verify callouts.
3688 2. When Exim is called to do synchronous delivery of a locally submitted
3689 message (the -odf or -odi options), it no longer closes stderr before doing
3692 3. Implemented the mua_wrapper option.
3694 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3696 5. Implemented the functions header_remove(), header_testname(),
3697 header_add_at_position(), and receive_remove_recipient(), and exported them
3700 6. If an ACL "warn" statement specified the addition of headers, Exim already
3701 inserted X-ACL-Warn: at the start if there was no header name. However, it
3702 was not making this test for the second and subsequent header lines if
3703 there were newlines in the string. This meant that an invalid header could
3704 be inserted if Exim was badly configured.
3706 7. Allow an ACL "warn" statement to add header lines at the start or after all
3707 the Received: headers, as well as at the end.
3709 8. Added the rcpt_4xx retry error code.
3711 9. Added postmaster_mailfrom=xxx to callout verification option.
3713 10. Added mailfrom=xxxx to the callout verification option, for verify=
3716 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3717 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3719 12. Inserted some casts to stop certain compilers warning when using pointer
3720 differences as field lengths or precisions in printf-type calls (mostly
3721 affecting debugging statements).
3723 13. Added optional readline() support for -be (dynamically loaded).
3725 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3726 same clock tick as a message's arrival, so that its received time was the
3727 same as the "first fail" time on the retry record, and that message
3728 remained on the queue past the ultimate address timeout, every queue runner
3729 would try a delivery (because it was past the ultimate address timeout) but
3730 after another failure, the ultimate address timeout, which should have then
3731 bounced the address, did not kick in. This was a "< instead of <=" error;
3732 in most cases the first failure would have been in the next clock tick
3733 after the received time, and all would be well.
3735 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3736 being recognized when the domain list was tested by the match_domain
3737 condition in an expansion string.
3739 16. Added the ${str2b64: operator.
3741 17. Exim was always calling setrlimit() to set a large limit for the number of
3742 processes, without checking whether the existing limit was already
3743 adequate. (It did check for the limit on file descriptors.) Furthermore,
3744 errors from getrlimit() and setrlimit() were being ignored. Now they are
3745 logged to the main and panic logs, but Exim does carry on, to try to do its
3746 job under whatever limits there are.
3748 18. Imported PCRE 5.0.
3750 19. Trivial typo in log message " temporarily refused connection" (the leading
3753 20. If the log selector return_path_on_delivery was set and an address was
3754 redirected to /dev/null, the delivery process crashed because it assumed
3755 that a return path would always be set for a "successful" delivery. In this
3756 case, the whole delivery is bypassed as an optimization, and therefore no
3759 21. Internal re-arrangement: the function for sending a challenge and reading
3760 a response while authentication was assuming a zero-terminated challenge
3761 string. It's now changed to take a pointer and a length, to allow for
3762 binary data in such strings.
3764 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3766 23. Exim was not respecting finduser_retries when seeking the login of the
3767 uid under which it was called; it was always trying 10 times. (The default
3768 setting of finduser_retries is zero.) Also, it was sleeping after the final
3769 failure, which is pointless.
3771 24. Implemented tls_on_connect_ports.
3773 25. Implemented acl_smtp_predata.
3775 26. If the domain in control=submission is set empty, Exim assumes that the
3776 authenticated id is a complete email address when it generates From: or
3777 Sender: header lines.
3779 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3780 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3781 chown and chgrp in /bin and hostname in /usr/bin.
3783 28. Exim was keeping the "process log" file open after each use, just as it
3784 does for the main log. This opens the possibility of it remaining open for
3785 long periods when the USR1 signal hits a daemon. Occasional processlog
3786 errors were reported, that could have been caused by this. Anyway, it seems
3787 much more sensible not to leave this file open at all, so that is what now
3790 29. The long-running daemon process does not normally write to the log once it
3791 has entered its main loop, and it closes the log before doing so. This is
3792 so that log files can straightforwardly be renamed and moved. However,
3793 there are a couple of unusual error situations where the daemon does write
3794 log entries, and I had neglected to close the log afterwards.
3796 30. The text of an SMTP error response that was received during a remote
3797 delivery was being truncated at 512 bytes. This is too short for some of
3798 the long messages that one sometimes sees. I've increased the limit to
3801 31. It is now possible to make retry rules that apply only when a message has a
3802 specific sender, in particular, an empty sender.
3804 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3805 it possible to be selective about when SMTP synchronization is enforced.
3807 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3809 32. Implemented hosts_connection_nolog.
3811 33. Added an ACL for QUIT.
3813 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3816 35. Added mailbox_size and mailbox_filecount to appendfile.
3818 36. Added control = no_multiline_responses to ACLs.
3820 37. There was a bug in the logic of the code that waits for the clock to tick
3821 in the case where the clock went backwards by a substantial amount such
3822 that the microsecond fraction of "now" was more than the microsecond
3823 fraction of "then" (but the whole seconds number was less).
3825 38. Added support for the libradius Radius client library this is found on
3826 FreeBSD (previously only the radiusclient library was supported).
3832 1. When certain lookups returned multiple values in the form name=value, the
3833 quoting of the values was not always being done properly. Specifically:
3834 (a) If the value started with a double quote, but contained no whitespace,
3836 (b) If the value contained whitespace other than a space character (i.e.
3837 tabs or newlines or carriage returns) it was not quoted.
3838 This fix has been applied to the mysql and pgsql lookups by writing a
3839 separate quoting function and calling it from the lookup code. The fix
3840 should probably also be applied to nisplus, ibase and oracle lookups, but
3841 since I cannot test any of those, I have not disturbed their existing code.
3843 2. A hit in the callout cache for a specific address caused a log line with no
3844 reason for rejecting RCPT. Now it says "Previous (cached) callout
3845 verification failure".
3847 3. There was an off-by-one bug in the queryprogram router. An over-long
3848 return line was truncated at 256 instead of 255 characters, thereby
3849 overflowing its buffer with the terminating zero. As well as fixing this, I
3850 have increased the buffer size to 1024 (and made a note to document this).
3852 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3853 when Exim is waiting for an SMTP response from a remote server, Exim
3854 restarts its select() call on the socket, thereby resetting its timeout.
3855 This is not a problem when such interrupts are rare. Somebody set up a cron
3856 job to run exiwhat every 2 minutes, which is less than the normal select()
3857 timeout (5 or 10 minutes). This meant that the select() timeout never
3858 kicked in because it was always reset. I have fixed this by comparing the
3859 time when an interrupt arrives with the time at the start of the first call
3860 to select(). If more time than the timeout has elapsed, the interrupt is
3861 treated as a timeout.
3863 5. Some internal re-factoring in preparation for the addition of Sieve
3864 extensions (by MH). In particular, the "personal" test is moved to a
3865 separate function, and given an option for scanning Cc: and Bcc: (which is
3866 not set for Exim filters).
3868 6. When Exim created an email address using the login of the caller as the
3869 local part (e.g. when creating a From: or Sender: header line), it was not
3870 quoting the local part when it contained special characters such as @.
3872 7. Installed new OpenBSD configuration files.
3874 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3875 try to make them clearer.
3877 9. Callout options, other than the timeout value, were being ignored when
3878 verifying sender addresses in header lines. For example, when using
3880 verify = header_sender/callout=no_cache
3882 the cache was (incorrectly) being used.
3884 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3885 only the Cygwin environment.
3887 11. When return_path_on_delivery was set as a log selector, if different remote
3888 addresses in the same message used different return paths and parallel
3889 remote delivery occurred, the wrong values would sometimes be logged.
3890 (Whenever a remote delivery process finished, the return path value from
3891 the most recently started remote delivery process was logged.)
3893 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3894 lines when AUTH and/or TLS are in use. This is the "received protocol"
3895 field. Exim used to use "asmtp" for authenticated SMTP, without any
3896 indication (in the protocol name) for TLS use. Now it follows the RFC and
3897 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3898 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3899 names appear in log lines as well as in Received: header lines.
3901 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3902 extensions, and comparison tests, and to fix some bugs.
3904 14. Changes to the "personal" filter test:
3906 (1) The test was buggy in that it was just doing the equivalent of
3907 "contains" tests on header lines. For example, if a user's address was
3908 anne@some.where, the "personal" test would incorrectly be true for
3910 To: susanne@some.where
3912 This test is now done by extracting each address from the header in turn,
3913 and checking the entire address. Other tests that are part of "personal"
3914 are now done using regular expressions (for example, to check local parts
3915 of addresses in From: header lines).
3917 (2) The list of non-personal local parts in From: addresses has been
3918 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3919 taken from the Sieve specification recommendations.
3921 (3) If the message contains any header line starting with "List-" it is
3922 treated as non-personal.
3924 (4) The test for "circular" in the Subject: header line has been removed
3925 because it now seems ill-conceived.
3927 15. Minor typos in src/EDITME comments corrected.
3929 16. Installed latest exipick from John Jetmore.
3931 17. If headers_add on a router specified a text string that was too long for
3932 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3933 of string_sprintf() is now avoided.
3935 18. $message_body_size was not set (it was always zero) when running the DATA
3936 ACL and the local_scan() function.
3938 19. For the "mail" command in an Exim filter, no default was being set for
3939 the once_repeat time, causing a random time value to be used if "once" was
3940 specified. (If the value happened to be <= 0, no repeat happened.) The
3941 default is now 0s, meaning "never repeat". The "vacation" command was OK
3942 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3943 (I found it when inspecting the code).
3945 20. There is now an overall timeout for performing a callout verification. It
3946 defaults to 4 times the callout timeout, which applies to individual SMTP
3947 commands during the callout. The overall timeout applies when there is more
3948 than one host that can be tried. The timeout is checked before trying the
3949 next host. This prevents very long delays if there are a large number of
3950 hosts and all are timing out (e.g. when the network connections are timing
3951 out). The value of the overall timeout can be changed by specifying an
3952 additional sub-option for "callout", called "maxwait". For example:
3954 verify = sender/callout=5s,maxwait=20s
3956 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3957 to the end before writing, but this should make it even safer).
3959 22. Exim was forgetting that it had advertised PIPELINING for the second and
3960 subsequent messages on an SMTP connection. It was also not resetting its
3961 memory on STARTTLS and an internal HELO.
3963 23. When Exim logs an SMTP synchronization error within a session, it now
3964 records whether PIPELINING has been advertised or not.
3966 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3967 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3968 rather than long int.
3970 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3972 26. Added the never_mail option to autoreply.
3978 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3979 crash if the getsockname() call failed; this can happen if a connection is
3980 closed very soon after it is established. The problem was simply in the
3981 order in which certain operations were done, causing Exim to try to write
3982 to the SMTP stream before it had set up the file descriptor. The bug has
3983 been fixed by making things happen in the correct order.
3989 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3990 before the connection was closed, thus losing the rejection response.
3992 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3993 some early Solaris releases, but causes trouble in current releases where
3994 socklen_t is defined.
3996 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3999 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4002 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4003 that releasing the top of it at the end releases what was used for sub-
4004 expansions (unless the block got too big). However, discard this block if
4005 the first thing is a variable or header, so that we can use its block when
4006 it is dynamic (useful for very large $message_headers, for example).
4008 6. Lookups now cache *every* query, not just the most recent. A new, separate
4009 store pool is used for this. It can be recovered when all lookup caches are
4010 flushed. Lookups now release memory at the end of their result strings.
4011 This has involved some general refactoring of the lookup sources.
4013 7. Some code has been added to the store_xxx() functions to reduce the amount
4014 of flapping under certain conditions.
4016 8. log_incoming_interface used to affect only the <= reception log lines. Now
4017 it causes the local interface and port to be added to several more SMTP log
4018 lines, for example "SMTP connection from", and rejection lines.
4020 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4022 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4024 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4025 could be overwritten at the end of the current message (or the start of a
4026 new message if it was set in a HELO ACL). The value is now preserved for
4027 the duration of the SMTP connection.
4029 12. If a transport had a headers_rewrite setting, and a matching header line
4030 contained an unqualified address, that address was qualified, even if it
4031 did not match any rewriting rules. The underlying bug was that the values
4032 of the flags that permit the existence of unqualified sender and recipient
4033 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4034 non-local messages, and by -bnq for local messages) were not being
4035 preserved with the message after it was received.
4037 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4038 "next input=" as part of the text comprising the host identity instead of
4039 the correct text. The code was using the same buffer for two different
4040 strings. However, depending on which order the printing function evaluated
4041 its arguments, the bug did not always show up. Under Linux, for example, my
4042 test suite worked just fine.
4044 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4045 doesn't work with some older versions of Perl. It has been changed to "my",
4046 which in any case is probably the better facility to use.
4048 15. A really picky compiler found some instances of statements for creating
4049 error messages that either had too many or two few arguments for the format
4052 16. The size of the buffer for calls to the DNS resolver has been increased
4053 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4054 for addresses that have a lot of PTR records. This alleviates a problem; it
4055 does not fully solve it.
4057 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4058 buffer now truncates the list and logs the incident, which is the same
4059 action as happens when Exim is looking up a host name and its aliases.
4060 Previously in this situation something unpredictable would happen;
4061 sometimes it was "internal error: store_reset failed".
4063 18. If a server dropped the connection unexpectedly when an Exim client was
4064 using GnuTLS and trying to read a response, the client delivery process
4065 crashed while trying to generate an error log message.
4067 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4068 string, for example:
4070 warn message = H1: something\nH2: something
4072 the text was added as a single header line from Exim's point of view
4073 though it ended up OK in the delivered message. However, searching for the
4074 second and subsequent header lines using $h_h2: did not work. This has been
4075 fixed. Similarly, if a system filter added multiple headers in this way,
4076 the routers could not see them.
4078 20. Expanded the error message when iplsearch is called with an invalid key to
4079 suggest using net-iplsearch in a host list.
4081 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4082 ACLs are no longer actually imposed (and a message to that effect is
4085 22. If a "gecos" field in a passwd entry contained escaped characters, in
4086 particular, if it contained a \" sequence, Exim got it wrong when building
4087 a From: or a Sender: header from that name. A second bug also caused
4088 incorrect handling when an unquoted " was present following a character
4089 that needed quoting.
4091 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4092 was not being matched caselessly.
4094 24. Arranged for all hyphens in the exim.8 source to be escaped with
4097 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4098 properly. Recipient callout cache records were still being keyed to include
4099 the sender, even when use_sender was set false. This led to far more
4100 callouts that were necessary. The sender is no longer included in the key
4101 when use_sender is false.
4103 26. Added "control = submission" modifier to ACLs.
4105 27. Added the ${base62d: operator to decode base 62 numbers.
4107 28. dnsdb lookups can now access SRV records.
4109 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4110 the configuration file.
4112 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4113 (-v) mode. This makes the output for a verbose queue run more intelligible.
4115 31. Added a use_postmaster feature to recipient callouts.
4117 32. Added the $body_zerocount variable, containing the number of binary zero
4118 bytes in the message body.
4120 33. The time of last modification of the "new" subdirectory is now used as the
4121 "mailbox time last read" when there is a quota error for a maildir
4124 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4126 35. Added +ignore_unknown as a special item in host lists.
4128 36. Code for decoding IPv6 addresses in host lists is now included, even if
4129 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4130 address was recognized as an IP address, but was then not correctly decoded
4131 into binary, causing unexpected and incorrect effects when compared with
4138 1. Very minor rewording of debugging text in manualroute to say "list of
4139 hosts" instead of "hostlist".
4141 2. If verify=header_syntax was set, and a header line with an unqualified
4142 address (no domain) and a large number of spaces between the end of the
4143 name and the colon was received, the reception process suffered a buffer
4144 overflow, and (when I tested it) crashed. This was caused by some obsolete
4145 code that should have been removed. The fix is to remove it!
4147 3. When running in the test harness, delay a bit after writing a bounce
4148 message to get a bit more predictability in the log output.
4150 4. Added a call to search_tidyup() just before forking a reception process. In
4151 theory, someone could use a lookup in the expansion of smtp_accept_max_
4152 per_host which, without the tidyup, could leave open a database connection.
4154 5. Added the variables $recipient_data and $sender_data which get set from a
4155 lookup success in an ACL "recipients" or "senders" condition, or a router
4156 "senders" option, similar to $domain_data and $local_part_data.
4158 6. Moved the writing of debug_print from before to after the "senders" test
4161 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4162 problems for message scanning, either using a data ACL, or using
4163 local_scan() because the Received: header was not generated till after they
4164 were called (in order to set the time as the time of reception completion).
4165 I have revised the way this works. The header is now generated after the
4166 body is received, but before the ACL or local_scan() are called. After they
4167 are run, the timestamp in the header is updated.
4173 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4174 before starting a queue runner without re-exec. This happened only when
4175 deliver_drop_privilege was set or when the Exim user was set to root. The
4176 effect of the bug was that timeouts during subsequent deliveries caused
4177 crashes instead of being properly handled. The handler is now left at its
4178 default (and expected) setting.
4180 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4181 message, again when deliver_drop_privilege is set or Exim is run as root.
4182 The bug described in (1) was not present in this case, but the tidying up
4183 of the other signals was missing. I have made the two cases consistent.
4185 3. The ignore_target_hosts setting on a manualroute router was being ignored
4186 for hosts that were looked up using the /MX notation.
4188 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4191 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4192 operated on the sender address. After changing the $sender_address to <>
4193 for the sender address verify, Exim was re-instated it as the original
4194 (before rewriting) address, but remembering that it had rewritten it, so it
4195 wasn't rewriting it again. This bug also had the effect of breaking the
4196 sender address verification caching when the sender address was rewritten.
4198 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4199 This has been changed so that if the ip literal address matches
4200 ignore_target_hosts, the router declines.
4202 7. Added expansion conditions match_domain, match_address, and match_local_
4203 part (NOT match_host).
4205 8. The placeholder for the Received: header didn't have a length field set.
4207 9. Added code to Exim itself and to exim_lock to test for a specific race
4208 condition that could lead to file corruption when using MBX delivery. The
4209 issue is with the lockfile that is created in /tmp. If this file is removed
4210 after a process has opened it but before that process has acquired a lock,
4211 there is the potential for a second process to recreate the file and also
4212 acquire a lock. This could lead to two Exim processes writing to the file
4213 at the same time. The added code performs the same test as UW imapd; it
4214 checks after acquiring the lock that its file descriptor still refers to
4215 the same named file.
4217 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4218 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4221 11. Added the smtp_active_hostname option. If used, this will typically be made
4222 to depend on the incoming interface address. Because $interface_address is
4223 not set up until the daemon has forked a reception process, error responses
4224 that can happen earlier (such as "too many connections") no longer contain
4227 12. If an expansion in a condition on a "warn" statement fails because a lookup
4228 defers, the "warn" statement is abandoned, and the next ACL statement is
4229 processed. Previously this caused the whole ACL to be aborted.
4231 13. Added the iplsearch lookup type.
4233 14. Added ident_timeout as a log selector.
4235 15. Added tls_certificate_verified as a log selector.
4237 16. Added a global option tls_require_ciphers (compare the smtp transport
4238 option of the same name). This controls incoming TLS connections.
4240 17. I finally figured out how to make tls_require_ciphers do a similar thing
4241 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4242 before starting the TLS session.
4244 18. Tabs are now shown as \t in -bP output.
4246 19. If the log selector return_path_on_delivery was set, Exim crashed when
4247 bouncing a message because it had too many Received: header lines.
4249 20. If two routers both had headers_remove settings, and the first one included
4250 a superfluous trailing colon, the final name in the first list and the
4251 first name in the second list were incorrectly joined into one item (with a
4252 colon in the middle).
4258 1. Added -C and -D options to the exinext utility, mainly to make it easier
4259 to include in the automated testing, but these could be helpful when
4260 multiple configurations are in use.
4262 2. The exinext utility was not formatting the output nicely when there was
4263 an alternate port involved in the retry record key, nor when there was a
4264 message id as well (for retries that were specific to a specific message
4265 and a specific host). It was also confused by IPv6 addresses, because of
4266 the additional colons they contain. I have fixed the IPv4 problem, and
4267 patched it up to do a reasonable job for IPv6.
4269 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4270 delivery, the log line now contains "pipelined" if PIPELINING was used.
4272 4. An SMTP transport process used to panic and die if the bind() call to set
4273 an explicit outgoing interface failed. This has been changed; it is now
4274 treated in the same way as a connect() failure.
4276 5. A reference to $sender_host_name in the part of a conditional expansion
4277 that was being skipped was still causing a DNS lookup. This no longer
4280 6. The def: expansion condition was not recognizing references to header lines
4281 that used bh_ and bheader_.
4283 7. Added the _cache feature to named lists.
4285 8. The code for checking quota_filecount in the appendfile transport was
4286 allowing one more file than it should have been.
4288 9. For compatibility with Sendmail, the command line option
4296 and sets the incoming protocol and host name (for trusted callers). The
4297 host name and its colon can be omitted when only the protocol is to be set.
4298 Note the Exim already has two private options, -pd and -ps, that refer to
4299 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4300 "s", but I don't think that's a major issue.
4302 10. A number of refactoring changes to the code, none of which should affect
4305 (a) The number of logging options was getting close to filling up the
4306 32-bit word that was used as a bit map. I have split them into two classes:
4307 those that are passed in the argument to log_write(), and those that are
4308 only ever tested independently outside of that function. These are now in
4309 separate 32-bit words, so there is plenty of room for expansion again.
4310 There is no change in the user interface or the logging behaviour.
4312 (b) When building, for example, log lines, the code previously used a
4313 macro that called string_cat() twice, in order to add two strings. This is
4314 not really sufficiently general. Furthermore, there was one instance where
4315 it was actually wrong because one of the argument was used twice, and in
4316 one call a function was used. (As it happened, calling the function twice
4317 did not affect the overall behaviour.) The macro has been replaced by a
4318 function that can join an arbitrary number of extra strings onto a growing
4321 (c) The code for expansion conditions now uses a table and a binary chop
4322 instead of a serial search (which was left over from when there were very
4323 few conditions). Also, it now recognizes conditions like "pam" even when
4324 the relevant support is not compiled in: a suitably worded error message is
4325 given if an attempt is made to use such a condition.
4327 11. Added ${time_interval:xxxxx}.
4329 12. A bug was causing one of the ddress fields not to be passed back correctly
4330 from remote delivery subprocesses. The field in question was not being
4331 subsequently used, so this caused to problems in practice.
4333 13. Added new log selectors queue_time and deliver_time.
4335 14. Might have fixed a bug in maildirsizefile handling that threw up
4336 "unexpected character" debug warnings, and recalculated the data
4337 unnecessarily. In any case, I expanded the warning message to give more
4340 15. Added the message "Restricted characters in address" to the statements in
4341 the default ACL that block characters like @ and % in local parts.
4343 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4344 Three changes have been made:
4346 (a) There was a serious bug; a negative response to MAIL caused the whole
4347 recipient domain to be cached as invalid, thereby blocking all messages
4348 to all local parts at the same domain, from all senders. This bug has
4349 been fixed. The domain is no longer cached after a negative response to
4350 MAIL if the sender used is not empty.
4352 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4355 (c) A new callout option, "use_sender" has been added for people who want
4356 the modified behaviour.
4362 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4365 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4366 indeed breaks things for older releases.
4368 3. Added additional logging to the case where there is a problem reading data
4369 from a filter that is running in a subprocess using a pipe, in order to
4370 try to track down a specific problem.
4372 4. Testing facility fudge: when running in the test harness and attempting
4373 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4374 getting "No route to host". Convert this to a timeout.
4376 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4379 6. Some OS don't have socklen_t but use size_t instead. This affects the
4380 fifth argument of getsockopt() amongst other things. This is now
4381 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4382 can be set for individual OS. I have set it for SunOS5, OSF1, and
4383 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4384 some earlier ones do not.
4386 7. Change 4.30/15 was not doing the test caselessly.
4388 8. The standard form for an IPv6 address literal was being rejected by address
4389 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4390 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4391 this, as well as the form without the "IPv6" on the front (but only when
4392 address literals are enabled, of course).
4394 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4396 10. Exim crashed if a message with an empty sender address specified by -f
4397 encountered a router with an errors_to setting. This could be provoked only
4398 by a command such as
4402 where an empty string was supplied; "<>" did not hit this bug.
4404 11. Installed PCRE release 4.5.
4406 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4407 remained set. It is now erased.
4409 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4410 times from message ids (which are base 36 rather than the normal 62).
4412 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4413 were being counted as actual protocol errors, and logged if the log
4414 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4415 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4416 and DATA following a set of rejected RCPTs do not count as protocol errors.
4417 In other words, Exim assumes they were pipelined, though this may not
4418 actually be the case. Of course, in all cases the client gets an
4419 appropriate error code.
4421 15. If a lookup fails in an ACL condition, a message about the failure may
4422 be available; it is used if testing the ACL cannot continue, because most
4423 such messages specify what the cause of the deferral is. However, some
4424 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4425 that caused an old message to be retained and used if a later statement
4426 caused a defer, replacing the real cause of the deferral.
4428 16. If an IP address had so many PTR records that the DNS lookup buffer
4429 was not large enough to hold them, Exim could crash while trying to process
4430 the truncated data. It now detects and logs this case.
4432 17. Further to 4.21/58, another change has been made: if (and only if) the
4433 first line of a message (the first header line) ends with CRLF, a bare LF
4434 in a subsequent header line has a space inserted after it, so as not to
4435 terminate the header.
4437 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4438 unnecessarily, used atoi() instead of strtol(), and didn't check the
4439 termination when getting file sizes from file names by regex.
4441 19. Completely re-implemented the support for maildirsize files, in the light
4442 of a number of problems with the previous contributed implementation
4443 (4.30/29). In particular:
4445 . If the quota is zero, the maildirsize file is maintained, but no quota is
4448 . If the maildir directory does not exist, it is created before any attempt
4449 to write a maildirsize file.
4451 . The quota value in the file is just a cache; if the quota is changed in
4452 the transport, the new value overrides.
4454 . A regular expression is available for excluding directories from the
4457 20. The autoreply transport checks the characters in options that define the
4458 message's headers; it allows continued headers, but it was checking with
4459 isspace() after an embedded newline instead of explicitly looking for a
4462 21. If all the "regular" hosts to which an address was routed had passed their
4463 expiry times, and had not reached their retry times, the address was
4464 bounced, even if fallback hosts were defined. Now Exim should go on to try
4467 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4468 equivalent code in the SMTP transport. Some hosts send humungous responses
4469 to HELO/EHLO, more than 1024 it seems.
4471 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4472 gives compiler warnings in some environments. I've now done it "properly",
4475 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4476 (because of problems with the built-in one) was declared to return uschar *
4477 instead of char *, causing compiler failure.
4479 25. Fixed a file descriptor leak when processing alias/forward files.
4481 26. Fixed a minor format string issue in dbfn.c.
4483 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4485 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4486 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4489 29. When a sender address is verified, it is cached, to save repeating the test
4490 when there is more than one recipient in a message. However, when the
4491 verification involves a callout, it is possible for different callout
4492 options to be set for different recipients. It is too complicated to keep
4493 track of this in the cache, so now Exim always runs a verification when a
4494 callout is required, relying on the callout cache for the optimization.
4495 The overhead is duplication of the address routing, but this should not be
4498 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4499 to be verified with callout=postmaster, and the main callout worked but the
4500 postmaster check failed, the verification correctly failed. However, if a
4501 subsequent RCPT command asked for sender verification *without* the
4502 postmaster check, incorrect caching caused this verification also to fail,
4505 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4506 it was not caching the DNS options (qualify_single, search_parents) that
4507 were used when the lookup failed. A subsequent lookup with different
4508 options therefore always gave the same answer, though there were cases
4509 where it should not have. (Example: a "domains = !$mx_any" option on a
4510 dnslookup router: the "domains" option is always processed without any
4511 widening, but the router might have qualify_single set.) Now Exim uses the
4512 cached value only when the same options are set.
4514 32. Added John Jetmore's "exipick" utility to the distribution.
4516 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4517 than a timeout (e.g. a certificate is required, and is not provided), an
4518 Exim server now closes the connection immediately. Previously it waited for
4519 the client to close - but if the client is SSL, it seems that they each
4520 wait for each other, leading to a delay before one of them times out.
4522 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4523 maintained 0.8.x compatibility because I don't think many are using it, and
4524 it is clearly obsolete.
4526 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4529 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4530 client certificate was expired. A simple patch fixes this, though I don't
4531 understand the full logic of why the verify callback is called multiple
4534 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4535 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4536 which causes problems with some clients (such as the Certicom SSL Plus
4537 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4538 disables the coutermeasure allowing Eudora to connect."
4540 38. Exim was not checking that a write() to a log file succeeded. This could
4541 lead to Bad Things if a log got too big, in particular if it hit a file
4542 size limit. Exim now panics and dies if it cannot write to a log file, just
4543 as it does if it cannot open a log file.
4545 39. Modified OS/Makefile-Linux so that it now contains
4547 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4549 The two -D definitions ensure that Exim is compiled with large file
4550 support, which makes it possible to handle log files that are bigger than
4553 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4554 instance) a domain was checked against a named list that involved a lookup,
4555 causing $domain_data to be set, then another domain was checked against the
4556 same list, then the first domain was re-checked, the value of $domain_data
4557 after the final check could be wrong. In particular, if the second check
4558 failed, it could be set empty. This bug probably also applied to
4561 41. The strip_trailing_dot option was not being applied to the address given
4562 with the -f command-line option.
4564 42. The code for reading a message's header from the spool was incrementing
4565 $received_count, but never initializing it. This meant that the value was
4566 incorrect (doubled) while delivering a message in the same process in which
4567 it was received. In the most common configuration of Exim, this never
4568 happens - a fresh exec is done - but it can happen when
4569 deliver_drop_privilege is set.
4571 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4572 it now includes up to 150 characters of the unexpected data in the log
4575 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4576 and building data strings. The size of both of these buffers was 10 000
4577 bytes - far larger than anybody would *ever* want, thought I. Needless to
4578 say, somebody hit the limit. I have increased the maximum line length to
4579 20 000 and the maximum data length of concatenated lines to 100 000. I have
4580 also fixed two bugs, because there was no checking on these buffers. Tsk,
4581 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4582 buffer is too small.
4584 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4585 lsearch lookups. Now it does.
4587 46. When parsing a route_list item in a manualroute router, a fixed-length
4588 buffer was used for the list of hosts. I made this 1024 bytes long,
4589 thinking that nobody would ever have a list of hosts that long. Wrong.
4590 Somebody had a whole pile of complicated expansion conditions, and the
4591 string was silently truncated, leading to an expansion error. It turns out
4592 that it is easier to change to an unlimited length (owing to other changes
4593 that have happened since this code was originally written) than to build
4594 structure for giving a limitation error. The length of the item that
4595 expands into the list of hosts is now unlimited.
4597 47. The lsearch lookup could not handle data where the length of text line was
4598 more than 4095 characters. Such lines were truncated, leading to shortened
4599 data being returned. It should now handle lines of any length.
4601 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4602 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4605 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4606 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4607 no longer happens in comment lines. A list of replacements is now placed
4608 at the head of all of the source files, except those whose only change is
4609 to replace PERL_COMMAND in the very first #! line.
4611 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4612 messages on the queue, with a bottom-up merge sort, using code contributed
4613 by Michael Haardt. This should make operations like -bp somewhat faster on
4614 large queues. It won't affect queue runners, except when queue_run_in_order
4617 51. Installed eximstats 1.31 in the distribution.
4619 52. Added support for SRV lookups to the dnslookup router.
4621 53. If an ACL referred to $message_body or $message_body_end, the value was not
4622 reset for any messages that followed in the same SMTP session.
4624 54. The store-handling optimization for building very long strings was not
4625 differentiating between the different store pools. I don't think this
4626 actually made any difference in practice, but I've tidied it.
4628 55. While running the routers to verify a sender address, $sender_address
4629 was still set to the sender address. This is wrong, because when routing to
4630 send a bounce to the sender, it would be empty. Therefore, I have changed
4631 it so that, while verifying a sender address, $sender_address is set to <>.
4632 (There is no change to what happens when verifying a recipient address.)
4634 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4635 target A or AAAA records (if not already returned) without resetting the
4636 qualify_single or search_parents options of the DNS resolver. These are
4637 inappropriate in this case because the targets of MX and SRV records must
4638 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4639 target that, when qualified, matched something in the local domain. These
4640 two options are now turned off when doing these lookups.
4642 57. It seems that at least some releases of Reiserfs (which does not have the
4643 concept of a fixed number of inodes) returns zero and not -1 for the
4644 number of available inodes. This interacted badly with check_spool_inodes,
4645 which assumed that -1 was the "no such thing" setting. What I have done is
4646 to check that the total number of inodes is greater than zero before doing
4647 the test of how many are available.
4649 58. When a "warn" ACL statement has a log_message modifier, the message is
4650 remembered, and not repeated. This is to avoid a lot of repetition when a
4651 message has many recipients that cause the same warning to be written.
4652 Howewer, Exim was preserving the list of already written lines for an
4653 entire SMTP session, which doesn't seem right. The memory is now reset if a
4654 new message is started.
4656 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4657 debugging output unless log_rewrite was also set.
4659 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4660 of (int)(handle) when we know that handle contains (void *)(-1).
4662 61. The Exim daemon panic-logs an error return when it closes the incoming
4663 connection. However "connection reset by peer" seems to be common, and
4664 isn't really an error worthy of noting specially, so that particular error
4667 62. When Exim is trying to find all the local interfaces, it used to panic and
4668 die if the ioctl to get the interface flags failed. However, it seems that
4669 on at least one OS (Solaris 9) it is possible to have an interface that is
4670 included in the list of interfaces, but for which you get a failure error
4671 for this call. This happens when the interface is not "plumbed" into a
4672 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4673 failure of the "get flags" call assumes that the interface is down.
4675 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4676 makes life easier for people who are doing arithmetic on fields extracted
4677 from dates, where you often get leading zeros that should not be
4678 interpreted as octal.
4680 64. Added qualify_domain to the redirect router, to override the global
4683 65. If a pathologically long header line contained very many addresses (the
4684 report of this problem mentioned 10 000) and each of them was rewritten,
4685 Exim could use up a very large amount of memory. (It kept on making new
4686 copies of the header line as it rewrote, and never released the old ones.)
4687 At the expense of a bit more processing, the header rewriting function has
4688 been changed so that it no longer eats memory in this way.
4690 66. The generation of the Received: header has been moved from the time that a
4691 message starts to be received, to the time that it finishes. The timestamp
4692 in the Received: header should now be very close to that of the <= log
4693 line. There are two side-effects of this change:
4695 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4696 logged header lines no longer include the local Received: line, because
4697 it has not yet been created. The same applies to a copy of the message
4698 that is returned to a non-SMTP sender when a message is rejected.
4700 (b) When a filter file is tested using -bf, no additional Received: header
4701 is added to the test message. After some thought, I decided that this
4704 This change does not affect the value of $received_for. It is still set
4705 after address rewriting, but before local_scan() is called.
4707 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4709 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4710 gave an unhelpful panic error message, and a defer error. I have managed to
4711 change this behaviour so that it now rejects any supplied certificate,
4712 which seems right, as the list of acceptable certificates is empty.
4714 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4715 gave an unhelpful defer error. I have not managed to make this reject any
4716 supplied certificates, but the error message it gives is "no certificate
4717 supplied", which is not helpful.
4719 70. exigrep's output now also includes lines that are not associated with any
4720 message, but which match the given pattern. Implemented by a patch from
4721 Martin Sluka, which also tidied up the Perl a bit.
4723 71. Recipient callout verification, like sender verification, was using <> in
4724 the MAIL FROM command. This isn't really the right thing, since the actual
4725 sender may affect whether the remote host accepts the recipient or not. I
4726 have changed it to use the actual sender in the callout; this means that
4727 the cache record is now keyed on a recipient/sender pair, not just the
4728 recipient address. There doesn't seem to be a real danger of callout loops,
4729 since a callout by the remote host to check the sender would use <>.
4730 [SEE ABOVE: changed after hitting problems.]
4732 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4733 temporary errors. However, in the case of such a code being given after
4734 the end of a data transmission (i.e. after ".") Exim was failing to write
4735 a retry record for the message. (Yes, there was some broken host that was
4736 actually sending 8xx at this point.)
4738 73. An unknown lookup type in a host list could cause Exim to panic-die when
4739 the list was checked. (An example that provoked this was putting <; in the
4740 middle of a list instead of at the start.) If this happened during a DATA
4741 ACL check, a -D file could be left lying around. This kind of configuration
4742 error no longer causes Exim to die; instead it causes a defer errror. The
4743 incident is still logged to the main and panic logs.
4745 74. Buglet left over from Exim 3 conversion. The message "too many messages
4746 in one connection" was written to the rejectlog but not the mainlog, except
4747 when address rewriting (yes!) was being logged.
4749 75. Added write_rejectlog option.
4751 76. When a system filter was run not as root (that is, when system_filter_user
4752 was set), the values of the $n variables were not being returned to the
4753 main process; thus, they were not subsequently available in the $sn
4756 77. Added +return_path_on_delivery log selector.
4758 78. A connection timeout was being treated differently from recipients deferred
4759 when testing hosts_max_try with a message that was older than the host's
4760 retry timeout. (The host should not be counted, thus allowing all hosts to
4761 be tried at least once before bouncing.) This may have been the cause of an
4762 occasionally reported bug whereby a message would remain on the queue
4763 longer than the retry timeout, but would be bounced if a delivery was
4764 forced. I say "may" because I never totally pinned down the problem;
4765 setting up timeout/retry tests is difficult. See also the next item.
4767 79. The ultimate address timeout was not being applied to errors that involved
4768 a combination of host plus message (for example, a timeout on a MAIL
4769 command). When an address resolved to a number of possible hosts, and they
4770 were not all tried for each delivery (e.g. because of hosts_max_try), a
4771 message could remain on the queue longer than the retry timeout.
4773 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4776 81. Fixed an obscure SMTP outgoing bug which required at least the following
4777 conditions: (a) there was another message waiting for the same server;
4778 (b) the server returned 5xx to all RCPT commands in the first message so
4779 that the message was not completed; (c) the server dropped the connection
4780 or gave a negative response to the RSET that Exim sends to abort the
4781 transaction. The observed case was a dropped connection after DATA that had
4782 been sent in pipelining mode. That is, the server had advertised PIPELINING
4783 but was not implementing it correctly. The effect of the bug was incorrect
4784 behaviour, such as trying another host, and this could lead to a crash.
4790 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4791 and daemon.c were passed as pointers to ints; they should have been
4792 pointers to socklen_t variables (which are typically unsigned ints).
4794 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4797 3. Fixed a really odd bug that affected only the testing scheme; patching a
4798 certain fixed string in the binary changed the value of another string that
4799 happened to be identical to the end of the original first string.
4801 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4802 name", it returns that address as the IP address. On some operating
4803 systems (e.g. Solaris), it also passes back the IP address string as the
4804 "host name". However, on others (e.g. Linux), it passes back an empty
4805 string. Exim wasn't checking for this, and was changing the host name to an
4806 empty string, assuming it had been canonicized.
4808 5. Although rare, it is permitted to have more than one PTR record for a given
4809 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4810 all the names associated with an address, because they do in Solaris.
4811 However, it seems that they do not in Linux for data that comes from the
4812 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4813 I found this out when I moved to a new Linux workstation and tried to run
4814 the Exim test suite.
4816 To get round this problem I have changed the code so that it now does its
4817 own call to the DNS to look up PTR records when searching for a host name.
4818 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4819 addresses that are only in /etc/hosts are still found.
4821 This behaviour is, however, controlled by an option called host_lookup_
4822 order, which defaults to "bydns:byaddr". If people want to use the other
4823 order, or indeed, just use one or the other means of lookup, they can
4824 specify it in this variable.
4826 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4827 some operating systems, this comes back from gethostbyaddr() as an empty
4828 string, and this is what Exim used to test for. However, it seems that in
4829 other systems, "." is yielded. Exim now tests for this case too.
4831 7. The values of check_spool_space and check_log_space are now held internally
4832 as a number of kilobytes instead of an absolute number of bytes. If a
4833 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4834 kilobyte. This means that much larger values can be stored.
4836 8. Exim monitor: an attempt to get the action menu when not actually pointing
4837 at a message produces an empty menu entitled "No message selected". This
4838 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4839 no entries in it ("Shell widget menu has zero width and/or height"). So I
4840 have added a single, blank menu entry in this case.
4842 9. Added ${quote_local_part.
4844 10. MIME decoding is now applied to the contents of Subject: header lines when
4847 11. Now that a reference to $sender_host_address automatically causes a reverse
4848 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4849 host lookup before query-style lookups in lists that might use this
4850 variable. This has therefore been abolished, and the "net-" prefix is no
4851 longer necessary for query-style lookups.
4853 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4854 have been a typo for LFLAGS, so it has been changed.
4856 13. The install script calls Exim with "-C /dev/null" in order to find the
4857 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4858 to be output. Howeve, since Exim outputs its version number before the
4859 error, it didn't break the script. It just looked ugly. I fixed this by
4860 always allowing "-C /dev/null" if the caller is root.
4862 14. Ignore overlarge ACL variable number when reading spool file - insurance
4863 against a later release with more variables having written the file.
4865 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4866 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4867 this, as well as the form without the "IPv6" on the front.
4869 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4870 OS/Makefile-Darwin file.
4872 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4873 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4875 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4876 variable. However, some OS use ints for inodes. I've added cast to long int
4877 to get rid of the compiler warning.
4879 19. I had forgotten to lock out "/../" in configuration file names when
4880 ALT_CONFIG_PREFIX was set.
4882 20. Routers used for verification do not need to specify transports. However,
4883 if such a router generated a host list, and callout was configured, Exim
4884 crashed, because it could not find a port number from the (non-existent)
4885 transport. It now assumes port 25 in this circumstance.
4887 21. Added the -t option to exigrep.
4889 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4890 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4891 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4892 them is set, LOOKUP_LSEARCH is forced.
4894 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4896 24. Added sender and host information to the "rejected by local_scan()" log
4897 line; previously there was no indication of these.
4899 25. Added .include_if_exists.
4901 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4902 Linux. It turns out that not all file systems support this. Apparently some
4903 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4904 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4905 not supported on the file descriptor, is now ignored when Exim is trying to
4906 sync a directory. This applies only to Linux.
4908 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4910 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4911 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4914 29. Added support for maildirsize files from supplied patch (modified a bit).
4916 30. The use of :fail: followed by an empty string could lead Exim to respond to
4917 sender verification failures with (e.g.):
4919 550 Verification failed for <xxx>
4920 550 Sender verify failed
4922 where the first response line was missing the '-' that indicates it is not
4923 the final line of the response.
4925 31. The loop for finding the name of the user that called Exim had a hardwired
4926 limit of 10; it now uses the value of finduser_retries, which is used for
4927 all other user lookups.
4929 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4932 33. Exim was neglecting to zero errno before one call of strtol() when
4933 expanding a string and expecting an integer value. On some systems this
4934 resulted in spurious "integer overflow" errors. Also, it was casting the
4935 result into an int without checking.
4937 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4938 did not work. The code looks as if it has *never* worked, though it appears
4939 to have been documented since at least releast 1.62. I have made it work.
4941 35. The "timeout_DNS" error in retry rules, also documented since at least
4942 1.62, also never worked. As it isn't clear exactly what this means, and
4943 clearly it isn't a major issue, I have abolished the feature by treating it
4944 as "timeout", and writing a warning to the main and panic logs.
4946 36. The display of retry rules for -brt wasn't always showing the error code
4949 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4950 timeout_connect_A, timeout_connect_MX.
4952 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4953 to the empty sender.
4955 39. The daemon was not analysing the content of -oX till after it had closed
4956 stderr and disconnected from the controlling terminal. This meant that any
4957 syntax errors were only noted on the panic log, and the return code from
4958 the command was 0. By re-arranging the code a little, I've made the
4959 decoding happen first, so such errors now appear on stderr, and the return
4960 code is 1. However, the actual setting up of the sockets still happens in
4961 the disconnected process, so errors there are still only recorded on the
4964 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4965 connections (as happens on some IP stacks) was logged at start up time as
4966 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4967 it from "IPv6 and IPv4", which means that two separate sockets are being
4970 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4971 says whether AF_INET or AF_INET6 was passed as an argument.
4973 42. Exiwhat output was messed up when time zones were included in log
4976 43. Exiwhat now gives more information about the daemon's listening ports,
4977 and whether -tls-on-connect was used.
4979 44. The "port" option of the smtp transport is now expanded.
4981 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4982 silently ignored. Now an error message is written to the main and panic
4985 46. There's a new ACL modifier called "logwrite" which writes to a log file
4986 as soon as it is encountered.
4988 47. Added $local_user_uid and $local_user_gid at routing time.
4990 48. Exim crashed when trying to verify a sender address that was being
4993 49. Exim was recognizing only a space character after ".include". It now also
4994 recognizes a tab character.
4996 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4997 extracting the relevant information from the specification. The man page no
4998 longer contains scrambled data for the -d option, and I've added a section
4999 at the front about calling Exim under different names.
5001 51. Added "extra_headers" argument to the "mail" command in filter files.
5003 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5006 53. Installed eximstats 1.29.
5008 54. Added transport_filter_timeout as a generic transport option.
5010 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5011 Cc: header lines. This was required by RFC 822, but it not required by RFC
5014 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5015 incoming messages that did not have them. Now it does so only if the
5016 message originates locally, that is, if there is no associated remote host
5017 address. When Resent- header lines are present, this applies to the Resent-
5018 lines rather than the non-Resent- lines.
5020 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5021 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5023 58. Messages for configuration errors now include the name of the main
5024 configuration file - useful now that there may be more than one file in a
5025 list (.included file names were always shown).
5027 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5028 for those rare installations that do not start the daemon as root or run it
5029 setuid root. I've cut out the call to initgroups() if the daemon is not
5032 60. The Exim user and group can now be bound into the binary as text strings
5033 that are looked up at the start of Exim's processing.
5035 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5037 62. Added $mailstore_basename variable.
5039 63. Installed patch to sieve.c from Michael Haardt.
5041 64. When Exim failed to open the panic log after failing to open the main log,
5042 the original message it was trying to log was written to stderr and debug
5043 output, but if they were not available (the usual case in production), it
5044 was lost. Now it is written to syslog before the two lines that record the
5045 failures to open the logs.
5047 65. Users' Exim filters run in subprocesses under the user's uid. It is
5048 possible for a "deliver" command or an alias in a "personal" command to
5049 provoke an address rewrite. If logging of address rewriting is configured,
5050 this fails because the process is not running as root or exim. There may be
5051 a better way of dealing with this, but for the moment (because 4.30 needs
5052 to be released), I have disabled address rewrite logging when running a
5053 filter in a non-root, non-exim process.
5059 1. The buildconfig auxiliary program wasn't quoting the value set for
5060 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5061 not defined. This bug was masked in 4.22 by the effect that was fixed in
5064 2. Some messages that were rejected after a message id was allocated were
5065 shown as "incomplete" by exigrep. It no longer does this for messages that
5066 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5068 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5069 have allow_domain_literals set, the ID did not get logged in the <= line.
5070 Domain literals are now always recognized in Message-ID: header lines.
5072 4. The first argument for a ${extract expansion item is the key name or field
5073 number. Leading and trailing spaces in this item were not being ignored,
5074 causing some misleading effects.
5076 5. When deliver_drop_privilege was set, single queue runner processes started
5077 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5078 same command in the process it spins off) were not dropping privilege.
5080 6. When the daemon running as "exim" started a queue runner, it always
5081 re-executed Exim in the spun-off process. This is a waste of effort when
5082 deliver_drop_privilege is set. The new process now just calls the
5083 queue-runner function directly.
5089 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5092 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5093 ignored. Though the use of -oP was forcing the writing of a pid file, it
5094 was always written to the default place.
5096 3. If the message "no IP address found for host xxxx" is generated during
5097 incoming verification, it is now followed by identification of the incoming
5098 connection (so you can more easily find what provoked it).
5100 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5102 5. Added some features to "harden" Exim a bit more against certain attacks:
5104 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5105 be put in Local/Makefile. This is like the never_users runtime option,
5106 but it cannot be overridden. The default setting is "root".
5108 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5109 prefix string with which any file named in a -C command line option
5112 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5113 is retained for -C and -D only if the caller of Exim is root. Without
5114 it, the exim user may also use -C and -D and retain privilege.
5116 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5117 command line option is disabled.
5119 6. Macro names set by the -D option must start with an upper case letter, just
5120 like macro names defined in the configuration file.
5122 7. Added "dereference=" facility to LDAP.
5124 8. Two instances of the typo "uknown" in the source files are fixed.
5126 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5127 the Configure-Makefile script screwed up while processing it.
5129 10. Incorporated PCRE 4.4.
5131 11. The SMTP synchronization check was not operating right at the start of an
5132 SMTP session. For example, it could not catch a HELO sent before the client
5133 waited for the greeting. There is now a check for outstanding input at the
5134 point when the greeting is written. Because of the duplex, asynchronous
5135 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5136 way, but not yet received, when the check is performed.
5138 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5139 on TCP/IP sockets, because this apparently causes some broken clients to
5142 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5143 unchanged) from the Cygwin maintainer.
5145 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5146 was supported instead of testing for mbx. Effectively a typo.
5148 15. The spa authenticator server code was not checking that the input it
5149 received was valid base64.
5151 16. The debug output line for the "set" modifier in ACLs was not showing the
5152 name of the variable that was being set.
5154 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5156 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5157 lookup on its own. Something else was needed to trigger the lookup. For
5158 example, a match in host_lookup or the need for a host name in a host list.
5159 Now, if $sender_host_name is referenced and the host name has not yet been
5160 looked up, a lookup is performed. If the lookup fails, the variable remains
5161 empty, and $host_lookup_failed is set to "1".
5163 19. Added "eqi" as a case-independent comparison operator.
5165 20. The saslauthd authentication condition could segfault if neither service
5166 nor realm was specified.
5168 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5169 error message that was logged was misleading, and incoming SMTP
5170 connections were dropped. The message is now more accurate, and temporary
5171 errors are given to SMTP connections.
5173 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5174 (except RSET) with a 421 error, until QUIT is received. However, it was
5175 failing to send a response to QUIT.
5177 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5178 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5179 ACL is now run after the helo_try_verify_hosts code.
5181 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5182 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5183 case-independent, so other case variants are also recognized). Apparently
5184 some systems use these upper case variants.
5186 25. If more than two messages were waiting for the same host, and a transport
5187 filter was specified for the transport, Exim sent two messages over the
5188 same TCP/IP connection, and then failed with "socket operation on non-
5189 socket" when it tried to send the third.
5191 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5193 27. The extern definition of crypt16() in expand.c was not being excluded when
5194 the OS had its own crypt16() function.
5196 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5197 as a preferred synonym for return_size_limit, both as an option and as an
5200 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5202 30. Changed the default configuration ACL to relax the local part checking rule
5203 for addresses that are not in any local domains. For these addresses,
5204 slashes and pipe symbols are allowed within local parts, but the sequence
5205 /../ is explicitly forbidden.
5207 31. SPA server authentication was not clearing the challenge buffer before
5210 32. log_message in a "warn" ACL statement was writing to the reject log as
5211 well as to the main log, which contradicts the documentation and doesn't
5212 seem right (because no rejection is happening). So I have stopped it.
5214 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5215 However, I am unable to do any testing of this.
5217 34. Fixed an infelicity in the appendfile transport. When checking directories
5218 for a mailbox, to see if any needed to be created, it was accidentally
5219 using path names with one or more superfluous leading slashes; tracing
5220 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5222 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5223 contents are added to the log line that is written for every discarded
5224 recipient. (Previously a log_message setting was ignored.)
5226 36. The ${quote: operator now quotes the string if it is empty.
5228 37. The install script runs exim in order to find its version number. If for
5229 some reason other than non-existence or emptiness, which it checks, it
5230 could not run './exim', it was installing it with an empty version number,
5231 i.e. as "exim-". This error state is now caught, and the installation is
5234 38. An argument was missing from the function that creates an error message
5235 when Exim fails to connect to the socket for saslauthd authentication.
5236 This could cause Exim to crash, or give a corrupted message.
5238 39. Added isip, isip4, and isip6 to ${if conditions.
5240 40. The ACL variables $acl_xx are now saved with the message, and can be
5241 accessed later in routers, transports, and filters.
5243 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5244 strings in the file are not string-expanded.
5246 42. If a MAIL command specified a SIZE value that was too large to fit into an
5247 int variable, the check against message_size_limit failed. Such values are
5248 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5249 day this will have to be increased, but I don't think I want to be around
5250 when emails are that large.
5257 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5258 iconv() is not standard in FreeBSD.
5260 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5261 IPv6 enabled. The observed symptom was a segmentation fault on return from
5262 the function os_common_find_running_interfaces() in src/os.c.
5264 3. In the check_special_case() function in daemon.c I had used "errno" as an
5265 argument name, which causes warnings on some systems. This was basically a
5266 typo, since it was named "eno" in the comments!
5268 4. The code that waits for the clock to tick (at a resolution of some fraction
5269 of a second) so as to ensure message-id uniqueness was always waiting for
5270 at least one whole tick, when it could have waited for less. [This is
5271 almost certainly not relevant at current processor speeds, where it is
5272 unlikely to ever wait at all. But we try to future-proof.]
5274 5. The function that sleeps for a time interval that includes fractions of a
5275 second contained a race. It did not block SIGALRM between setting the
5276 timer, and suspending (a couple of lines later). If the interval was short
5277 and the sigsuspend() was delayed until after it had expired, the suspension
5278 never ended. On busy systems this could lead to processes getting stuck for
5281 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5282 process, before it forks any delivery processes. The open lookup caching
5283 mechanism meant that the open file or database connection was passed into
5284 the delivery process. The problem was that delivery processes always tidy
5285 up cached lookup data. This could cause a problem for the next delivery
5286 process started by the queue runner, because the external queue runner
5287 process does not know about the closure. So the next delivery process
5288 still has data in the lookup cache. In the case of a file lookup, there was
5289 no problem because closing a file descriptor in a subprocess doesn't affect
5290 the parent. However, if the lookup was caching a connection to a database,
5291 the connection was closed, and the second delivery process was likely to
5292 see errors such as "PGSQL: query failed: server closed the connection
5293 unexpectedly". The problem has been fixed by closing all cached lookups
5294 in a queue runner before running a delivery process.
5296 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5297 seem to have the "const" qualifier which it has on other OS. I've
5300 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5301 *of the same type* (client or server) with the same public name that an
5302 error should be diagnosed.
5304 9. When Exim looked up a host name for an IP address, but failed to find the
5305 original IP address when looking up the host name (a safety check), it
5306 output the message "<ip address> does not match any IP for NULL", which was
5307 confusing, to say the least. The bug was that the host name should have
5308 appeared instead of "NULL".
5310 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5311 user that is built into the binary, and the -C or -D options is used, root
5312 privilege is dropped before the configuration file is read. In addition,
5313 logging is switched to stderr instead of the normal log files. If the
5314 configuration then re-defines the Exim user, the unprivileged environment
5315 is probably not what is expected, so Exim logs a panic warning message (but
5318 However, if deliver_drop_privilege is set, the unprivileged state may well
5319 be exactly what is intended, so the warning has been cut out in that case,
5320 and Exim is allowed to try to write to its normal log files.
5326 1. smtp_return_error_details was not giving details for temporary sender
5327 or receiver verification errors.
5329 2. Diagnose a configuration error if two authenticators have the same public
5332 3. Exim used not to create the message log file for a message until the first
5333 delivery attempt. This could be confusing when incoming messages were held
5334 for policy or load reasons. The message log file is now created at the time
5335 the message is received, and an initial "Received" line is written to it.
5337 4. The automatically generated man page for command line options had a minor
5338 bug that caused no ill effects; however, a more serious problem was that
5339 the procedure for building the man page automatically didn't always
5340 operate. Consequently, release 4.20 contains an out-of-date version. This
5341 shouldn't happen again.
5343 5. When building Exim with embedded Perl support, the script that builds the
5344 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5345 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5347 6. The freeze_tell option was not being used for messages that were frozen on
5348 arrival, either by an ACL or by local_scan().
5350 7. Added the smtp_incomplete_transaction log selector.
5352 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5353 was accepting AUTH without a new EHLO.
5355 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5356 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5359 10. Logging of TCP/IP connections (when configured) now happens in the main
5360 daemon process instead of the child process, so that the TCP/IP connection
5361 count is more accurate (but it can never be perfect).
5363 11. The use of "drop" in a nested ACL was not being handled correctly in the
5364 outer ACL. Now, if condition failure induced by the nested "drop" causes
5365 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5366 or "require"), the connection is dropped.
5368 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5369 that yield "discard" can now be used with an "accept" or a "discard" verb,
5370 but an error is generated for any others (because I can't see a useful way
5371 to define what should happen).
5373 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5374 are now processed in the same way as lines in the Exim configuration file.
5375 In particular, continuation lines are supported.
5377 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5379 15. Added -ti meaning -t -i.
5381 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5382 domains, and warn by logging if others are found.
5384 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5385 returned from GIFCONF when Exim is trying to find the list of interfaces on
5386 a host. The code in os.c has been modified to copy each ifreq to an aligned
5387 structure in all cases.
5389 Also, in some cases, the returned ifreq's were being copied to a 'struct
5390 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5391 means the last couple of bytes of an IPv6 address could be chopped if the
5392 ifreq contained only a normal sockaddr (14 bytes storage).
5394 18. Named domain lists were not supported in the hosts_treat_as_local option.
5395 An entry such as +xxxx was not recognized, and was treated as a literal
5398 19. Ensure that header lines added by a DATA ACL are included in the reject log
5399 if the ACL subsequently rejects the message.
5401 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5402 MD5 (which is deprecated).
5404 21. When testing a filter file using -bf, Exim was writing a message when it
5405 took the sender from a "From " line in the message, but it was not doing so
5406 when it took $return_path from a Return-Path: header line. It now does.
5408 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5409 with a valid header line field name (a series of printing characters
5410 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5412 23. Changed "disc" in the source to "disk" to conform to the documentation and
5413 the book and for uniformity.
5415 24. Ignore Sendmail's -Ooption=value command line item.
5417 25. When execve() failed while trying to run a command in a pipe transport,
5418 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5419 could be confused with a return value of 69 from the command itself. This
5420 has been changed to 127, the value the shell returns if it is asked to run
5421 a non-existent command. The wording for the related log line suggests a
5422 non-existent command as the problem.
5424 26. If received_header_text expands to an empty string, do not add a Received:
5425 header line to the message. (Well, it adds a token one on the spool, but
5426 marks it "old" so that it doesn't get used or transmitted.)
5428 27. Installed eximstats 1.28 (addition of -nt option).
5430 28. There was no check for failure on the call to getsockname() in the daemon
5431 code. This can fail if there is a shortage of resources on the system, with
5432 ENOMEM, for example. A temporary error is now given on failure.
5434 29. Contrary to the C standard, it seems that in some environments, the
5435 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5436 program. Exim now does this explicitly; it affects the formatting of
5437 timestamps using strftime().
5439 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5440 complaints. I've now initialized all the variables, to avoid this.
5442 32. Header lines added by a system filter were not being "seen" during
5443 transport-time rewrites.
5445 33. The info_callback() function passed to OpenSSL is set up with type void
5446 (*)(SSL *, int, int), as described somewhere. However, when calling the
5447 function (actually a macro) that sets it up, the type void(*)() is
5448 expected. I've put in a cast to prevent warnings from picky compilers.
5450 34. If a DNS black list lookup found a CNAME record, but there were no A
5451 records associated with the domain it pointed at, Exim crashed.
5453 35. If a DNS black list lookup returned more than one A record, Exim ignored
5454 all but the first. It now scans all returned addresses if a particular IP
5455 value is being sought. In this situation, the contents of the
5456 $dnslist_value variable are a list of all the addresses, separated by a
5459 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5460 to accept a host name and all its aliases if the forward lookup for any of
5461 them yielded the IP address of the incoming connection. Now it accepts only
5462 those names whose forward lookup yields the correct IP address. Any other
5463 names are discarded. This closes a loophole whereby a rogue DNS
5464 administrator could create reverse DNS records to break through a
5465 wildcarded host restriction in an ACL.
5467 37. If a user filter or a system filter that ran in a subprocess used any of
5468 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5469 the wrong values were passed to the pipe command ($thisaddress had the
5470 value of $0, $0 had the value of $1, etc). This bug was introduced by
5471 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5473 38. Improved the line breaking for long SMTP error messages from ACLs.
5474 Previously, if there was no break point between 40 and 75 characters, Exim
5475 left the rest of the message alone. Two changes have been made: (a) I've
5476 reduced the minimum length to 35 characters; (b) if it can't find a break
5477 point between 35 and 75 characters, it looks ahead and uses the first one
5478 that it finds. This may give the occasional overlong line, but at least the
5479 remaining text gets split now.
5481 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5482 file descriptors might be low, and that setting 1000 would always raise it.
5483 It turns out that in some environments, the limit is already over 1000 and
5484 that lowering it causes trouble. So now Exim takes care not to decrease it.
5486 40. When delivering a message, the value of $return_path is set to $sender_
5487 address at the start of routing (routers may change the value). By an
5488 oversight, this default was not being set up when an address was tested by
5489 -bt or -bv, which affected the outcome if any router or filter referred to
5492 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5493 log only when "message" or "log_message" are set. However, if one of the
5494 conditions was an address verification, or a call to a nested ACL, the
5495 messages generated by the underlying test were being passed through. This
5496 no longer happens. The underlying message is available in $acl_verify_
5497 message for both "message" and "log_message" expansions, so it can be
5498 passed through if needed.
5500 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5501 new expansion $bh_ to give the encoded byte string without charset
5502 translation. Translation happens only if iconv() is available; HAVE_ICONV
5503 indicates this at build time. HEADERS_CHARSET gives the charset to
5504 translate to; headers_charset can change it in the configuration, and
5505 "headers charset" can change it in an individual filter file.
5507 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5508 that creates RFC 2047 encoded "words" labels them as that charset instead
5509 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5510 expansion operator; (ii) when Exim creates a From: line for a local
5511 message; (iii) when a header line is rewritten to include a "phrase" part.
5513 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5514 buggy, causing it to skip the first lines of messages whose message ID
5515 ended in 'D'. This would not have bitten before Exim release 4.14, because
5516 message IDs were unlikely to end in 'D' before then. The effect was to have
5517 incorrect size information for certain domains.
5519 45. #include "config.h" was missing at the start of the crypt16.c module. This
5520 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5523 46. If there was a timeout during a "random" callout check, Exim treated it as
5524 a failure of the random address, and carried on sending RSET and the real
5525 address. If the delay was just some slowness somewhere, the response to the
5526 original RCPT would be taken as a response to RSET and so on, causing
5527 mayhem of various kinds.
5529 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5530 when I implemented it. It didn't allow for the fact that some option values
5531 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5532 the right test for positive values.
5534 48. Domain names in DNS records are case-independent. Exim always looks them up
5535 in lower case. Some resolvers return domain names in exactly the case they
5536 appear in the zone file, that is, they may contain uppercase letters. Not
5537 all resolvers do this - some return always lower case. Exim was treating a
5538 change of case by a resolver as a change of domain, similar to a widening
5539 of a domain abbreviation. This triggered its re-routing code and so it was
5540 trying to route what was effectively the same domain again. This normally
5541 caused routing to fail (because the router wouldn't handle the domain
5542 twice). Now Exim checks for this case specially, and just changes the
5543 casing of the domain that it ultimately uses when it transmits the message
5546 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5549 50. If a filter generated a file delivery with a non-absolute name (possible if
5550 no home directory exists for the router), the forbid_file option was not
5553 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5554 the existing equality matching.
5556 52. Exim was using ints instead of ino_t variables in some places where it was
5557 dealing with inode numbers.
5559 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5560 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5561 called TMPDIR, and if it finds it is different, it changes its value.
5563 54. The smtp_printf() function is now made available to local_scan() so
5564 additional output lines can be written before returning. There is also an
5565 smtp_fflush() function to enable the detection of a dropped connection.
5566 The variables smtp_input and smtp_batched_input are exported to
5569 55. Changed the default runtime configuration: the message "Unknown user"
5570 has been removed from the ACL, and instead placed on the localuser router,
5571 using the cannot_route_message feature. This means that any verification
5572 failures that generate their own messages won't get overridden. Similarly,
5573 the "Unrouteable address" message that was in the ACL for unverifiable
5574 relay addresses has also been removed.
5576 56. Added hosts_avoid_esmtp to the smtp transport.
5578 57. The exicyclog script was not checking for the esoteric option
5579 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5580 will work only if exicyclog is run under the appropriate euid.
5582 58. Following a discussion on the list, the rules by which Exim recognises line
5583 endings on incoming messages have been changed. The -dropcr and drop_cr
5584 options are now no-ops, retained only for backwards compatibility. The
5585 following line terminators are recognized: LF CRLF CR. However, special
5586 processing applies to CR:
5588 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5589 nor a local message in the state where . is a terminator.
5591 (ii) If a bare CR is encountered in a header line, an extra space is added
5592 after the line terminator so as not to end the header. The reasoning
5593 behind this is that bare CRs in header lines are most likely either
5594 to be mistakes, or people trying to play silly games.
5596 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5597 was being incorrectly given as 18 bytes larger than it should have been.
5598 This is a VOB (very old bug).
5600 60. This may never have affected anything current, but just in case it has:
5601 When the local host is found other than at the start of a list of hosts,
5602 the local host, those with the same MX, and any that follow, are discarded.
5603 When the list in question was part of a longer list of hosts, the following
5604 hosts (not currently being processed) were also being discarded. This no
5605 longer happens. I'm not sure if this situation could ever has previously
5608 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5611 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5612 header. This is something that is recommended in a new Internet Draft, and
5613 is something that is documented as being done by Sendmail. There are two
5614 possible values. For messages generated by the autoreply transport, Exim
5617 Auto-Submitted: auto-replied
5619 whereas for all other generated messages (e.g. bounces) it adds
5621 Auto-Submitted: auto-generated
5623 63. The "personal" condition in filters now includes a test for the
5624 Auto-Submitted: header. If it contains the string "auto-" the message it
5625 not considered personal.
5627 64. Added rcpt_include_affixes as a generic transport option.
5629 65. Added queue_only_override (default true).
5631 66. Added the syslog_duplication option.
5633 67. If what should have been the first header line of a message consisted of
5634 a space followed by a colon, Exim was mis-interpreting it as a header line.
5635 It isn't of course - it is syntactically invalid and should therefore be
5636 treated as the start of the message body. The misbehaviour could have
5637 caused a number of strange effects, including loss of data in subsequent
5638 header lines, and spool format errors.
5640 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5641 client host had authenticated. This control can now be exercised by an ACL
5642 for more flexibility.
5644 69. By default, callouts do not happen when testing with -bh. There is now a
5645 variant, -bhc, which does actually run the callout code, including
5646 consulting and updating the callout cache.
5648 70. Added support for saslauthd authentication, courtesy of Alexander
5651 71. If statvfs() failed on the spool or log directories while checking their
5652 size for availability, Exim confusingly gave the error "space shortage".
5653 Furthermore, in debugging mode it crashed with a floating point exception.
5654 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5655 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5656 serious problem, Exim now writes to the main and panic logs when this
5657 happens, with details of the failure. It then refuses to accept the
5658 incoming message, giving the message "spool directory problem" or "log
5659 directory problem" with a 421 code for SMTP messages.
5661 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5662 0, 1, and 2 exist, because some OS complain for execs without them (see
5663 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5664 descriptors. However, the code omitted to check that the open succeeded,
5665 causing mysterious errors if for some reason the permissions on /dev/null
5666 got screwed. Now Exim writes a message to the main and panic logs, and
5667 bombs out if it can't open /dev/null.
5669 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5670 interact so that it is all more flexible. It is supposed to remain
5671 backwards compatible. Also added extra_local_interfaces.
5673 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5674 to bomb out with an assertion failure - to the client this appears as a
5675 connection drop. This problem occurs in the part of the code that was taken
5676 from the Samba project. Fortunately, the assertion is in a very simple
5677 function, so I have fixed this by reproducing the function inline in the
5678 one place where it is called, and arranging for authentication to fail
5679 instead of killing the process with assert().
5681 75. The SPA client code was not working when the server requested OEM rather
5682 than Unicode encoding.
5684 76. Added code to make require_files with a specific uid setting more usable in
5685 the case where statting the file as root fails - usually a non-root-mounted
5686 NFS file system. When this happens and the failure is EACCES, Exim now
5687 forks a subprocess and does the per-uid checking as the relevant uid.
5689 77. Added process_log_path.
5691 78. If log_file_path was not explicitly set, a setting of check_log_space or
5692 check_log_inodes was ignored.
5694 79. If a space check for the spool or log partitions fails, the incident is now
5695 logged. Of course, in the latter case the data may get lost...
5697 80. Added the %p formatting code to string_format() so that it can be used to
5698 print addresses in debug_print(). Adjusted all the address printing in the
5699 debugging in store.c to use %p rather than %d.
5701 81. There was a concern that a line of code in smtp_in.c could overflow a
5702 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5703 initially expressed, the concern was not well-founded, because trailing
5704 spaces are removed early. However, if the trailing spaces were followed by
5705 a NULL, they did not get removed, so the overflow was possible. Two fixes
5708 (a) I re-wrote the offending code in a cleaner fashion.
5709 (b) If an incoming SMTP command contains a NULL character, it is rejected
5712 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5713 runs initgroups(), so that if the Exim user is in any additional groups,
5714 they will be used during message reception.
5720 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob