1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
9 PP/02 Make -n do something, by making it not do something.
10 When combined with -bP, the name of an option is not output.
12 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
15 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
16 $sender_host_name and config options to manage this, and basic check
19 PP/05 DSCP support for outbound connections and control modifier for inbound.
21 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
22 (Only plugin which currently uses this is kerberos4, which nobody should
23 be using, but we should make it available and other future plugins might
24 conceivably use it, even though it would break NAT; stuff *should* be
25 using channel bindings instead).
27 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
28 name; added for Sendmail compatibility; requires admin caller.
29 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
30 just ignore it); requires trusted caller.
31 Also parse but ignore: -Ac -Am -X<logfile>
34 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
36 TL/02 Add +smtp_confirmation as a default logging option.
38 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
39 Patch by Magnus Holmgren from 2007-02-20.
41 TL/04 Bugzilla 1281 - Spec typo.
42 Bugzilla 1283 - Spec typo.
43 Bugzilla 1290 - Spec grammar fixes.
45 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
47 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
49 JH/02 Support "G" suffix to numbers in ${if comparisons.
51 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
53 NM/01 Bugzilla 1197 - Spec typo
54 Bugzilla 1196 - Spec examples corrections
56 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
58 PP/09 Add gnutls_enable_pkcs11 option.
60 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
61 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
63 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
64 "acl {{name}{arg}...}", and optional args on acl condition
67 JH/05 Permit multiple router/transport headers_add/remove lines.
69 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
71 JH/07 Avoid using a waiting database for a single-message-only transport.
72 Performance patch from Paul Fisher. Bugzilla 1262.
74 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
77 JH/09 Add $headers_added variable, with content from use of ACL modifier
78 add_header (but not yet added to the message). Bugzilla 199.
80 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
81 Pulled from Bugzilla 817 by Wolfgang Breyha.
83 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
86 JH/11 Add A= logging on delivery lines, and a client_set_id option on
89 JH/12 Add optional authenticated_sender logging to A= and a log_selector
92 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
94 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
95 advertise SMTP AUTH mechanism to us, instead of a generic
96 protocol violation error. Also, make Exim more robust to bad
97 data from the Dovecot auth socket.
103 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
105 This, or similar/improved, will also be change PP/11 of 4.82.
111 PP/01 Handle short writes when writing local log-files.
112 In practice, only affects FreeBSD (8 onwards).
113 Bugzilla 1053, with thanks to Dmitry Isaikin.
115 NM/01 Bugzilla 949 - Documentation tweak
117 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
120 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
122 PP/02 Implemented gsasl authenticator.
124 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
126 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
127 `pkg-config foo` for cflags/libs.
129 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
130 with rest of GSASL and with heimdal_gssapi.
132 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
133 `pkg-config foo` for cflags/libs for the TLS implementation.
135 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
136 properties get this fed in as external SSF. A number of robustness
137 and debugging improvements to the cyrus_sasl authenticator.
139 PP/08 cyrus_sasl server now expands the server_realm option.
141 PP/09 Bugzilla 1214 - Log authentication information in reject log.
142 Patch by Jeremy Harris.
144 PP/10 Added dbmjz lookup type.
146 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
148 PP/12 MAIL args handles TAB as well as SP, for better interop with
149 non-compliant senders.
150 Analysis and variant patch by Todd Lyons.
152 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
153 Bug report from Lars Müller <lars@samba.org> (via SUSE),
154 Patch from Dirk Mueller <dmueller@suse.com>
156 PP/13 tls_peerdn now print-escaped for spool files.
157 Observed some $tls_peerdn in wild which contained \n, which resulted
158 in spool file corruption.
160 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
161 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
162 or write after TLS renegotiation, which otherwise led to messages
165 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
166 as a tracking header (ie: a signed header comes before the signature).
167 Patch from Wolfgang Breyha.
169 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
170 comma-sep list; embedded commas doubled.
172 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
174 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
176 Report and patch from Dmitry Banschikov.
178 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
179 Removed SSL_clear() after SSL_new() which led to protocol negotiation
180 failures. We appear to now support TLS1.1+ with Exim.
182 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
183 lets Exim select keys and certificates based upon TLS SNI from client.
184 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
185 before an outbound SMTP session. New log_selector, +tls_sni.
187 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
188 NULL dereference. Report and patch from Alun Jones.
190 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
191 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
192 Not seeing resolver debug output on NetBSD, but suspect this is a
193 resolver implementation change.
195 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
196 Left warnings. Added "eximon gdb" invocation mode.
198 PP/21 Defaulting "accept_8bitmime" to true, not false.
200 PP/22 Added -bw for inetd wait mode support.
202 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
203 locate the relevant includes and libraries. Made this the default.
205 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
206 Bugzilla 1246, report and most of solution from Tomasz Kusy.
208 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
209 This may cause build issues on older platforms.
211 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
212 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
213 gnutls_require_mac & gnutls_require_protocols (no longer supported).
214 Added SNI support via GnuTLS too.
215 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
217 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
219 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
220 Applied second patch from Janne, implementing suggestion to default
221 multiple-strings-in-record handling to match SPF spec.
223 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
225 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
226 read-only, out of scope).
227 Patch from Wolfgang Breyha, report from Stuart Northfield.
229 PP/29 Fix three issues highlighted by clang analyser static analysis.
230 Only crash-plausible issue would require the Cambridge-specific
231 iplookup router and a misconfiguration.
232 Report from Marcin Mirosław.
234 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
236 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
237 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
238 As part of this, removing so much warning spew let me fix some minor
239 real issues in debug logging.
241 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
242 assignment on my part. Fixed.
244 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
245 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
246 Janne Snabb (who went above and beyond: thank you).
248 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
249 string otherwise requires a connection and a bunch more work and it's
250 relatively easy to get wrong. Should also expose TLS library linkage
253 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
254 64-bit ${eval} (JH/03).
256 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
257 GNU libc to support some of the 64-bit stuff, should not lead to
258 conflicts. Defined before os.h is pulled in, so if a given platform
259 needs to override this, it can.
261 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
262 protection layer was required, which is not implemented.
263 Bugzilla 1254, patch from Wolfgang Breyha.
265 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
266 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
267 tls_dhparam take prime identifiers. Also unbreak combination of
268 OpenSSL+DH_params+TLSSNI.
270 PP/39 Disable SSLv2 by default in OpenSSL support.
276 PP/01 Solaris build fix for Oracle's LDAP libraries.
277 Bugzilla 1109, patch from Stephen Usher.
279 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
281 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
284 TF/02 Fix a couple more cases where we did not log the error message
285 when unlink() failed. See also change 4.74-TF/03.
287 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
288 lock up or crash if it happened to be inside a call to libc when it
289 got a SIGUSR1 from exiwhat.
291 The SIGUSR1 handler appends the current process status to the process
292 log which is later printed by exiwhat. It used to use the general
293 purpose logging code to do this, but several functions it calls are
294 not safe for signals.
296 The new output code in the SIGUSR1 handler is specific to the process
297 log, and simple enough that it's easy to inspect for signal safety.
298 Removing some special cases also simplifies the general logging code.
299 Removing the spurious timestamps from the process log simplifies
302 TF/04 Improved ratelimit ACL condition.
304 The /noupdate option has been deprecated in favour of /readonly which
305 has clearer semantics. The /leaky, /strict, and /readonly update modes
306 are mutually exclusive. The update mode is no longer included in the
307 database key; it just determines when the database is updated. (This
308 means that when you upgrde Exim will forget old rate measurements.)
310 Exim now checks that the per_* options are used with an update mode that
311 makes sense for the current ACL. For example, when Exim is processing a
312 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
313 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
314 must specify per_mail/readonly. If you omit the update mode it defaults to
315 /leaky where that makes sense (as before) or /readonly where required.
317 The /noupdate option is now undocumented but still supported for
318 backwards compatibility. It is equivalent to /readonly except that in
319 ACLs where /readonly is required you may specify /leaky/noupdate or
320 /strict/noupdate which are treated the same as /readonly.
322 A useful new feature is the /count= option. This is a generalization
323 of the per_byte option, so that you can measure the throughput of other
324 aggregate values. For example, the per_byte option is now equivalent
325 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
327 The per_rcpt option has been generalized using the /count= mechanism
328 (though it's more complicated than the per_byte equivalence). When it is
329 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
330 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
331 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
332 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
333 Note that using per_rcpt with a non-readonly update mode in more than
334 one ACL will cause the recipients to be double-counted. (The per_mail
335 and per_byte options don't have this problem.)
337 The handling of very low rates has changed slightly. If the computed rate
338 is less than the event's count (usually one) then this event is the first
339 after a long gap. In this case the rate is set to the same as this event's
340 count, so that the first message of a spam run is counted properly.
342 The major new feature is a mechanism for counting the rate of unique
343 events. The new per_addr option counts the number of different
344 recipients that someone has sent messages to in the last time period. It
345 behaves like per_rcpt if all the recipient addresses are different, but
346 duplicate recipient addresses do not increase the measured rate. Like
347 the /count= option this is a general mechanism, so the per_addr option
348 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
349 example, measure the rate that a client uses different sender addresses
350 with the options per_mail/unique=$sender_address. There are further
351 details in the main documentation.
353 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
355 TF/06 Removed a few PCRE remnants.
357 TF/07 Automatically extract Exim's version number from tags in the git
358 repository when doing development or release builds.
360 PP/02 Raise smtp_cmd_buffer_size to 16kB.
361 Bugzilla 879. Patch from Paul Fisher.
363 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
364 Heavily based on revision 40f9a89a from Simon Arlott's tree.
367 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
369 PP/05 Variable $av_failed, true if the AV scanner deferred.
370 Bugzilla 1078. Patch from John Horne.
372 PP/06 Stop make process more reliably on build failure.
373 Bugzilla 1087. Patch from Heiko Schlittermann.
375 PP/07 Make maildir_use_size_file an _expandable_ boolean.
376 Bugzilla 1089. Patch from Heiko Schlittermann.
378 PP/08 Handle ${run} returning more data than OS pipe buffer size.
379 Bugzilla 1131. Patch from Holger Weiß.
381 PP/09 Handle IPv6 addresses with SPF.
382 Bugzilla 860. Patch from Wolfgang Breyha.
384 PP/10 GnuTLS: support TLS 1.2 & 1.1.
386 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
389 PP/11 match_* no longer expand right-hand-side by default.
390 New compile-time build option, EXPAND_LISTMATCH_RHS.
391 New expansion conditions, "inlist", "inlisti".
393 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
395 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
397 PP/14 fix log_write() format string regression from TF/03.
398 Bugzilla 1152. Patch from Dmitry Isaikin.
404 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
406 PP/02 Harmonised TLS library version reporting; only show if debugging.
407 Layout now matches that introduced for other libraries in 4.74 PP/03.
409 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
411 PP/04 New "dns_use_edns0" global option.
413 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
416 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
417 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
419 TK/01 Updated PolarSSL code to 0.14.2.
420 Bugzilla 1097. Patch from Andreas Metzler.
422 PP/07 Catch divide-by-zero in ${eval:...}.
425 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
428 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
429 format-string attack -- SECURITY: remote arbitrary code execution.
431 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
432 time unintentionally subject to list matching rules, letting the header
433 cause arbitrary Exim lookups (of items which can occur in lists, *not*
434 arbitrary string expansion). This allowed for information disclosure.
436 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
437 INT_MIN/-1 -- value coerced to INT_MAX.
443 NM/01 Workround for PCRE version dependency in version reporting
446 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
447 This fixes portability to compilers other than gcc, notably
448 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
450 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
451 makefiles for portability to HP-UX and POSIX correctness.
453 PP/01 Permit LOOKUP_foo enabling on the make command-line.
454 Also via indented variable definition in the Makefile.
455 (Debugging by Oliver Heesakkers).
457 PP/02 Restore caching of spamd results with expanded spamd_address.
458 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
460 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
461 Improves build reliability. Fix from: Frank Elsner
463 NM/02 Fix wide character breakage in the rfc2047 coding
464 Fixes bug 1064. Patch from Andrey N. Oktyabrski
466 NM/03 Allow underscore in dnslist lookups
467 Fixes bug 1026. Patch from Graeme Fowler
469 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
470 Code patches from Adam Ciarcinski of NetBSD.
472 NM/04 Fixed exiqgrep to cope with mailq missing size issue
475 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
476 is logged, to avoid truncation. Patch from John Horne.
478 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
479 Patch from Jakob Hirsch.
481 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
482 SQL string expansion failure details.
483 Patch from Andrey Oktyabrski.
485 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
486 Patch from Simon Arlott.
488 PP/09 New lookups functionality failed to compile on old gcc which rejects
489 extern declarations in function scope.
490 Patch from Oliver Fleischmann
492 PP/10 Use sig_atomic_t for flags set from signal handlers.
493 Check getgroups() return and improve debugging.
494 Fixed developed for diagnosis in bug 927 (which turned out to be
497 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
498 Patch from Mark Zealey.
500 PP/12 Bugzilla 1056: Improved spamd server selection.
501 Patch from Mark Zealey.
503 PP/13 Bugzilla 1086: Deal with maildir quota file races.
504 Based on patch from Heiko Schlittermann.
506 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
507 Patch from Uwe Doering, sign-off by Michael Haardt.
509 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
510 variable declaration deep within a block. Bug and patch from
513 PP/15 lookups-Makefile IRIX compatibilty coercion.
515 PP/16 Make DISABLE_DKIM build knob functional.
517 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
518 Patch by Simon Arlott
520 TF/03 Fix valgrind.h portability to C89 compilers that do not support
521 variable argument macros. Our copy now differs from upstream.
527 TF/01 Failure to get a lock on a hints database can have serious
528 consequences so log it to the panic log.
530 TF/02 Log LMTP confirmation messages in the same way as SMTP,
531 controlled using the smtp_confirmation log selector.
533 TF/03 Include the error message when we fail to unlink a spool file.
535 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
536 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
537 for maintaining out-of-tree patches for some time.
539 PP/01 Bugzilla 139: Documentation and portability issues.
540 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
541 Handle per-OS dynamic-module compilation flags.
543 PP/02 Let /dev/null have normal permissions.
544 The 4.73 fixes were a little too stringent and complained about the
545 permissions on /dev/null. Exempt it from some checks.
546 Reported by Andreas M. Kirchwitz.
548 PP/03 Report version information for many libraries, including
549 Exim version information for dynamically loaded libraries. Created
550 version.h, now support a version extension string for distributors
551 who patch heavily. Dynamic module ABI change.
553 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
554 privilege escalation vulnerability whereby the Exim run-time user
555 can cause root to append content of the attacker's choosing to
558 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
561 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
562 If dropping privileges for untrusted macros, we disabled normal logging
563 on the basis that it would fail; for the Exim run-time user, this is not
564 the case, and it resulted in successful deliveries going unlogged.
565 Fixed. Reported by Andreas Metzler.
571 PP/01 Date: & Message-Id: revert to normally being appended to a message,
572 only prepend for the Resent-* case. Fixes regression introduced in
573 Exim 4.70 by NM/22 for Bugzilla 607.
575 PP/02 Include check_rfc2047_length in configure.default because we're seeing
576 increasing numbers of administrators be bitten by this.
578 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
580 PP/03 Bugzilla 994: added openssl_options main configuration option.
582 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
584 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
586 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
588 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
589 without a peer certificate, leading to a segfault because of an
590 assumption that peers always have certificates. Be a little more
591 paranoid. Problem reported by Martin Tscholak.
593 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
594 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
595 NB: ClamAV planning to remove STREAM in "middle of 2010".
596 CL also introduces -bmalware, various -d+acl logging additions and
597 more caution in buffer sizes.
599 PP/09 Implemented reverse_ip expansion operator.
601 PP/10 Bugzilla 937: provide a "debug" ACL control.
603 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
605 PP/12 Bugzilla 973: Implement --version.
607 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
609 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
611 PP/15 Bugzilla 816: support multiple condition rules on Routers.
613 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
614 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
615 ignore trailing whitespace.
617 JJ/02 prevent non-panic DKIM error from being sent to paniclog
619 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
622 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
623 Notification from Dr Andrew Aitchison.
625 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
626 ExtendedDetectionInfo response format.
627 Notification from John Horne.
629 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
632 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
633 XSL and documented dependency on system catalogs, with examples of how
636 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
639 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
640 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
641 configuration file which is writeable by the Exim user or group.
643 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
644 of configuration files to cover files specified with the -C option if
645 they are going to be used with root privileges, not just the default
648 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
649 option (effectively making it always true).
651 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
652 files to be used while preserving root privileges.
654 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
655 that rogue child processes cannot use them.
657 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
658 run-time user, instead of root.
660 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
661 Exim run-time user without dropping privileges.
663 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
664 result string, instead of calling string_vformat() twice with the same
667 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
668 for other users. Others should always drop root privileges if they use
669 -C on the command line, even for a whitelisted configure file.
671 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
673 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
679 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
680 $data_path, and $header_path variables; fixed documentation bugs and
683 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
684 exipick to access non-standard spools, including the "frozen" queue
687 NM/01 Bugzilla 965: Support mysql stored procedures.
688 Patch from Alain Williams
690 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
692 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
693 Patch from Andreas Metzler
695 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
696 Patch from Kirill Miazine
698 NM/05 Bugzilla 671: Added umask to procmail example.
700 JJ/03 installed exipick 20100323.0, fixing doc bug
702 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
703 directory. Notification and patch from Dan Rosenberg.
705 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
707 TK/02 Improve log output when DKIM signing operation fails.
709 MH/01 Treat the transport option dkim_domain as a colon separated
710 list, not as a single string, and sign the message with each element,
711 omitting multiple occurences of the same signer.
713 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
714 Bugzilla 985, 986. Patch by Simon Arlott
716 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
717 Patch by Simon Arlott
719 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
720 MBX locking. Notification from Dan Rosenberg.
726 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
728 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
730 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
732 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
734 NM/04 Bugzilla 915: Flag broken perl installation during build.
740 TK/01 Added patch by Johannes Berg that expands the main option
741 "spamd_address" if it starts with a dollar sign.
743 TK/02 Write list of recipients to X-Envelope-Sender header when building
744 the mbox-format spool file for content scanning (suggested by Jakob
747 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
748 (http://www.dcc-servers.net/) support via dccifd. Activated by
749 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
751 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
752 by Mark Daniel Reidel <mr@df.eu>.
754 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
755 When building exim an external PCRE library is now needed -
756 PCRE is a system library on the majority of modern systems.
757 See entry on PCRE_LIBS in EDITME file.
759 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
760 conversation. Added nologin parameter to request.
761 Patch contributed by Kirill Miazine.
763 TF/01 Do not log submission mode rewrites if they do not change the address.
765 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
767 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
768 log files in place. Contributed by Roberto Lima.
770 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
772 TF/03 Bugzilla 615: When checking the local_parts router precondition
773 after a local_part_suffix or local_part_prefix option, Exim now
774 does not use the address's named list lookup cache, since this
775 contains cached lookups for the whole local part.
777 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
778 Robert Millan. Documentation is in experimental-spec.txt.
780 TF/04 Bugzilla 668: Fix parallel build (make -j).
782 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
784 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
785 Patch provided by Jan Srzednicki.
787 TF/05 Leading white space used to be stripped from $spam_report which
788 wrecked the formatting. Now it is preserved.
790 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
791 that they are available at delivery time.
793 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
795 TF/08 TLS error reporting now respects the incoming_interface and
796 incoming_port log selectors.
798 TF/09 Produce a more useful error message if an SMTP transport's hosts
799 setting expands to an empty string.
801 NM/06 Bugzilla 744: EXPN did not work under TLS.
802 Patch provided by Phil Pennock.
804 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
805 Patch provided by Richard Godbee.
807 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
808 acl_smtp_notquit, added index entry.
810 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
811 Patch provided by Eugene Bujak.
813 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
814 accept(). Patch provided by Maxim Dounin.
816 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
817 Patch provided by Phil Pennock.
819 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
821 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
822 Patch provided by Brad "anomie" Jorsch.
824 NM/14 Bugzilla 622: Added timeout setting to transport filter.
825 Patch provided by Dean Brooks.
827 TK/05 Add native DKIM support (does not depend on external libraries).
829 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
830 Patch provided by Graeme Fowler.
832 NM/16 Bugzilla 851: Documentation example syntax fix.
834 NM/17 Changed NOTICE file to remove references to embedded PCRE.
836 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
839 NM/19 Bugzilla 745: TLS version reporting.
840 Patch provided by Phil Pennock.
842 NM/20 Bugzilla 167: bool: condition support.
843 Patch provided by Phil Pennock.
845 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
846 clients. Patch provided by Phil Pennock.
848 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
849 Patch provided by Brad "anomie" Jorsch.
851 NM/23 Bugzilla 687: Fix misparses in eximstats.
852 Patch provided by Heiko Schlittermann.
854 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
855 Patch provided by Heiko Schlittermann.
857 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
858 plus update to original patch.
860 NM/26 Bugzilla 799: Documentation correction for ratelimit.
862 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
863 Patch provided by David Brownlee.
865 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
867 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
869 NM/30 Bugzilla 888: TLS documentation bugfixes.
871 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
873 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
874 Unlike the original bugzilla I have changed all shell scripts in src tree.
876 NM/33 Bugzilla 898: Transport filter timeout fix.
877 Patch by Todd Rinaldo.
879 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
880 Patch by Serge Demonchaux.
882 NM/35 Bugzilla 39: Base64 decode bug fixes.
883 Patch by Jakob Hirsch.
885 NM/36 Bugzilla 909: Correct connect() call in dcc code.
887 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
889 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
891 NM/39 Bugzilla 911: Fixed MakeLinks build script.
897 TK/01 Add preliminary DKIM support. Currently requires a forked version of
898 ALT-N's libdkim that I have put here:
899 http://duncanthrax.net/exim-experimental/
901 Note to Michael Haardt: I had to rename some vars in sieve.c. They
902 were called 'true' and it seems that C99 defines that as a reserved
903 keyword to be used with 'bool' variable types. That means you could
904 not include C99-style headers which use bools without triggering
905 build errors in sieve.c.
907 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
908 as mailq or other aliases. Changed the --help handling significantly
909 to do whats expected. exim_usage() emits usage/help information.
911 SC/01 Added the -bylocaldomain option to eximstats.
913 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
915 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
917 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
923 PH/01 Another patch from the Sieve maintainer.
925 PH/02 When an IPv6 address is converted to a string for single-key lookup
926 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
927 dots are used instead of colons so that keys in lsearch files need not
928 contain colons. This was done some time before quoting was made available
929 in lsearch files. However, iplsearch files do require colons in IPv6 keys
930 (notated using the quote facility) so as to distinguish them from IPv4
931 keys. This meant that lookups for IP addresses in host lists did not work
932 for iplsearch lookups.
934 This has been fixed by arranging for IPv6 addresses to be expressed with
935 colons if the lookup type is iplsearch. This is not incompatible, because
936 previously such lookups could never work.
938 The situation is now rather anomolous, since one *can* have colons in
939 ordinary lsearch keys. However, making the change in all cases is
940 incompatible and would probably break a number of configurations.
942 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
945 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
946 conversion specification without a maximum field width, thereby enabling
947 a rogue spamd server to cause a buffer overflow. While nobody in their
948 right mind would setup Exim to query an untrusted spamd server, an
949 attacker that gains access to a server running spamd could potentially
950 exploit this vulnerability to run arbitrary code as the Exim user.
952 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
953 $primary_hostname instead of what libspf2 thinks the hosts name is.
955 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
956 a directory entry by the name of the lookup key. Previously, if a
957 symlink pointed to a non-existing file or a file in a directory that
958 Exim lacked permissions to read, a lookup for a key matching that
959 symlink would fail. Now it is enough that a matching directory entry
960 exists, symlink or not. (Bugzilla 503.)
962 PH/03 The body_linecount and body_zerocount variables are now exported in the
965 PH/04 Added the $dnslist_matched variable.
967 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
968 This means they are set thereafter only if the connection becomes
971 PH/06 Added the client_condition to authenticators so that some can be skipped
972 by clients under certain conditions.
974 PH/07 The error message for a badly-placed control=no_multiline_responses left
975 "_responses" off the end of the name.
977 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
979 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
980 (without spaces) instead of just copying the configuration text.
982 PH/10 Added the /noupdate option to the ratelimit ACL condition.
984 PH/11 Added $max_received_linelength.
986 PH/12 Added +ignore_defer and +include_defer to host lists.
988 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
989 way in which PCRE > 7.0 is built.
991 PH/14 Implemented queue_only_load_latch.
993 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
994 MAIL command. The effect was to mangle the value on 64-bit systems.
996 PH/16 Another patch from the Sieve maintainer.
998 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1000 PH/18 If a system quota error occurred while trying to create the file for
1001 a maildir delivery, the message "Mailbox is full" was not appended to the
1002 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1003 only to a quota excession during the actual writing of the file.
1005 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1006 characters?) which causes problems in log lines. The DN values are now
1007 passed through string_printing() before being added to log lines.
1009 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1010 and InterBase are left for another time.)
1012 PH/21 Added message_body_newlines option.
1014 PH/22 Guard against possible overflow in moan_check_errorcopy().
1016 PH/23 POSIX allows open() to be a macro; guard against that.
1018 PH/24 If the recipient of an error message contained an @ in the local part
1019 (suitably quoted, of course), incorrect values were put in $domain and
1020 $local_part during the evaluation of errors_copy.
1026 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1027 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1030 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1031 issue a MAIL command.
1033 PH/02 In an ACL statement such as
1035 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1037 if a client was not listed at all, or was listed with a value other than
1038 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1039 the condition was not true (as it should be), so access was not denied.
1040 The bug was that the ! inversion was incorrectly passed on to the second
1041 item. This has been fixed.
1043 PH/03 Added additional dnslists conditions == and =& which are different from
1044 = and & when the dns lookup returns more than one IP address.
1046 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1047 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1049 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1050 FSYNC, which compiles an option called disable_fsync that allows for
1051 bypassing fsync(). The documentation is heavily laced with warnings.
1053 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1055 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1056 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1057 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1058 including adding "make clean"; (3) Added -fPIC when compiling the test
1059 dynamically loaded module, to get rid of a warning.
1061 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1062 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1063 = 0s. The bug is otherwise harmless.
1065 PH/07 There was a bug in the dovecot authenticator such that the value of
1066 $auth1 could be overwritten, and so not correctly preserved, after a
1067 successful authentication. This usually meant that the value preserved by
1068 the server_setid option was incorrect.
1070 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1072 PH/09 Installed PCRE release 7.0.
1074 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1075 run for batched SMTP input. It is now run at the start of every message
1076 in the batch. While fixing this I discovered that the process information
1077 (output by running exiwhat) was not always getting set for -bs and -bS
1078 input. This is fixed, and it now also says "batched" for BSMTP.
1080 PH/11 Added control=no_pipelining.
1082 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1083 patch, slightly modified), and move the expansion of helo_data till after
1084 the connection is made in the smtp transport (so it can use these
1087 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1089 PH/14 Added log_selector = +pid.
1091 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1093 PH/16 Add ${if forany and ${if forall.
1095 PH/17 Added dsn_from option to vary the From: line in DSNs.
1097 PH/18 Flush SMTP output before performing a callout, unless control =
1098 no_callout_flush is set.
1100 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1101 was true (the default) a successful delivery failed to delete the retry
1102 item, thus causing premature timeout of the address. The bug is now
1105 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1107 PH/21 Long custom messages for fakedefer and fakereject are now split up
1108 into multiline reponses in the same way that messages for "deny" and
1109 other ACL rejections are.
1111 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1112 with slight modification.
1114 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1115 draft, changing the syntax and factoring some duplicate code".
1117 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1118 for deliveries of the second and subsequent messages over the same SMTP
1121 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1122 ${reduce, with only minor "tidies".
1124 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1126 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1127 expansion side effects.
1129 PH/27 When a message times out after an over-quota error from an Exim-imposed
1130 quota, the bounce message says "mailbox is full". This message was not
1131 being given when it was a system quota that was exceeded. It now should
1134 MH/03 Made $recipients available in local_scan(). local_scan() already has
1135 better access to the recipient list through recipients_list[], but
1136 $recipients can be useful in postmaster-provided expansion strings.
1138 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1139 in the case of a MAIL command with additional options following the
1140 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1141 were accidentally chopped off.
1143 PH/29 SMTP synchronization checks are implemented when a command is read -
1144 there is a check that no more input is waiting when there shouldn't be
1145 any. However, for some commands, a delay in an ACL can mean that it is
1146 some time before the response is written. In this time, more input might
1147 arrive, invalidly. So now there are extra checks after an ACL has run for
1148 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1149 pipelining has not been advertised.
1151 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1153 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1154 returned in the SMTP response when smtp_return_error_details was false.
1155 This has been fixed.
1157 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1158 instead of the C I/O that was originally supplied, because problems were
1159 reported on Solaris.
1161 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1162 Exim which did not show up earlier: it was assuming that a call to
1163 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1164 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1165 was a macro that became an assignment, so it seemed to work. This has
1166 changed to a proper function call with a void return, hence the compile
1167 error. Exim's code has been fixed.
1169 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1172 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1174 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1176 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1177 because it turns out to be too big or there is a timeout) there is no
1178 "Completed" line in the log. When some messages of this type were
1179 selected by exigrep, they were listed as "not completed". Others were
1180 picked up by some special patterns. I have improved the selection
1181 criteria to be more general.
1183 PH/38 The host_find_failed option in the manualroute router can now be set
1184 to "ignore", to completely ignore a host whose IP address cannot be
1185 found. If all hosts are ignored, the behaviour is controlled by the new
1186 host_all_ignored option.
1188 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1189 homing or because of multiple MX records with /mx) generated more than
1190 one IP address, and the following item turned out to be the local host,
1191 all the secondary addresses of the first item were incorrectly removed
1192 from the list, along with the local host and any following hosts (which
1193 is what is supposed to happen).
1195 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1196 whoever called Exim into the -H file. In the case of the daemon it was
1197 behaving confusingly. When first started, it used values for whoever
1198 started the daemon, but after a SIGHUP it used the Exim user (because it
1199 calls itself on a restart). I have changed the code so that it now always
1202 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1203 message are rejected with the same error (e.g. no authentication or bad
1204 sender address), and a DATA command is nevertheless sent (as can happen
1205 with PIPELINING or a stupid MUA), the error message that was given to the
1206 RCPT commands is included in the rejection of the DATA command. This is
1207 intended to be helpful for MUAs that show only the final error to their
1210 PH/42 Another patch from the Sieve maintainer.
1212 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1213 Eximstats - Fixed some broken HTML links and added missing column headers
1215 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1216 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1218 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1220 PH/43 Yet another patch from the Sieve maintainer.
1222 PH/44 I found a way to check for a TCP/IP connection going away before sending
1223 the response to the final '.' that terminates a message, but only in the
1224 case where the client has not sent further data following the '.'
1225 (unfortunately, this is allowed). However, in many cases there won't be
1226 any further data because there won't be any more messages to send. A call
1227 to select() can be used: if it shows that the input is "ready", there is
1228 either input waiting, or the socket has been closed. An attempt to read
1229 the next input character can distinguish the two cases. Previously, Exim
1230 would have sent an OK response which the client would never have see.
1231 This could lead to message repetition. This fix should cure that, at
1232 least in a lot of common cases.
1234 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1235 advertised in response to EHLO.
1241 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1242 fixed by 4.65/MH/01 (is this a record?) are fixed:
1244 (i) An empty string was always treated as zero by the numeric comparison
1245 operators. This behaviour has been restored.
1247 (ii) It is documented that the numeric comparison operators always treat
1248 their arguments as decimal numbers. This was broken in that numbers
1249 starting with 0 were being interpreted as octal.
1251 While fixing these problems I realized that there was another issue that
1252 hadn't been noticed. Values of message_size_limit (both the global option
1253 and the transport option) were treated as octal if they started with 0.
1254 The documentation was vague. These values are now always treated as
1255 decimal, and I will make that clear in the documentation.
1261 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1262 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1265 MH/01 Don't check that the operands of numeric comparison operators are
1266 integers when their expansion is in "skipping" mode (fixes bug
1267 introduced by 4.64-PH/07).
1269 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1270 child addresses, Exim now panics and dies. Previously, because the count
1271 is held in a short int, deliveries were likely to be lost. As such a
1272 large number of recipients for a single message is ridiculous
1273 (performance will be very, very poor), I have chosen to impose a limit
1274 rather than extend the field.
1280 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1281 leftover -K file (the existence of which was triggered by #402).
1282 While we were at it, introduced process PID as part of the -K
1283 filename. This should rule out race conditions when creating
1286 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1287 processing considerably. Previous code took too long for large mails,
1288 triggering a timeout which in turn triggers #401.
1290 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1291 in the DK code in transports.c. sendfile() is not really portable,
1292 hence the _LINUX specificness.
1294 TF/01 In the add_headers option to the mail command in an Exim filter,
1295 there was a bug that Exim would claim a syntax error in any
1296 header after the first one which had an odd number of characters
1299 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1300 callout verification, Exim cached a "reject" for the entire domain. This
1301 is correct for most verifications, but it is not correct for a recipient
1302 verification with use_sender or use_postmaster set, because in that case
1303 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1304 case of MAIL FROM:<> rejection from other early rejections (e.g.
1305 rejection of HELO). When verifying a recipient using a non-null MAIL
1306 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1307 Whatever the result of the callout, the value of the domain cache is
1308 left unchanged (for any other kind of callout, getting as far as trying
1309 RCPT means that the domain itself is ok).
1311 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1314 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1315 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1316 session, the daemon ignores ECONNECT errors and logs others; it now
1317 ignores EPIPE as well.
1319 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1320 (quoted-printable decoding).
1322 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1323 later the small subsequent patch to fix an introduced bug.
1325 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1327 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1329 PH/08 An error is now given if message_size_limit is specified negative.
1331 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1332 to be given (somewhat) arbitrary names.
1334 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1337 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1338 miscellaneous code fixes
1340 PH/10 Added the log_reject_target ACL modifier to specify where to log
1343 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1344 hostname. This is wrong, because it relates to the incoming message (and
1345 probably the interface on which it is arriving) and not to the outgoing
1346 callout (which could be using a different interface). This has been
1347 changed to use the value of the helo_data option from the smtp transport
1348 instead - this is what is used when a message is actually being sent. If
1349 there is no remote transport (possible with a router that sets up host
1350 addresses), $smtp_active_hostname is used.
1352 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1353 tweaks were necessary in order to get it to work (see also 21 below):
1354 (a) The code assumed that strncpy() returns a negative number on buffer
1355 overflow, which isn't the case. Replaced with Exim's string_format()
1357 (b) There were several signed/unsigned issues. I just did the minimum
1358 hacking in of casts. There is scope for a larger refactoring.
1359 (c) The code used strcasecmp() which is not a standard C function.
1360 Replaced with Exim's strcmpic() function.
1361 (d) The code set only $1; it now sets $auth1 as well.
1362 (e) A simple test gave the error "authentication client didn't specify
1363 service in request". It would seem that Dovecot has changed its
1364 interface. Fortunately there's a specification; I followed it and
1365 changed what the client sends and it appears to be working now.
1367 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1370 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1371 address A is aliased to B and C, where B exists and C does not. Without
1372 -v the output is "A verified" because verification stops after a
1373 successful redirection if more than one address is generated. However,
1374 with -v the child addresses are also verified. Exim was outputting "A
1375 failed to verify" and then showing the successful verification for C,
1376 with its parentage. It now outputs "B failed to verify", showing B's
1377 parentage before showing the successful verification of C.
1379 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1380 look up a TXT record in a specific list after matching in a combined
1383 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1384 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1385 they consult the DNS. I had assumed they would set it the way they
1386 wanted; and indeed my experiments on Linux seem to show that in some
1387 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1388 To be on the safe side, however, I have now made the interface to
1389 host_find_byname() similar to host_find_bydns(), with an argument
1390 containing the DNS resolver options. The host_find_byname() function now
1391 sets these options at its start, just as host_find_bydns() does. The smtp
1392 transport options dns_qualify_single and dns_search_parents are passed to
1393 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1394 of host_find_byname() use the default settings of RES_DEFNAMES
1395 (qualify_single) but not RES_DNSRCH (search_parents).
1397 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1398 spool_read_header() do less string testing, by means of a preliminary
1399 switch on the second character of optional "-foo" lines. (This is
1400 overdue, caused by the large number of possibilities that now exist.
1401 Originally there were few.) While I was there, I also converted the
1402 str(n)cmp tests so they don't re-test the leading "-" and the first
1403 character, in the hope this might squeeze out yet more improvement.
1405 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1406 flag allowing group syntax was set by the header_syntax check but not
1407 turned off, possible causing trouble later; (2) The flag was not being
1408 set at all for the header_verify test, causing "group"-style headers to
1409 be rejected. I have now set it in this case, and also caused header_
1410 verify to ignore an empty address taken from a group. While doing this, I
1411 came across some other cases where the code for allowing group syntax
1412 while scanning a header line wasn't quite right (mostly, not resetting
1413 the flag correctly in the right place). These bugs could have caused
1414 trouble for malformed header lines. I hope it is now all correct.
1416 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1417 with the "reply" argument non-NULL. The code, however (which originally
1418 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1419 but it didn't always do it. This confused somebody who was copying the
1420 code for some other use. I have removed all the tests.
1422 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1423 feature that was used to support insecure browsers during the U.S. crypto
1424 embargo. It requires special client support, and Exim is probably the
1425 only MTA that supported it -- and would never use it because real RSA is
1426 always available. This code has been removed, because it had the bad
1427 effect of slowing Exim down by computing (never used) parameters for the
1428 RSA_EXPORT functionality.
1430 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1431 authenticator to fail if there's a tab character in the incoming data
1432 (there should never be unless someone is messing about, as it's supposed
1433 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1434 if the connection is using TLS or if the remote IP is the same as the
1435 local IP, and the "valid-client-cert option" if a client certificate has
1438 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1439 authenticators. This can be used for authorization after authentication
1440 succeeds. (In the case of plaintext, it servers for both authentication
1443 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1444 if any retry times were supplied.
1446 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1447 connection, where there is no client IP address to check. In this
1448 situation, the verify now always succeeds.
1450 PH/25 Applied John Jetmore's -Mset patch.
1452 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1454 PH/27 In a string expansion for a processed (not raw) header when multiple
1455 headers of the same name were present, leading whitespace was being
1456 removed from all of them, but trailing whitespace was being removed only
1457 from the last one. Now trailing whitespace is removed from each header
1458 before concatenation. Completely empty headers in a concatenation (as
1459 before) are ignored.
1461 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1462 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1464 PH/29 [Removed. This was a change that I later backed out, and forgot to
1465 correct the ChangeLog entry (that I had efficiently created) before
1466 committing the later change.]
1468 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1469 address errors (4xx response to RCPT) over the same connection as other
1470 messages routed to the same hosts. Such deliveries are always "forced",
1471 so retry times are not inspected. This resulted in far too many retries
1472 for the affected addresses. The effect occurred only when there were more
1473 hosts than the hosts_max_try setting in the smtp transport when it had
1474 the 4xx errors. Those hosts that it had tried were not added to the list
1475 of hosts for which the message was waiting, so if all were tried, there
1476 was no problem. Two fixes have been applied:
1478 (i) If there are any address or message errors in an SMTP delivery, none
1479 of the hosts (tried or untried) are now added to the list of hosts
1480 for which the message is waiting, so the message should not be a
1481 candidate for sending over the same connection that was used for a
1482 successful delivery of some other message. This seems entirely
1483 reasonable: after all the message is NOT "waiting for some host".
1484 This is so "obvious" that I'm not sure why it wasn't done
1485 previously. Hope I haven't missed anything, but it can't do any
1486 harm, as the worst effect is to miss an optimization.
1488 (ii) If, despite (i), such a delivery is accidentally attempted, the
1489 routing retry time is respected, so at least it doesn't keep
1490 hammering the server.
1492 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1493 in ${readsocket because some servers need this prod.
1495 PH/32 Added some extra debug output when updating a wait-xxx database.
1497 PH/33 The hint "could be header name not terminated by colon", which has been
1498 given for certain expansion errors for a long time, was not being given
1499 for the ${if def:h_colon_omitted{... case.
1501 PH/34 The spec says: "With one important exception, whenever a domain list is
1502 being scanned, $domain contains the subject domain." There was at least
1503 one case where this was not true.
1505 PH/35 The error "getsockname() failed: connection reset by peer" was being
1506 written to the panic log as well as the main log, but it isn't really
1507 panic-worthy as it just means the connection died rather early on. I have
1508 removed the panic log writing for the ECONNRESET error when getsockname()
1511 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1512 runs only) independently of the message's sender address. This meant
1513 that, if the 4xx error was in fact related to the sender, a different
1514 message to the same recipient with a different sender could confuse
1515 things. In particualar, this can happen when sending to a greylisting
1516 server, but other circumstances could also provoke similar problems.
1517 I have changed the default so that the retry time for these errors is now
1518 based a combination of the sender and recipient addresses. This change
1519 can be overridden by setting address_retry_include_sender=false in the
1522 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1523 remote server are returned as part of bounce messages. This was not
1524 happening for LMTP over a pipe (the lmtp transport), but now it is the
1525 same for both kinds of LMTP.
1527 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1528 in header lines that were in fact CNAMEs. This is no longer the case.
1530 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1531 and queue runs started by the daemon processed all messages. This has
1532 been fixed so that -R and -S can now usefully be given with -q<time>.
1534 PH/40 Import PCRE release 6.7 (fixes some bugs).
1536 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1538 PH/42 Give an error if -q is specified more than once.
1540 PH/43 Renamed the variables $interface_address and $interface_port as
1541 $received_ip_address and $received_port, to make it clear that these
1542 values apply to message reception, and not to the outgoing interface when
1543 a message is delivered. (The old names remain recognized, of course.)
1545 PH/44 There was no timeout on the connect() call when using a Unix domain
1546 socket in the ${readsocket expansion. There now is.
1548 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1549 be meaningful with "accept".
1551 SC/01 Eximstats V1.43
1552 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1554 SC/02 Eximstats V1.44
1555 Use a glob alias rather than an array ref in the generated
1556 parser. This improves both readability and performance.
1558 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1559 Collect SpamAssassin and rejection statistics.
1560 Don't display local sender or destination tables unless
1561 there is data to show.
1562 Added average volumes into the top table text output.
1564 SC/04 Eximstats V1.46
1565 Collect data on the number of addresses (recipients)
1566 as well as the number of messages.
1568 SC/05 Eximstats V1.47
1569 Added 'Message too big' to the list of mail rejection
1570 reasons (thanks to Marco Gaiarin).
1572 SC/06 Eximstats V1.48
1573 Mainlog lines which have GMT offsets and are too short to
1574 have a flag are now skipped.
1576 SC/07 Eximstats V1.49 (Alain Williams)
1577 Added the -emptyok flag.
1579 SC/08 Eximstats V1.50
1580 Fixes for obtaining the IP address from reject messages.
1582 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1583 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1584 whitesspace changes from 4.64-PH/27
1586 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1589 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1590 are found, allow negative numbers in numeric criteria)
1592 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1594 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1595 to match changes made in 4.64-PH/43
1597 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1599 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1600 "domain" is actually an IP address, and force a failure. This locks out
1601 those revolvers/nameservers that support "A-for-A" lookups, in
1602 contravention of the specifications.
1604 PH/48 When a host name was looked up from an IP address, and the subsequent
1605 forward lookup of the name timed out, the host name was left in
1606 $sender_host_name, contrary to the specification.
1608 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1609 restricted to single-key lookups, Exim was not diagnosing an error if
1610 * or *@ was used with a query-style lookup.
1612 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1614 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1615 long ago, but noone interested enough thought of it. Let's just say that
1616 the "1.1" means that there are some new functions that weren't there at
1617 some point in the past.
1619 PH/51 Error processing for expansion failure of helo_data from an smtp
1620 transport during callout processing was broken.
1622 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1623 tested/used via the -bh/-bhc/-bs options.
1625 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1626 bug, fixed in subsequent PCRE releases).
1628 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1629 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1631 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1637 SC/01 Use a glob alias rather than an array ref in eximstats generated
1638 parser. This improves both readability and performance.
1640 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1641 Don't display local sender or destination tables in eximstats unless
1642 there is data to show.
1643 Added average volumes into the eximstats top table text output.
1645 SC/03 Collect data on the number of addresses (recipients) as well
1646 as the number of messages in eximstats.
1648 TF/01 Correct an error in the documentation for the redirect router. Exim
1649 does (usually) call initgroups() when daemonizing.
1651 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1652 with consistent privilege compared to when running as a daemon.
1654 TF/03 Note in the spec that $authenticated_id is not set for local
1655 submissions from trusted users.
1657 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1658 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1660 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1661 by adding some example configuration directives to the default
1662 configuration file. A little bit of work is required to uncomment the
1663 directives and define how usernames and passwords are checked, but
1664 there is now a framework to start from.
1666 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1667 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1668 without this. I don't know how relevant this is to other LDAP libraries.
1670 PH/02 Add the verb name to the "unknown ACL verb" error.
1672 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1674 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1676 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1677 directory not expanded when it should be if an expanded home directory
1678 was set for the address (which is overridden by the transport).
1680 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1683 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1684 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1685 because it is too late at that time, and has no effect.
1687 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1688 security issue with \' (bugzilla #107). I could not use the
1689 PQescapeStringConn() function, because it needs a PGconn value as one of
1692 PH/08 When testing addresses using -bt, indicate those final addresses that
1693 are duplicates that would not cause an additional delivery. At least one
1694 person was confused, thinking that -bt output corresponded to deliveries.
1695 (Suppressing duplicates isn't a good idea as you lose the information
1696 about possibly different redirections that led to the duplicates.)
1698 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1699 systems where poll() doesn't work, in particular OS X.
1701 PH/10 Added more information to debugging output for retry time not reached.
1703 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1704 operations in malware.c.
1706 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1709 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1710 syslog_duplication set false, log lines that would normally be written
1711 both the the main log and to the reject log were not written to syslog at
1714 PH/14 In the default configuration, change the use of "message" in ACL warn
1715 statements to "add_header".
1717 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1718 not followed by a command (e.g. "seen endif").
1720 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1721 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1724 PH/17 Added extra conditions to the default value of delay_warning_condition
1728 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1729 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1730 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1733 The Auto-Submitted: and various List- headers are standardised, whereas I
1734 don't think Precedence: ever was.
1736 PH/18 Refactored debugging code in route_finduser() to show more information,
1737 in particular, the error code if getpwnam() issues one.
1739 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1740 This is apparently needed in addition to the PH/07 change above to avoid
1741 any possible encoding problems.
1743 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1744 but not after initializing Perl.
1746 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1747 output them only if debugging. By default they are written stderr,
1748 apparently, which is not desirable.
1750 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1753 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1756 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1758 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1759 authenticated or an ident call has been made. Suppress the default
1760 values for $authenticated_id and $authenticated_sender (but permit -oMai
1761 and -oMas) when testing with -bh.
1763 PH/24 Re-jigged the order of the tests in the default configuration so that the
1764 tests for valid domains and recipients precede the DNS black list and CSA
1765 tests, on the grounds that those ones are more expensive.
1767 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1768 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1769 This bug exists in every version of Exim that I still have, right back to
1772 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1773 However, an attempt to turn on case-sensitivity in a regex key by
1774 including (?-i) didn't work because the subject string was already
1775 lowercased, and the effects were non-intuitive. It turns out that a
1776 one-line patch can be used to allow (?-i) to work as expected.
1782 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1783 other effects) broke the use of negated acl sub-conditions.
1785 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1788 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1789 "Deny" causes Exim to reject the incoming connection with a 554 error.
1790 Unfortunately, if there is a major crisis, such as a disk failure,
1791 tcp-wrappers gives "deny", whereas what one would like would be some
1792 kind of temporary error. A kludge has been added to help with this.
1793 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1794 554 error is used if errno is still zero or contains ENOENT (which occurs
1795 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1798 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1800 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1801 errors. Otherwise a message that provokes a temporary error (when other
1802 messages do not) can cause a whole host to time out.
1804 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1805 addresses were routed directly to files or pipes from a redirect router.
1806 File deliveries just didn't batch; pipe deliveries might have suffered
1809 PH/06 A failure to get a lock for a hints database would erroneously always say
1810 "Failed to get write lock", even when it was really a read lock.
1812 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1813 of 0600. This has been changed to use the value of the lockfile_mode
1814 option (which defaults to 0600).
1816 PH/08 Applied small patch from the Sieve maintainer.
1818 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1819 folder from quota calculations, a direct delivery into this folder messed
1820 up the contents of the maildirsize file. This was because the regex was
1821 used only to exclude .Trash (or whatever) when the size of the mailbox
1822 was calculated. There was no check that a delivery was happening into an
1823 excluded directory. This bug has been fixed by ignoring all quota
1824 processing for deliveries into excluded directories.
1826 PH/10 Added the maildirfolder_create_regex option to appendfile.
1832 PH/01 The code for finding all the local interface addresses on a FreeBSD
1833 system running IPv6 was broken. This may well have applied to all BSD
1834 systems, as well as to others that have similar system calls. The broken
1835 code found IPv4 interfaces correctly, but gave incorrect values for the
1836 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1837 that it would not match correctly against @[] and not recognize the IPv6
1840 PH/02 The ipliteral router was not recognizing addresses of the form user@
1841 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1843 PH/03 Added disable_ipv6.
1845 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1846 decoded form, because it is most often used to construct To: headers
1847 lines in autoreplies, and the decoded form may well be syntactically
1848 invalid. However, $reply_address has leading white space removed, and all
1849 newlines turned into spaces so that the autoreply transport does not
1852 PH/05 If group was specified without a user on a router, and no group or user
1853 was specified on a transport, the group from the router was ignored.
1855 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1856 for visible compile-time settings that can be used to change these
1857 numbers, for those that want even more. Backwards compatibility with old
1858 spool files has been maintained. However, going back to a previous Exim
1859 release will lost any variables that are in spool files.
1861 PH/07 Two small changes when running in the test harness: increase delay when
1862 passing a TCP/IP connection to a new process, in case the original
1863 process has to generate a bounce, and remove special handling of
1864 127.0.0.2 (sic), which is no longer necessary.
1866 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1867 be the same on different OS.
1869 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1872 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1873 whether --show-vars was specified or not
1875 JJ/02 exipick: Added support for new ACL variable spool format introduced
1878 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1879 syntactically invalid From: or Reply-to: line, and a filter used this to
1880 generate an autoreply, and therefore failed to obtain an address for the
1881 autoreply, Exim could try to deliver to a non-existent relative file
1882 name, causing unrelated and misleading errors. What now happens is that
1883 it logs this as a hard delivery error, but does not attempt to create a
1886 PH/11 The exinext utility has a -C option for testing purposes, but although
1887 the given file was scanned by exinext itself; it wasn't being passed on
1888 when Exim was called.
1890 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1891 an end-of-file indication when reading a command response.
1893 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1894 compiled. In many other places in Exim, IPv6 addresses are always
1895 recognized, so I have changed this. It also means that IPv4 domain
1896 literals of the form [IPV4:n.n.n.n] are now always recognized.
1898 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1899 used if the router is not running as root, for example, when verifying at
1900 ACL time, or when using -bh. The debugging output from this situation was
1901 non-existent - all you got was a failure to exec. I have made two
1904 (a) Failures to set uid/gid, the current directory, or a process leader
1905 in a subprocess such as that created by queryprogram now generate
1906 suitable debugging ouput when -d is set.
1908 (b) The queryprogram router detects when it is not running as root,
1909 outputs suitable debugging information if -d is set, and then runs
1910 the subprocess without attempting to change uid/gid.
1912 PH/15 Minor change to Makefile for building test_host (undocumented testing
1915 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1916 additional section of a DNS packet that returns MX or SRV records.
1917 Instead, it always explicitly searches for A/AAAA records. This avoids
1918 major problems that occur when a DNS server includes only records of one
1919 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1920 fixed another bug: if SRV records were looked up and the corresponding
1921 address records were *not* found in the additional section, the port
1922 values from the SRV records were lost.
1924 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1925 using the correct key (the original address) when searching the retry
1926 rules in order to find which one to use for generating the retry hint.
1928 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1929 adding the default one. Similarly, if it contains a Reply-To: header, the
1930 errors_reply_to option, if set, is not used.
1932 PH/19 When calculating a retry time, Exim used to measure the "time since
1933 failure" by looking at the "first failed" field in the retry record. Now
1934 it does not use this if it is later than than the arrival time of the
1935 message. Instead it uses the arrival time. This makes for better
1936 behaviour in cases where some deliveries succeed, thus re-setting the
1937 "first failed" field. An example is a quota failure for a huge message
1938 when small messages continue to be delivered. Without this change, the
1939 "time since failure" will always be short, possible causing more frequent
1940 delivery attempts for the huge message than are intended.
1941 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1943 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1944 $1, $2, $3) because the numerical variables can be reset during some
1945 expansion items (e.g. "match"), thereby losing the authentication data.
1947 PH/21 Make -bV show the size of off_t variables so that the test suite can
1948 decide whether to run tests for quotas > 2G.
1950 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1951 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1952 filecount value is greater than 2G or if a quota value is greater than 2G
1953 on a system where the size of off_t is not greater than 4, a panic error
1956 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1957 never match. The debug and -bh output now contains an explicit error
1958 message indicating a malformed IPv4 address or mask.
1960 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1961 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1962 PH/23 above applies.
1964 PH/25 Do not write to syslog when running in the test harness. The only
1965 occasion when this arises is a failure to open the main or panic logs
1966 (for which there is an explicit test).
1968 PH/26 Added the /no_tell option to "control=freeze".
1970 PH/27 If a host name lookup failed very early in a connection, for example, if
1971 the IP address matched host_lookup and the reverse lookup yielded a name
1972 that did not have a forward lookup, an error message of the form "no IP
1973 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1974 could be logged. Now it outputs the IP address instead of "NULL".
1976 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1977 allows the sender and the authenticated sender to be set when
1978 submitting a message from within Exim. Since child_open_exim() is
1979 documented for local_scan(), the new function should be too.
1981 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1982 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1983 results in an empty string is now treated as unset.
1985 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1987 PH/31 Added sender_verify_fail logging option.
1989 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1990 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1991 tidied the source and removed it altogether.
1993 PH/33 When a queue run was abandoned because the load average was too high, a
1994 log line was always written; now it is written only if the queue_run log
1995 selector is set. In addition, the log line for abandonment now contains
1996 information about the queue run such as the pid. This is always present
1997 in "start" and "stop" lines but was omitted from the "abandon" line.
1999 PH/34 Omit spaces between a header name and the colon in the error message that
2000 is given when verify = headers_syntax fails (if there are lots of them,
2001 the message gets confusing).
2003 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2004 names, as there are now some PTR records that contain slashes. This check
2005 is only to protect against broken name servers that fall over on strange
2006 characters, so the fact that it applies to all lookups doesn't matter.
2008 PH/36 Now that the new test suite is complete, we can remove some of the
2009 special code in Exim that was needed for the old test suite. For example,
2010 sorting DNS records because real resolvers return them in an arbitrary
2011 order. The new test suite's fake resolver always returns records in the
2014 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2015 bounces) except when queue_only is set, to avoid logging races between
2016 the different processes.
2018 PH/38 Panic-die if .include specifies a non-absolute path.
2020 PH/39 A tweak to the "H" retry rule from its user.
2022 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2023 a label. They prevented compilation on older perls.
2025 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2026 a warning to be raised on newish perls.
2028 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2029 on queue. Changes to match documented behaviour of showing count of
2030 messages matching specified criteria.
2032 PH/40 Changed the default ident timeout from 30s to 5s.
2034 PH/41 Added support for the use of login_cap features, on those BSD systems
2035 that have them, for controlling the resources used by pipe deliveries.
2037 PH/42 The content-scanning code uses fopen() to create files in which to put
2038 message data. Previously it was not paying any attention to the mode of
2039 the files. Exim runs with umask(0) because the rest of the code creates
2040 files with open(), and sets the required mode explicitly. Thus, these
2041 files were ending up world-writeable. This was not a big issue, because,
2042 being within the spool directory, they were not world-accessible. I have
2043 created a function called modefopen, which takes an additional mode
2044 argument. It sets umask(777), creates the file, chmods it to the required
2045 mode, then resets the umask. All the relevant calls to fopen() in the
2046 content scanning code have been changed to use this function.
2048 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2049 to 24 hours. This avoids potential overflow problems when processing G
2050 and H retry rules. I suspect nobody ever tinkers with this value.
2052 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2054 PH/45 When the plaintext authenticator is running as a client, the server's
2055 challenges are checked to ensure they are valid base64 strings. By
2056 default, the authentication attempt is cancelled if an invalid string is
2057 received. Setting client_ignore_invalid_base64 true ignores these errors.
2058 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2059 they are received. Thus, the responses can be made to depend on the
2060 challenges. If an invalid string is ignored, an empty string is placed in
2063 PH/46 Messages that are created by the autoreply transport now contains a
2064 References: header, in accordance with RFCs 2822 and 3834.
2066 PH/47 Added authenticated_sender_force to the smtp transport.
2068 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2070 PH/49 Installed latest patch from the Sieve maintainer.
2072 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2073 was also set, the appendfile transport was unnecessarily scanning a
2074 directory of message files (e.g. for maildir delivery) to find the count
2075 of files (along with the size), even though it did not need this
2076 information. It now does the scan only if it needs to find either the
2077 size of the count of files.
2079 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2081 PH/52 Two bugs concerned with error handling when the smtp transport is
2084 (i) Exim was not creating retry information for temporary errors given
2085 for individual recipients after the DATA command when the smtp transport
2086 was used in LMTP mode. This meant that they could be retried too
2087 frequently, and not timed out correctly.
2089 (ii) Exim was setting the flag that allows error details to be returned
2090 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2091 recipients that were returned after the DATA command.
2093 PH/53 This is related to PH/52, but is more general: for any failing address,
2094 when detailed error information was permitted to be returned to the
2095 sender, but the error was temporary, then after the final timeout, only
2096 "retry timeout exceeded" was returned. Now it returns the full error as
2097 well as "retry timeout exceeded".
2099 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2100 do this, and (what is worse) MTAs that accept it.
2102 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2103 will now be deprecated.
2105 PH/56 New os.c-cygwin from the Cygwin maintainer.
2107 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2108 formats (previously only available in exim formats via -bpr, -bpru,
2109 and -bpra. Now also available in native and exiqgrep formats)
2111 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2112 with very large, slow to parse queues
2114 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2116 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2118 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2119 responses to authentication challenges, though it was showing the
2120 challenges; (ii) I've removed the CR characters from the debug output for
2123 PH/58 Allow for the insertion of a newline as well as a space when a string
2124 is turned into more than one encoded-word during RFC 2047 encoding. The
2125 Sieve code now uses this.
2127 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2128 data_4xx, lost_connection, tls_required.
2130 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2131 message was being sent as an SMTP response.
2133 PH/61 Add -l and -k options to exicyclog.
2135 PH/62 When verifying, if an address was redirected to one new address, so that
2136 verification continued, and the new address failed or deferred after
2137 having set something in $address_data, the value of $address_data was not
2138 passed back to the ACL. This was different to the case when no
2139 redirection occurred. The value is now passed back in both cases.
2141 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2142 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2143 use login_cap.h, so on its own it isn't the distinguishing feature. The
2144 new name refers directly to the setclassresources() function.
2146 PH/65 Added configuration files for NetBSD3.
2148 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2150 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2151 is preferred over IPv4.
2153 PH/68 The bounce_return_message and bounce_return_body options were not being
2154 honoured for bounces generated during the reception of non-SMTP messages.
2155 In particular, this applied to messages rejected by the ACL. This bug has
2156 been fixed. However, if bounce_return_message is true and bounce_return_
2157 body is false, the headers that are returned for a non-SMTP message
2158 include only those that have been read before the error was detected.
2159 (In the case of an ACL rejection, they have all been read.)
2161 PH/69 The HTML version of the specification is now built in a directory called
2162 spec_html instead of spec.html, because the latter looks like a path with
2163 a MIME-type, and this confuses some software.
2165 PH/70 Catch two compiler warnings in sieve.c.
2167 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2168 function verify_get_ident() calls ip_connect() to connect a socket, but
2169 if the "connect()" function timed out, ip_connect() used to close the
2170 socket. However, verify_get_ident() also closes the socket later, and in
2171 between Exim writes to the log, which may get opened at this point. When
2172 the socket was closed in ip_connect(), the log could get the same file
2173 descriptor number as the socket. This naturally causes chaos. The fix is
2174 not to close the socket in ip_connect(); the socket should be closed by
2175 the function that creates it. There was only one place in the code where
2176 this was missing, in the iplookup router, which I don't think anybody now
2177 uses, but I've fixed it anyway.
2179 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2180 well as to direct DNS lookups. Otherwise the handling of names in host
2181 lists is inconsistent and therefore confusing.
2187 PH/01 Two changes to the default runtime configuration:
2189 (1) Move the checks for relay_from_hosts and authenticated clients from
2190 after to before the (commented out) DNS black list checks.
2192 (2) Add control=submission to the relay_from_hosts and authenticated
2193 clients checks, on the grounds that messages accepted by these
2194 statements are most likely to be submissions.
2196 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2198 (1) Generate an error if the third argument for the ${prvs expansion is
2201 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2204 (3) Reset the variables that are obtained from the first argument of
2205 ${prvscheck and used in the second argument before leaving the code,
2206 because their memory is reclaimed, so using them afterwards may do
2209 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2210 one (it's much easier than Tom thought :-).
2212 (5) Because of (4), we can now allow for the use of $prvscheck_result
2213 inside the third argument.
2215 PH/03 For some reason, the default setting of PATH when running a command from
2216 a pipe transport was just "/usr/bin". I have changed it to
2219 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2220 anything to be listed in the output from -bV.
2222 PH/05 When a filter generated an autoreply, the entire To: header line was
2223 quoted in the delivery log line, like this:
2225 => >A.N.Other <ano@some.domain> <original@ddress> ...
2227 This has been changed so that it extracts the operative address. There
2228 may be more than one such address. If so, they are comma-separated, like
2231 => >ano@some.domain,ona@other.domain <original@ddress> ...
2233 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2234 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2235 not being looked up in the rDNS to get a host name, Exim was showing the
2236 IP address twice in Received: lines, even though the IP addresses were
2237 identical. For example:
2239 Received: from [1.2.3.4] (helo=[1.2.3.4])
2241 However, if the real host name was known, it was omitting the HELO data
2242 if it matched the actual IP address. This has been tidied up so that it
2243 doesn't show the same IP address twice.
2245 PH/07 When both +timestamp and +memory debugging was on, the value given by
2246 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2247 called by the debug printing, thereby overwriting the timestamp buffer.
2248 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2250 PH/08 When the original message was included in an autoreply transport, it
2251 always said "this is a copy of the message, including all the headers",
2252 even if body_only or headers_only was set. It now gives an appropriate
2255 PH/09 Applied a patch from the Sieve maintainer which:
2257 o fixes some comments
2258 o adds the (disabled) notify extension core
2259 o adds some debug output for the result of if/elsif tests
2260 o points to the current vacation draft in the documentation
2261 and documents the missing references header update
2265 o fixes a bug in processing the envelope test (when testing
2266 multiple envelope elements, the last element determinted the
2269 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2270 Electronic Mail") by including:
2272 Auto-submitted: auto-generated
2274 in the messages that it generates (bounce messages and others, such as
2275 warnings). In the case of bounce messages for non-SMTP mesages, there was
2276 also a typo: it was using "Auto_submitted" (underscore instead of
2277 hyphen). Since every message generated by Exim is necessarily in response
2278 to another message, thes have all been changed to:
2280 Auto-Submitted: auto-replied
2282 in accordance with these statements in the RFC:
2284 The auto-replied keyword:
2286 - SHOULD be used on messages sent in direct response to another
2287 message by an automatic process,
2289 - MUST NOT be used on manually-generated messages,
2291 - MAY be used on Delivery Status Notifications (DSNs) and Message
2292 Disposition Notifications (MDNs),
2294 - MUST NOT be used on messages generated by automatic or periodic
2295 processes, except for messages which are automatic responses to
2298 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2299 to the default Received: header definition.
2301 PH/12 Added log selector acl_warn_skipped (default on).
2303 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2304 variables because (a) they are in the wrong storage pool and (b) even if
2305 they were copied, it wouldn't work properly because of the caching.
2307 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2308 checking when decoding. Apparently there are clients that generate
2309 overlong encoded strings. Why am I not surprised?
2311 PH/15 If the first argument of "${if match_address" was not empty, but did not
2312 contain an "@" character, Exim crashed. Now it writes a panic log message
2313 and treats the condition as false.
2315 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2317 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2318 extension "envelope-auth". The code is finished and in agreement with
2319 other implementations, but there is no documentation so far and in fact,
2320 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2321 not changing the active code.
2323 Print executed "if" and "elsif" statements when debugging is used. This
2324 helps a great deal to understand what a filter does.
2326 Document more things not specified clearly in RFC3028. I had all this
2327 sorted out, when out of a sudden new issues came to my mind. Oops."
2329 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2332 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2333 canonical form (as documented). However, after a host name lookup from
2334 the IP address, check_host() was doing a simple string comparison with
2335 addresses acquired from the DNS when checking that the found name did
2336 have the original IP as one of its addresses. Since any found IPv6
2337 addresses are likely to be in abbreviated form, the comparison could
2338 fail. Luckily, there already exists a function for doing the comparison
2339 by converting both addresses to binary, so now that is used instead of
2340 the text comparison.
2342 PH/20 There was another similar case to PH/19, when a complete host name was
2343 given in a host list; looking up its IP address could give an abbreviated
2344 form, whereas the current host's name might or might not be abbreviated.
2345 The same fix has been applied.
2351 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2352 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2355 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2356 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2358 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2360 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2361 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2362 Also, for "Auto-Submitted", treat anything other than "no" as
2363 non-personal, in accordance with RFC 3834. (Previously it treated
2364 anything starting "auto-" as non-personal.)
2366 TF/01 The control=submission/name=... option had a problem with syntax
2367 errors if the name included a slash character. The /name= option
2368 now slurps the rest of the string, so it can include any characters
2369 but it must come last in the list of options (after /sender_retain
2372 PH/05 Some modifications to the interface to the fake nameserver for the new
2380 TK/01 Added the "success_on_redirect" address verification option. See
2381 NewStuff for rationale and an example.
2383 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2385 PH/02 Patch to exigrep to allow it to work on syslog lines.
2387 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2388 fread() to skip over the body file's header line, because in Cygwin the
2389 header line is locked and is inaccessible.
2391 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2392 co-exist for some time) to make it clear that it is the Exim ID that is
2393 referenced, not the Message-ID: header line.
2395 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2396 string_format() function, because snprintf() does not exist on all
2399 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2400 ${stat: expansion item.
2402 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2403 protocol synchronization error", to keep the pedants happy.
2405 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2406 well as for IRIX systems, when gcc is being used. See the host.c source
2409 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2411 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2414 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2415 transport and to the smtp transport in LMTP mode.
2417 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2419 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2420 run a filter in a subprocess. This could lead to confusion in subsequent
2421 lookups in the parent process. There should also be a search_tidyup() at
2422 the end of the subprocess.
2424 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2425 only if the host matched helo_try_verify_hosts, which caused the
2426 verification to occur when the EHLO/HELO command was issued. The ACL just
2427 tested the remembered result. Now, if a previous verification attempt has
2428 not happened, "verify = helo" does it there and then.
2430 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2432 TK/03 Fix log output including CR from clamd.
2434 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2435 exist provoked a memory error which could cause a segfault.
2437 PH/15 Installed PCRE 6.2
2439 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2441 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2442 of the problem. Specifically, suggested +O2 rather than +O1 for the
2445 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2447 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2448 record, the port was not correctly logged when the outgoing_port log
2449 selector was set (it logged the transort's default port).
2451 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2452 fallback_hosts, and "hosts" in the smtp transport.
2454 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2455 host errors such as "Connection refused".
2457 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2458 authentication with radiusclient 0.4.9:
2460 - Error returned from rc_read_config was caught wrongly
2461 - Username/password not passed on to radius server due to wrong length.
2463 The presumption is that some radiusclient API changes for 4.51/PH/17
2464 were not taken care of correctly. The code is still untested by me (my
2465 Linux distribution still has 0.3.2 of radiusclient), but it was
2466 contributed by a Radius user.
2468 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2469 expanding the "port" option of the smtp transport.
2471 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2472 while reading a MIME header. Thanks to Tom Hughes for a patch.
2474 PH/24 Include config.h inside local_scan.h so that configuration settings are
2477 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2478 that in an ACL for RCPT (for example), you can examine exactly what was
2481 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2482 commands, but it was not correctly comparing the address with the actual
2483 client host address. Thus, it would show the EHLO address in Received:
2484 header lines when this was not necessary.
2486 PH/27 Added the % operator to ${eval:}.
2488 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2489 it should be ignoring failures (because with -C, for example, it has lost
2490 privilege). It wasn't ignoring creation failures other than "already
2493 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2496 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2497 because an input line was too long, either on its own, or by virtue of
2498 too many continuations, the temporary file was not being removed, and the
2499 return code was incorrect.
2501 PH/31 Missing "BOOL" in function definition in filtertest.c.
2503 PH/32 Applied Sieve patches from the maintainer.
2505 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2507 PH/33 Added "verify = not_blind".
2509 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2510 Local/Makefile (with some defaults set). These are used in built scripts
2511 such as exicyclog, but they have never been used in the exim_install
2512 script (though there are many overriding facilities there). I have
2513 arranged that the exim_install script now takes note of these two
2516 PH/35 Installed configuration files for Dragonfly.
2518 PH/36 When a locally submitted message by a trusted user did not contain a
2519 From: header, and the sender address was obtained from -f or from an SMTP
2520 MAIL command, and the trusted user did not use -F to supply a sender
2521 name, $originator_name was incorrectly used when constructing a From:
2522 header. Furthermore, $originator_name was used for submission mode
2523 messages from external hosts without From: headers in a similar way,
2524 which is clearly wrong.
2526 PH/37 Added control=suppress_local_fixups.
2528 PH/38 When log_selector = +received_sender was set, and the addition of the
2529 sender made the log line's construction buffer exactly full, or one byte
2530 less than full, an overflow happened when the terminating "\n" was
2533 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2534 when the result of a list match is failure because a DNS lookup failed.
2536 PH/40 RM_COMMAND is now used in the building process.
2538 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2539 the "build-* directories that it finds.
2541 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2542 address in a domain literal was a prefix of an interface address.
2544 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2545 when verifying a sender address, unless rewrite_headers is false.
2547 PH/44 Wrote a long comment about why errors_to addresses are verified as
2548 recipients, not senders.
2550 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2551 the ratelimit ACL was added.
2553 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2555 PH/46 Added extra information about PostgreSQL errors to the error string.
2557 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2558 avoiding the need to install special zones in a real server. This is
2559 backwards compatible; if it can't find the fake resolver, it drops back.
2560 Thus, both old and new test suites can be run.
2562 TF/02 Added util/ratelimit.pl
2564 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2565 clock is set back in time.
2567 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2568 Candler <B.Candler@pobox.com>.
2570 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2571 OK for addresses that are the result of redirections.
2573 PH/48 A number of further additions for the benefit of the new test suite,
2574 including a fake gethostbyname() that interfaces to the fake DNS resolver
2577 TF/06 The fix for widen_domains has also been applied to qualify_single and
2578 search_parents which are the other dnslookup options that can cause
2581 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2584 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2586 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2587 service" if it encountered an MX record with an empty target hostname.
2588 The message is now "an MX or SRV record indicated no SMTP service".
2590 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2591 if the DNS of the sending site is misconfigured. This is quite a
2592 common situation. This change restores the behaviour of treating a
2593 helo verification defer as a failure.
2595 PH/51 If self=fail was set on a router, the bounce message did not include the
2596 actual error message.
2602 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2604 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2605 command itself ended in error, the underlying message about the transport
2606 filter timeout was being overwritten with the pipe command error. Now the
2607 underlying error message should be appended to the second error message.
2609 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2611 PH/02 Reduce the amount of output that "make" produces by default. Full output
2612 can still be requested.
2614 PH/03 The warning log line about a condition test deferring for a "warn" verb
2615 was being output only once per connection, rather than after each
2616 occurrence (because it was using the same function as for successful
2617 "warn" verbs). This seems wrong, so I have changed it.
2619 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2620 it should not have, which might have caused a crash in the right
2621 circumstances, but probably never did.
2623 PH/04 Installed a modified version of Tony Finch's patch to make submission
2624 mode fix the return path as well as the Sender: header line, and to
2625 add a /name= option so that you can make the user's friendly name appear
2628 TF/03 Added the control = fakedefer ACL modifier.
2630 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2631 Mark Lowes for thorough testing.
2633 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2635 TK/03 Merged latest SRS patch from Miles Wilton.
2637 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2638 with the definition in sysexits.h (which is #included earlier).
2639 Fortunately, Exim does not actually use EX_OK. The code used to try to
2640 preserve the sysexits.h value, by assumimg that macro definitions were
2641 scanned for macro replacements. I have been disabused of this notion,
2642 so now the code just undefines EX_OK before #including unistd.h.
2644 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2645 in the smtp transport. When a block could not be written in a single
2646 write() function, the timeout was being re-applied to each part-write.
2647 This seems wrong - if the receiver was accepting one byte at a time it
2648 would take for ever. The timeout is now adjusted when this happens. It
2649 doesn't have to be particularly precise.
2651 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2652 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2654 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2655 without a domain if the check to <postmaster@domain> fails.
2657 SC/01 Eximstats: added -xls and the ability to specify output files
2658 (patch written by Frank Heydlauf).
2660 SC/02 Eximstats: use FileHandles for outputing results.
2662 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2664 SC/04 Eximstats: fixed display of large numbers with -nvr option
2666 SC/05 Eximstats: fixed merging of reports with empty tables.
2668 SC/06 Eximstats: added the -include_original_destination flag
2670 SC/07 Eximstats: removed tabs and trailing whitespace.
2672 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2674 TK/06 MBOX spool code: Add real "From " MBOX separator line
2675 so the .eml file is really in mbox format (even though
2676 most programs do not really care). Patch from Alex Miller.
2678 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2679 The latter is generated from $received_to and is only set if the
2680 message has one envelope recipient. SA can use these headers,
2681 obviously out-of-the-box. Patch from Alex Miller.
2683 PH/08 The ${def test on a variable was returning false if the variable's
2684 value was "0", contrary to what the specification has always said!
2685 The result should be true unless the variable is empty.
2687 PH/09 The syntax error of a character other than { following "${if
2688 def:variable_name" (after optional whitespace) was not being diagnosed.
2689 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2690 accidental colon was present, for example, could give incorrect results.
2692 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2693 result is used (not including appendfile, where other changes are about
2696 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2697 This involved changing a lot of size variables from int to off_t. It
2698 should work with maildirs and everything.
2700 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2701 spamd dying while we are connected to it.
2703 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2706 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2707 function for BDB 4.3.
2709 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2711 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2712 the sources for PCRE have been re-arranged and more files are now
2715 PH/15 The code I had for printing potentially long long variables in PH/11
2716 above was not the best (it lost precision). The length of off_t variables
2717 is now inspected at build time, and an appropriate printing format (%ld
2718 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2719 to be "long long int" or "long int". This is needed for the internal
2720 formatting function string_vformat().
2722 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2723 the configuration file to be ":syslog", then the script "guesses" where
2724 the logs files are, rather than using the compiled in default. In our
2725 case the guess is not the same as the compiled default, so the script
2726 suddenly stopped working when I started to use syslog. The patch checks
2727 to see if log_file_path is "". If so, it attempts to read it from exim
2728 with no configuration file to get the compiled in version, before it
2729 falls back to the previous guessing code."
2731 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2732 implementing BATV in an Exim configuration. See NewStuff for the gory
2735 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2736 Makefile that are specific to HP-UX.
2738 PH/18 If the "use_postmaster" option was set for a recipient callout together
2739 with the "random" option, the postmaster address was used as the MAIL
2740 FROM address for the random test, but not for the subsequent recipient
2741 test. It is now used for both.
2743 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2744 patch removes a few documentation additions to RFC 3028, because the
2745 latest draft now contains them. It adds the new en;ascii-case comparator
2746 and a new error check for 8bit text in MIME parts. Comparator and
2747 require names are now matched exactly. I enabled the subaddress
2748 extension, but it is not well tested yet (read: it works for me)."
2750 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2751 rework some of the code of TK/09 above to avoid the hardwired use of
2752 "%lld" and "long long". Replaced the call to snprintf() with a call to
2755 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2756 records point to non-existent hosts", "retry timeout exceeded", and
2757 "retry time not reached for any host after a long failure period".
2759 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2760 experimental DomainKeys support:
2762 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2763 (2) On an error such as an illegally used "control", the wrong name for
2764 the control was given.
2766 These problems did NOT occur unless DomainKeys support was compiled.
2768 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2770 PH/24 Added ${if match_ip condition.
2772 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2773 databases so that it will be absolutely obvious if a crash occurs in the
2774 DB library. This is a regular occurrence (often caused by mis-matched
2777 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2778 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2779 picked up on a user's system that detects such things. There doesn't seem
2780 to be a gcc warning option for this - only an attribute that has to be
2781 put on the function's prototype. It seems that in Fedora Core 4 they have
2782 set this on a number of new functions. No doubt there will be more in due
2785 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2786 specify a transport. However, if an address that was verified by such a
2787 router was the subject of a callout, Exim crashed because it tried to
2788 read the rcpt_include_affixes from the non-existent transport. Now it
2789 just assumes that the setting of that option is false. This bug was
2790 introduced by 4.51/PH/31.
2792 PH/28 Changed -d+all to exclude +memory, because that information is very
2793 rarely of interest, but it makes the output a lot bigger. People tend to
2794 do -d+all out of habit.
2796 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2797 code in os-type was giving problems when libc.so lives in lib64, like on
2800 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2801 aren't the modern standard, and it seems that some systems' include files
2802 don't always have them. Exim was already checking for some of the newer
2803 ones like T_AAAA, and defining it itself. I've added checks for all the
2804 record types that Exim uses.
2806 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2807 not automatically generating a new one, as it is supposed to. This
2808 prevented TLS from working. If the file did exist, but contained invalid
2809 data, a new version was generated, as expected. It was only the case of a
2810 non-existent file that was broken.
2812 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2813 with a change in libdomainkeys > 0.64.
2815 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2816 from DNS. If the selector record carries the flag, it now has
2817 precedence over the domain-wide flag.
2819 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2821 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2822 the use of a port name that isn't defined in /etc/services) occurred, the
2823 message was deferred as in a normal delivery, and thus remained on the
2824 spool, instead of being failed because of the mua_wrapper setting. This
2825 is now fixed, and I tidied up some of the mua_wrapper messages at the
2828 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2829 the messages in a hash of arrays rather than using individual hashes.
2830 This is a bit cleaner and results in dramatic memory savings, albeit
2831 at a slight CPU cost.
2833 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2834 as requested by Marc Sherman.
2836 SC/10 Eximstats: added histograms for user specified patterns as requested
2839 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2841 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2842 fopen() in the content-scanning modules that did not already have it.
2848 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2849 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2851 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2853 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2855 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2856 to test Sieve filters that use "vacation".
2858 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2859 that changes the way the GnuTLS parameters are stored in the cache file.
2860 The new format can be generated externally. For backward compatibility,
2861 if the data in the cache doesn't make sense, Exim assumes it has read an
2862 old-format file, and it generates new data and writes a new file. This
2863 means that you can't go back to an older release without removing the
2866 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2867 work if there are any delivery delays because "one_time" forces the
2868 parent to be marked "delivered", so its unseen clone is never tried
2869 again. For this reason, Exim now forbids the simultaneous setting of
2872 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2873 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2874 entry if you want to know the details. The fix, however, neglected to
2875 consider the case when local delivery batching is involved. The test for
2876 "previously delivered" was not happening when checking to see if an
2877 address could be batched with a previous (undelivered) one; under
2878 certain circumstances this could lead to multiple deliveries to the same
2881 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2882 in its include files, and this causes problems building Exim.
2884 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2885 header_syntax) but Exim was just ignoring anything given after a slash.
2886 In particular, this caused confusion with an attempt to use "verify =
2887 reverse_host_lookup/defer_ok". An error is now given when options are
2888 supplied for verify items that do not have them. (Maybe reverse_host_
2889 lookup should have a defer_ok option, but that's a different point.)
2891 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2892 defined by RFC 821) to 2048, because there were problems with some AUTH
2893 commands, and RFC 1869 says the size should be increased for extended
2894 SMTP commands that take arguments.
2896 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2899 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2900 "unknown" error; now it says that the functionality isn't in the binary.
2902 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2903 an address' error message when a string expansion fails (syntax or
2904 whatever). Otherwise the password may appear in the log. Following change
2905 PH/42 below, there is no longer a chance of it appearing in a bounce
2908 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2910 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2911 of its addresses was ever tried. (Bugzilla bug #2.)
2913 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2914 the result incorrectly in the debug output. (It correctly added a newline
2915 to what was transported.)
2917 TF/01 Added $received_time.
2919 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2920 commented out examples of how to interface to a virus scanner and to
2921 SpamAssassin. Also added commented examples of av_scanner and
2922 spamd_address settings.
2924 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2925 and controls are allowed in which ACLs. There were a couple of minor
2926 errors. Some of the entries in the conditions table (which is a table of
2927 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2928 negation of where the condition IS allowed.
2930 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2932 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2933 header file does not have a version number, so I've had to invent a new
2934 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2935 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2936 radiusclient), but it was contributed by a Radius user.
2938 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2939 files or directories, for OpenSSL.
2941 PH/19 When an Exim process that is running as root has to create an Exim log
2942 file, it does so in a subprocess that runs as exim:exim so as to get the
2943 ownership right at creation (otherwise, other Exim processes might see
2944 the file with the wrong ownership). There was no test for failure of this
2945 fork() call, which would lead to the process getting stuck as it waited
2946 for a non-existent subprocess. Forks do occasionally fail when resources
2947 run out. I reviewed all the other calls to fork(); they all seem to check
2950 PH/20 When checking for unexpected SMTP input at connect time (before writing
2951 the banner), Exim was not dealing correctly with a non-positive return
2952 from the read() function. If the client had disconnected by this time,
2953 the result was a log entry for a synchronization error with an empty
2954 string after "input=" when read() returned zero. If read() returned -1
2955 (an event I could not check), uninitialized data bytes were printed.
2956 There were reports of junk text (parts of files, etc) appearing after
2959 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2961 PH/22 Added support for macro redefinition, and (re)definition in between
2962 driver and ACL definitions.
2964 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2965 forgetting to use the resulting value; it was using the unexpanded value.
2967 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2968 hadn't been configured. The fix is from Juergen Kreileder, who
2969 understands it better than I do:
2971 "Here's what I see happening with three configured cyrus_sasl
2972 authenticators configured (plain, login, cram-md5):
2974 On startup auth_cyrus_sasl_init() gets called for each of these.
2975 This means three calls to sasl_listmech() without a specified mech_list.
2976 => SASL tests which mechs of all available mechs actually work
2977 => three warnings about OTP not working
2978 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2980 With the patch, sasl_listmech() also gets called three times. But now
2981 SASL's mech_list option is set to the server_mech specified in the the
2982 authenticator. Or in other words, the answer from sasl_listmech()
2983 gets limited to just the mech you're testing for (which is different
2985 => the return list contains just 'plain' or 'login', 'cram-md5' or
2986 nothing depending on the value of ob->server_mech.
2988 I've just tested the patch: Authentication still works fine,
2989 unavailable mechs specified in the exim configuration are still
2990 caught, and the auth.log warnings about OTP are gone."
2992 PH/25 When debugging is enabled, the contents of the command line are added
2993 to the debugging output, even when log_selector=+arguments is not
2996 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2997 answer is "GNU", and only if the return is "GNU/something" is the answer
3000 PH/27 $acl_verify_message is now set immediately after the failure of a
3001 verification in an ACL, and so is available in subsequent modifiers. In
3002 particular, the message can be preserved by coding like this:
3004 warn !verify = sender
3005 set acl_m0 = $acl_verify_message
3007 Previously, $acl_verify_message was set only while expanding "message"
3008 and "log_message" when a very denied access.
3010 PH/28 Modified OS/os.c-Linux with
3012 -#ifndef OS_LOAD_AVERAGE
3013 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3015 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3016 nomenclature these days.)
3018 PH/29 Installed patch from the Sieve maintainer that adds the options
3019 sieve_useraddress and sieve_subaddress to the redirect router.
3021 PH/30 In these circumstances:
3022 . Two addresses routed to the same list of hosts;
3023 . First host does not offer TLS;
3024 . First host accepts first address;
3025 . First host gives temporary error to second address;
3026 . Second host offers TLS and a TLS session is established;
3027 . Second host accepts second address.
3028 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3029 and peerdn, if requested) that were in fact used only for the second
3032 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3033 attention to any local part prefix or suffix that was matched by the
3034 router that accepted the address. It now behaves in the same way as it
3035 does for delivery: the affixes are removed from the local part unless
3036 rcpt_include_affixes is set on the transport.
3038 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3039 timeout during the DATA phase of an incoming message.
3041 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3042 applied a patch sanctioned by the Sieve maintainer.
3044 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3045 the uid or gid is negative. A case of a negative gid caused this to be
3046 noticed. The fix allows for either to be negative.
3048 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3049 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3052 PH/36 The change PH/12 above was broken. Fixed it.
3054 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3055 the grounds that routing the same address twice would always produce the
3056 same answer. This might have been true once, but it is certainly no
3057 longer true now. Routing a child address may depend on the previous
3058 routing that produced that child. Some complicated redirection strategies
3059 went wrong when messages had multiple recipients, and made Exim's
3060 behaviour dependent on the order in which the addresses were given.
3062 I have moved the duplicate checking until after the routing is complete.
3063 Exim scans the addresses that are assigned to local and remote
3064 transports, and removes any duplicates. This means that more work will be
3065 done, as duplicates will always all be routed, but duplicates are
3066 presumably rare, so I don't expect this is of any significance.
3068 For deliveries to pipes, files, and autoreplies, the duplicate checking
3069 still happens during the routing process, since they are not going to be
3072 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3073 It corrects a timeout issue with spamd. This is Ian's comment: "The
3074 background is that sometimes spamd either never reads data from a
3075 connection it has accepted, or it never writes response data. The exiscan
3076 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3077 blindly assumes that writes won't block so it may never time out."
3079 PH/39 Allow G after quota size as well as K and M.
3081 PH/40 The value set for $authenticated_id in an authenticator may not contain
3082 binary zeroes or newlines because the value is written to log lines and
3083 to spool files. There was no check on this. Now the value is run through
3084 the string_printing() function so that such characters are converted to
3085 printable escape sequences.
3087 PH/41 $message_linecount is a new variable that contains the total number of
3088 lines in the message. Compare $body_linecount, which is the count for the
3091 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3092 bounce and delay warning messages, except in certain special cases, which
3095 (a) An SMTP error message from a remote host;
3096 (b) A message specified in a :fail: redirection;
3097 (c) A message specified in a "fail" command in a system filter;
3098 (d) A message specified in a FAIL return from the queryprogram router;
3099 (e) A message specified by the cannot_route_message router option.
3101 In these cases only, Exim does include the error details in bounce and
3102 warning messages. There are also a few cases where bland messages such
3103 as "unrouteable address" or "local delivery error" are given.
3105 PH/43 $value is now also set for the "else" part of a ${run expansion.
3107 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3108 being worked on, but at least Exim now implements the latest version to
3111 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3112 process to complete was treated as a delivery failure, a timeout while
3113 writing the message to the pipe was logged, but erroneously treated as a
3114 successful delivery. Such timeouts include transport filter timeouts. For
3115 consistency with the overall process timeout, these timeouts are now
3116 treated as errors, giving rise to delivery failures by default. However,
3117 there is now a new Boolean option for the pipe transport called
3118 timeout_defer, which, if set TRUE, converts the failures into defers for
3119 both kinds of timeout. A transport filter timeout is now identified in
3122 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3123 systems where "make" and "gmake" are different, calling "gmake" at top
3124 level broke things. I've arranged for the value of $(MAKE) to be passed
3125 from the Makefile to this script so that it can call the same version of
3129 A note about Exim versions 4.44 and 4.50
3130 ----------------------------------------
3132 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3133 changes of various kinds. As a consequence, a big documentation update was
3134 needed. This delayed the release for rather longer than seemed good, especially
3135 in the light of a couple of (minor) security issues. Therefore, the changes
3136 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3137 release. So 4.44 and 4.50 are in effect two different branches that both start
3140 I have left the 4.50 change log unchanged; it contains all the changes since
3141 4.43. The change log for 4.44 is below; many of its items are identical to
3142 those for 4.50. This seems to be the most sensible way to preserve the
3143 historical information.
3149 1. Minor wording change to the doc/README.SIEVE file.
3151 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3152 computation of the current number of files was incorrect.
3154 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3155 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3158 4. Give more explanation in the error message when the command for a transport
3159 filter fails to execute.
3161 5. There are several places where Exim runs a non-Exim command in a
3162 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3163 was being done only for the command run by the queryprogram router. It is
3164 now done for all such subprocesses. The other cases are: ${run, transport
3165 filters, and the commands run by the lmtp and pipe transports.
3167 6. Added CONFIGURE_GROUP build-time option.
3169 7. Some older OS have a limit of 256 on the maximum number of file
3170 descriptors. Exim was using setrlimit() to set 1000 as a large value
3171 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3172 systems. I've change it so that if it can't get 1000, it tries for 256.
3174 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3175 was an oversight, and furthermore, ever since the addition of extra
3176 controls (e.g. 4.43/32), the checks on when to allow different forms of
3177 "control" were broken. There should now be diagnostics for all cases when a
3178 control that does not make sense is encountered.
3180 9. Added the /retain_sender option to "control=submission".
3182 10. $recipients is now available in the predata ACL (oversight).
3184 11. Tidy the search cache before the fork to do a delivery from a message
3185 received from the command line. Otherwise the child will trigger a lookup
3186 failure and thereby defer the delivery if it tries to use (for example) a
3187 cached ldap connection that the parent has called unbind on.
3189 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3190 of $address_data from the recipient verification was clobbered by the
3191 sender verification.
3193 13. The value of address_data from a sender verification is now available in
3194 $sender_address_data in subsequent conditions in the ACL statement.
3196 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3198 15. Added a new option "connect=<time>" to callout options, to set a different
3201 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3202 was its contents. (It was OK if the option was not defined at all.)
3204 17. A "Completed" log line is now written for messages that are removed from
3205 the spool by the -Mrm option.
3207 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3208 information about exactly what failed.
3210 19. Added -dd to debug only the daemon process.
3212 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3213 handles timeouts, both on the server side and network timeouts. Renamed the
3214 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3216 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3217 It is now set to "smtps".
3219 22. $host_address is now set to the target address during the checking of
3220 ignore_target_hosts.
3222 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3223 being passed; this would have caused $sender_host_name to have been used if
3224 matching the list had actually called for a host name (not very likely,
3225 since this list is usually IP addresses). A host name is now passed as
3228 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3229 code that specifies a non-restarting handler (typically sigaction() in
3230 modern systems) in an attempt to fix a rare and obscure crash bug.
3232 25. Narrowed the window for a race in the daemon that could cause it to ignore
3233 SIGCHLD signals. This is not a major problem, because they are used only to
3234 wake it up if nothing else does.
3236 26. A malformed maildirsize file could cause Exim to calculate negative values
3237 for the mailbox size or file count. Odd effects could occur as a result.
3238 The maildirsize information is now recalculated if the size or filecount
3241 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3242 support for a long time. Removed HAVE_SYS_VFS_H.
3244 28. Installed the latest version of exipick from John Jetmore.
3246 29. In an address list, if the pattern was not a regular expression, an empty
3247 subject address (from a bounce message) matched only if the pattern was an
3248 empty string. Non-empty patterns were not even tested. This was the wrong
3249 because it is perfectly reasonable to use an empty address as part of a
3250 database query. An empty address is now tested by patterns that are
3251 lookups. However, all the other forms of pattern expect the subject to
3252 contain a local part and a domain, and therefore, for them, an empty
3253 address still always fails if the pattern is not itself empty.
3255 30. Exim went into a mad DNS loop when attempting to do a callout where the
3256 host was specified on an smtp transport, and looking it up yielded more
3257 than one IP address.
3259 31. Re-factored the code for checking spool and log partition space into a
3260 function that finds that data and another that does the check. The former
3261 is then used to implement four new variables: $spool_space, $log_space,
3262 $spool_inodes, and $log_inodes.
3264 32. The RFC2047 encoding function was originally intended for short strings
3265 such as real names; it was not keeping to the 75-character limit for
3266 encoded words that the RFC imposes. It now respects the limit, and
3267 generates multiple encoded words if necessary. To be on the safe side, I
3268 have increased the buffer size for the ${rfc2047: expansion operator from
3271 33. It is now permitted to omit both strings after an "if" condition; if the
3272 condition is true, the result is "true". As before, when the second string
3273 is omitted, a false condition yields an empty string. This makes it less
3274 cumbersome to write custom ACL and router conditions.
3276 34. Failure to deliver a bounce message always caused it to be frozen, even if
3277 there was an errors_to setting on the router. The errors_to setting is now
3280 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3281 canonical form (fully expanded) before being placed in
3282 $sender_host_address.
3284 36. The table in the code that translates DNS record types into text (T_A to
3285 "A" for instance) was missing entries for NS and CNAME. It is just possible
3286 that this could have caused confusion if both these types were looked up
3287 for the same domain, because the text type is used as part of Exim's
3288 per-process caching. But the chance of anyone hitting this buglet seems
3291 37. The dnsdb lookup has been extended in a number of ways.
3293 (1) There is a new type, "zns", which walks up the domain tree until it
3294 finds some nameserver records. It should be used with care.
3296 (2) There is a new type, "mxh", which is like "mx" except that it returns
3297 just the host names, not the priorities.
3299 (3) It is now possible to give a list of domains (or IP addresses) to be
3300 looked up. The behaviour when one of the lookups defers can be
3301 controlled by a keyword.
3303 (4) It is now possible to specify the separator character for use when
3304 multiple records are returned.
3306 38. The dnslists ACL condition has been extended: it is now possible to supply
3307 a list of IP addresses and/or domains to be looked up in a particular DNS
3310 39. Added log_selector=+queue_time_overall.
3312 40. When running the queue in the test harness, wait just a tad after forking a
3313 delivery process, to get repeatability of debugging output.
3315 41. Include certificate and key file names in error message when GnuTLS fails
3316 to set them up, because the GnuTLS error message doesn't include the name
3317 of the failing file when there is a problem reading it.
3319 42. Allow both -bf and -bF in the same test run.
3321 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3323 44. The "Exiscan patch" is now merged into the mainline Exim source.
3325 45. Sometimes the final signoff response after QUIT could fail to get
3326 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3327 < 0 before doing a fflush(). This bug looks as though it goes back to the
3328 introduction of TLS in release 3.20, but "sometimes" must have been rare
3329 because the tests only now provoked it.
3331 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3332 (this can affect the format of dates).
3334 47. exim_tidydb, when checking for the continued existence of a message for
3335 which it has found a message-specific retry record, was not finding
3336 messages that were in split spool directories. Consequently, it was
3337 deleting retry records that should have stayed in existence.
3339 48. Steve fixed some bugs in eximstats.
3341 49. The SPA authentication driver was not abandoning authentication and moving
3342 on to the next authenticator when an expansion was forced to fail,
3343 contradicting the general specification for all authenticators. Instead it
3344 was generating a temporary error. It now behaves as specified.
3346 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3347 (the order specifies the preference for clients). The order is now AES256,
3348 AES128, 3DES, ARCFOUR128.
3350 51. Small patch to Sieve code - explicitly set From: when generating an
3353 52. Exim crashed if a remote delivery caused a very long error message to be
3354 recorded - for instance if somebody sent an entire SpamAssassin report back
3355 as a large number of 550 error lines. This bug was coincidentally fixed by
3356 increasing the size of one of Exim's internal buffers (big_buffer) that
3357 happened as part of the Exiscan merge. However, to be on the safe side, I
3358 have made the code more robust (and fixed the comments that describe what
3361 53. Now that there can be additional text after "Completed" in log lines (if
3362 the queue_time_overall log selector is set), a one-byte patch to exigrep
3363 was needed to allow it to recognize "Completed" as not the last thing in
3366 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3367 patch that reportedly fixes this has been added. I am not expert enough to
3368 create a test for it. This is what the patch creator wrote:
3370 "I found a little strange behaviour of ldap code when working with
3371 Windows 2003 AD Domain, where users was placed in more than one
3372 Organization Units. When I tried to give exim partial DN, the exit code
3373 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3374 But simultaneously result of request was absolutely normal ldap result,
3375 so I produce this patch..."
3377 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3378 REFERENCE, so I have modified the code to exclude the patch when that macro
3381 55. Some experimental protocols are using DNS PTR records for new purposes. The
3382 keys for these records are domain names, not reversed IP addresses. The
3383 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3384 leaves it alone. Component reversal etc. now happens only for IP addresses.
3387 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3389 57. Double the size of the debug message buffer (to 2048) so that more of very
3390 long debug lines gets shown.
3392 58. The exicyclog utility now does better if the number of log files to keep
3393 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3395 59. Two changes related to the smtp_active_hostname option:
3397 (1) $smtp_active_hostname is now available as a variable.
3398 (2) The default for smtp_banner uses $smtp_active_hostname instead
3399 of $primary_hostname.
3401 60. The host_aton() function is supposed to be passed a string that is known
3402 to be a valid IP address. However, in the case of IPv6 addresses, it was
3403 not checking this. This is a hostage to fortune. Exim now panics and dies
3404 if the condition is not met. A case was found where this could be provoked
3405 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3406 components; fortuitously, this particular loophole had already been fixed
3407 by change 4.50/55 above.
3409 If there are any other similar loopholes, the new check in host_aton()
3410 itself should stop them being exploited. The report I received stated that
3411 data on the command line could provoke the exploit when Exim was running as
3412 exim, but did not say which command line option was involved. All I could
3413 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3414 running as the user.
3417 61. There was a buffer overflow vulnerability in the SPA authentication code
3418 (which came originally from the Samba project). I have added a test to the
3419 spa_base64_to_bits() function which I hope fixes it.
3422 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3423 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3425 63. The daemon start-up calls getloadavg() while still root for those OS that
3426 need the first call to be done as root, but it missed one case: when
3427 deliver_queue_load_max is set with deliver_drop_privilege. This is
3428 necessary for the benefit of the queue runner, because there is no re-exec
3429 when deliver_drop_privilege is set.
3431 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3432 This has been fixed.
3434 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3435 was in use, was not putting the data itself into the right store pool;
3436 consequently, it could be overwritten for a subsequent message in the same
3437 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3440 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3442 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3443 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3444 the function were treating the return as a boolean value, which happened to
3445 work because 0=false and not-0=true, but is not correct code.
3447 68. The host_aton() function was not handling scoped IPv6 addresses (those
3448 with, for example, "%eth0" on the end) correctly.
3450 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3451 negated items (that is, ~something) in unsigned ints. Some compilers
3452 apparently mutter when there is no cast.
3454 70. If an address verification called from an ACL failed, and did not produce a
3455 user-specific message (i.e. there was only a "system" message), nothing was
3456 put in $acl_verify_message. In this situation, it now puts the system
3459 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3460 session; change 4.31/43 added the unwanted input to the log line - except
3461 that it did not do this in the start of session case. It now does.
3463 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3464 This is wrong and can cause the other end to generate a synchronization
3465 error if it is another Exim or anything else that does the synchronization
3466 check. A QUIT command is no longer sent after a timeout.
3468 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3469 during host lookups.
3471 74. The defer_ok option of callout verification was not working if it was used
3472 when verifying addresses in header lines, that is, for this case:
3474 verify = header_sender/callout=defer_ok
3476 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3477 those file descriptors could be used for SMTP connections. If anything
3478 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3479 could be sent to the SMTP client, causing chaos. The daemon now opens
3480 stdin, stdout, and stderr to /dev/null when it puts itself into the
3483 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3484 log by default. The user can override this with suitable Perl magic.
3486 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3487 the log message when discard triggers, was not working for the DATA ACL or
3488 for the non-SMTP ACL.
3490 78. Error message wording change in sieve.c.
3492 79. If smtp_accept_max_per_host was set, the number of connections could be
3493 restricted to fewer than expected, because the daemon was trying to set up
3494 a new connection before checking whether the processes handling previous
3495 connections had finished. The check for completed processes is now done
3496 earlier. On busy systems, this bug wouldn't be noticed because something
3497 else would have woken the daemon, and it would have reaped the completed
3500 80. If a message was submitted locally by a user whose login name contained one
3501 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3502 It caused a spool format error. I have fixed the spool reading code. A
3503 related problem was that the "from" clause in the Received: line became
3504 illegal because of the space(s). It is now covered by ${quote_local_part.
3506 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3509 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3511 83. Patch from Sieve maintainer to fix unterminated string problem in
3512 "vacation" handling.
3514 84. Some minor changes to the Linux configuration files to help with other
3515 OS variants using glibc.
3517 85. One more patch for Sieve to update vacation handling to latest spec.
3520 ----------------------------------------------------
3521 See the note above about the 4.44 and 4.50 releases.
3522 ----------------------------------------------------
3528 1. Change 4.43/35 introduced a bug that caused file counts to be
3529 incorrectly computed when quota_filecount was set in an appendfile
3532 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3533 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3536 3. Give more explanation in the error message when the command for a transport
3537 filter fails to execute.
3539 4. There are several places where Exim runs a non-Exim command in a
3540 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3541 was being done only for the command run by the queryprogram router. It is
3542 now done for all such subprocesses. The other cases are: ${run, transport
3543 filters, and the commands run by the lmtp and pipe transports.
3545 5. Some older OS have a limit of 256 on the maximum number of file
3546 descriptors. Exim was using setrlimit() to set 1000 as a large value
3547 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3548 systems. I've change it so that if it can't get 1000, it tries for 256.
3550 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3551 was an oversight, and furthermore, ever since the addition of extra
3552 controls (e.g. 4.43/32), the checks on when to allow different forms of
3553 "control" were broken. There should now be diagnostics for all cases when a
3554 control that does not make sense is encountered.
3556 7. $recipients is now available in the predata ACL (oversight).
3558 8. Tidy the search cache before the fork to do a delivery from a message
3559 received from the command line. Otherwise the child will trigger a lookup
3560 failure and thereby defer the delivery if it tries to use (for example) a
3561 cached ldap connection that the parent has called unbind on.
3563 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3564 of $address_data from the recipient verification was clobbered by the
3565 sender verification.
3567 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3568 was its contents. (It was OK if the option was not defined at all.)
3570 11. A "Completed" log line is now written for messages that are removed from
3571 the spool by the -Mrm option.
3573 12. $host_address is now set to the target address during the checking of
3574 ignore_target_hosts.
3576 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3577 being passed; this would have caused $sender_host_name to have been used if
3578 matching the list had actually called for a host name (not very likely,
3579 since this list is usually IP addresses). A host name is now passed as
3582 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3583 code that specifies a non-restarting handler (typically sigaction() in
3584 modern systems) in an attempt to fix a rare and obscure crash bug.
3586 15. Narrowed the window for a race in the daemon that could cause it to ignore
3587 SIGCHLD signals. This is not a major problem, because they are used only to
3588 wake it up if nothing else does.
3590 16. A malformed maildirsize file could cause Exim to calculate negative values
3591 for the mailbox size or file count. Odd effects could occur as a result.
3592 The maildirsize information is now recalculated if the size or filecount
3595 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3596 support for a long time. Removed HAVE_SYS_VFS_H.
3598 18. Updated exipick to current release from John Jetmore.
3600 19. Allow an empty sender to be matched against a lookup in an address list.
3601 Previously the only cases considered were a regular expression, or an
3604 20. Exim went into a mad DNS lookup loop when doing a callout where the
3605 host was specified on the transport, if the DNS lookup yielded more than
3608 21. The RFC2047 encoding function was originally intended for short strings
3609 such as real names; it was not keeping to the 75-character limit for
3610 encoded words that the RFC imposes. It now respects the limit, and
3611 generates multiple encoded words if necessary. To be on the safe side, I
3612 have increased the buffer size for the ${rfc2047: expansion operator from
3615 22. Failure to deliver a bounce message always caused it to be frozen, even if
3616 there was an errors_to setting on the router. The errors_to setting is now
3619 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3620 canonical form (fully expanded) before being placed in
3621 $sender_host_address.
3623 24. Updated eximstats to version 1.33
3625 25. Include certificate and key file names in error message when GnuTLS fails
3626 to set them up, because the GnuTLS error message doesn't include the name
3627 of the failing file when there is a problem reading it.
3629 26. Expand error message when OpenSSL has problems setting up cert/key files.
3632 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3633 (this can affect the format of dates).
3635 28. exim_tidydb, when checking for the continued existence of a message for
3636 which it has found a message-specific retry record, was not finding
3637 messages that were in split spool directories. Consequently, it was
3638 deleting retry records that should have stayed in existence.
3640 29. eximstats updated to version 1.35
3641 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3642 1.35 - bugfix such that pie charts by volume are generated correctly
3644 30. The SPA authentication driver was not abandoning authentication and moving
3645 on to the next authenticator when an expansion was forced to fail,
3646 contradicting the general specification for all authenticators. Instead it
3647 was generating a temporary error. It now behaves as specified.
3649 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3650 (the order specifies the preference for clients). The order is now AES256,
3651 AES128, 3DES, ARCFOUR128.
3653 31. Small patch to Sieve code - explicitly set From: when generating an
3656 32. Exim crashed if a remote delivery caused a very long error message to be
3657 recorded - for instance if somebody sent an entire SpamAssassin report back
3658 as a large number of 550 error lines. This bug was coincidentally fixed by
3659 increasing the size of one of Exim's internal buffers (big_buffer) that
3660 happened as part of the Exiscan merge. However, to be on the safe side, I
3661 have made the code more robust (and fixed the comments that describe what
3664 33. Some experimental protocols are using DNS PTR records for new purposes. The
3665 keys for these records are domain names, not reversed IP addresses. The
3666 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3667 leaves it alone. Component reversal etc. now happens only for IP addresses.
3670 34. The host_aton() function is supposed to be passed a string that is known
3671 to be a valid IP address. However, in the case of IPv6 addresses, it was
3672 not checking this. This is a hostage to fortune. Exim now panics and dies
3673 if the condition is not met. A case was found where this could be provoked
3674 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3675 components; fortuitously, this particular loophole had already been fixed
3676 by change 4.50/55 or 4.44/33 above.
3678 If there are any other similar loopholes, the new check in host_aton()
3679 itself should stop them being exploited. The report I received stated that
3680 data on the command line could provoke the exploit when Exim was running as
3681 exim, but did not say which command line option was involved. All I could
3682 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3683 running as the user.
3686 35. There was a buffer overflow vulnerability in the SPA authentication code
3687 (which came originally from the Samba project). I have added a test to the
3688 spa_base64_to_bits() function which I hope fixes it.
3691 36. The daemon start-up calls getloadavg() while still root for those OS that
3692 need the first call to be done as root, but it missed one case: when
3693 deliver_queue_load_max is set with deliver_drop_privilege. This is
3694 necessary for the benefit of the queue runner, because there is no re-exec
3695 when deliver_drop_privilege is set.
3697 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3698 was in use, was not putting the data itself into the right store pool;
3699 consequently, it could be overwritten for a subsequent message in the same
3700 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3703 38. Sometimes the final signoff response after QUIT could fail to get
3704 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3705 < 0 before doing a fflush(). This bug looks as though it goes back to the
3706 introduction of TLS in release 3.20, but "sometimes" must have been rare
3707 because the tests only now provoked it.
3713 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3714 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3715 Now it returns an integer. A number of calls were still expecting a T/F
3716 return. Fortuitously, in all cases, the tests worked in OK situations,
3717 which is the norm. However, things would have gone wrong on any write
3718 failures on the smtp file descriptor. This function is used when sending
3719 messages over SMTP and also when doing verify callouts.
3721 2. When Exim is called to do synchronous delivery of a locally submitted
3722 message (the -odf or -odi options), it no longer closes stderr before doing
3725 3. Implemented the mua_wrapper option.
3727 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3729 5. Implemented the functions header_remove(), header_testname(),
3730 header_add_at_position(), and receive_remove_recipient(), and exported them
3733 6. If an ACL "warn" statement specified the addition of headers, Exim already
3734 inserted X-ACL-Warn: at the start if there was no header name. However, it
3735 was not making this test for the second and subsequent header lines if
3736 there were newlines in the string. This meant that an invalid header could
3737 be inserted if Exim was badly configured.
3739 7. Allow an ACL "warn" statement to add header lines at the start or after all
3740 the Received: headers, as well as at the end.
3742 8. Added the rcpt_4xx retry error code.
3744 9. Added postmaster_mailfrom=xxx to callout verification option.
3746 10. Added mailfrom=xxxx to the callout verification option, for verify=
3749 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3750 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3752 12. Inserted some casts to stop certain compilers warning when using pointer
3753 differences as field lengths or precisions in printf-type calls (mostly
3754 affecting debugging statements).
3756 13. Added optional readline() support for -be (dynamically loaded).
3758 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3759 same clock tick as a message's arrival, so that its received time was the
3760 same as the "first fail" time on the retry record, and that message
3761 remained on the queue past the ultimate address timeout, every queue runner
3762 would try a delivery (because it was past the ultimate address timeout) but
3763 after another failure, the ultimate address timeout, which should have then
3764 bounced the address, did not kick in. This was a "< instead of <=" error;
3765 in most cases the first failure would have been in the next clock tick
3766 after the received time, and all would be well.
3768 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3769 being recognized when the domain list was tested by the match_domain
3770 condition in an expansion string.
3772 16. Added the ${str2b64: operator.
3774 17. Exim was always calling setrlimit() to set a large limit for the number of
3775 processes, without checking whether the existing limit was already
3776 adequate. (It did check for the limit on file descriptors.) Furthermore,
3777 errors from getrlimit() and setrlimit() were being ignored. Now they are
3778 logged to the main and panic logs, but Exim does carry on, to try to do its
3779 job under whatever limits there are.
3781 18. Imported PCRE 5.0.
3783 19. Trivial typo in log message " temporarily refused connection" (the leading
3786 20. If the log selector return_path_on_delivery was set and an address was
3787 redirected to /dev/null, the delivery process crashed because it assumed
3788 that a return path would always be set for a "successful" delivery. In this
3789 case, the whole delivery is bypassed as an optimization, and therefore no
3792 21. Internal re-arrangement: the function for sending a challenge and reading
3793 a response while authentication was assuming a zero-terminated challenge
3794 string. It's now changed to take a pointer and a length, to allow for
3795 binary data in such strings.
3797 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3799 23. Exim was not respecting finduser_retries when seeking the login of the
3800 uid under which it was called; it was always trying 10 times. (The default
3801 setting of finduser_retries is zero.) Also, it was sleeping after the final
3802 failure, which is pointless.
3804 24. Implemented tls_on_connect_ports.
3806 25. Implemented acl_smtp_predata.
3808 26. If the domain in control=submission is set empty, Exim assumes that the
3809 authenticated id is a complete email address when it generates From: or
3810 Sender: header lines.
3812 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3813 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3814 chown and chgrp in /bin and hostname in /usr/bin.
3816 28. Exim was keeping the "process log" file open after each use, just as it
3817 does for the main log. This opens the possibility of it remaining open for
3818 long periods when the USR1 signal hits a daemon. Occasional processlog
3819 errors were reported, that could have been caused by this. Anyway, it seems
3820 much more sensible not to leave this file open at all, so that is what now
3823 29. The long-running daemon process does not normally write to the log once it
3824 has entered its main loop, and it closes the log before doing so. This is
3825 so that log files can straightforwardly be renamed and moved. However,
3826 there are a couple of unusual error situations where the daemon does write
3827 log entries, and I had neglected to close the log afterwards.
3829 30. The text of an SMTP error response that was received during a remote
3830 delivery was being truncated at 512 bytes. This is too short for some of
3831 the long messages that one sometimes sees. I've increased the limit to
3834 31. It is now possible to make retry rules that apply only when a message has a
3835 specific sender, in particular, an empty sender.
3837 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3838 it possible to be selective about when SMTP synchronization is enforced.
3840 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3842 32. Implemented hosts_connection_nolog.
3844 33. Added an ACL for QUIT.
3846 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3849 35. Added mailbox_size and mailbox_filecount to appendfile.
3851 36. Added control = no_multiline_responses to ACLs.
3853 37. There was a bug in the logic of the code that waits for the clock to tick
3854 in the case where the clock went backwards by a substantial amount such
3855 that the microsecond fraction of "now" was more than the microsecond
3856 fraction of "then" (but the whole seconds number was less).
3858 38. Added support for the libradius Radius client library this is found on
3859 FreeBSD (previously only the radiusclient library was supported).
3865 1. When certain lookups returned multiple values in the form name=value, the
3866 quoting of the values was not always being done properly. Specifically:
3867 (a) If the value started with a double quote, but contained no whitespace,
3869 (b) If the value contained whitespace other than a space character (i.e.
3870 tabs or newlines or carriage returns) it was not quoted.
3871 This fix has been applied to the mysql and pgsql lookups by writing a
3872 separate quoting function and calling it from the lookup code. The fix
3873 should probably also be applied to nisplus, ibase and oracle lookups, but
3874 since I cannot test any of those, I have not disturbed their existing code.
3876 2. A hit in the callout cache for a specific address caused a log line with no
3877 reason for rejecting RCPT. Now it says "Previous (cached) callout
3878 verification failure".
3880 3. There was an off-by-one bug in the queryprogram router. An over-long
3881 return line was truncated at 256 instead of 255 characters, thereby
3882 overflowing its buffer with the terminating zero. As well as fixing this, I
3883 have increased the buffer size to 1024 (and made a note to document this).
3885 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3886 when Exim is waiting for an SMTP response from a remote server, Exim
3887 restarts its select() call on the socket, thereby resetting its timeout.
3888 This is not a problem when such interrupts are rare. Somebody set up a cron
3889 job to run exiwhat every 2 minutes, which is less than the normal select()
3890 timeout (5 or 10 minutes). This meant that the select() timeout never
3891 kicked in because it was always reset. I have fixed this by comparing the
3892 time when an interrupt arrives with the time at the start of the first call
3893 to select(). If more time than the timeout has elapsed, the interrupt is
3894 treated as a timeout.
3896 5. Some internal re-factoring in preparation for the addition of Sieve
3897 extensions (by MH). In particular, the "personal" test is moved to a
3898 separate function, and given an option for scanning Cc: and Bcc: (which is
3899 not set for Exim filters).
3901 6. When Exim created an email address using the login of the caller as the
3902 local part (e.g. when creating a From: or Sender: header line), it was not
3903 quoting the local part when it contained special characters such as @.
3905 7. Installed new OpenBSD configuration files.
3907 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3908 try to make them clearer.
3910 9. Callout options, other than the timeout value, were being ignored when
3911 verifying sender addresses in header lines. For example, when using
3913 verify = header_sender/callout=no_cache
3915 the cache was (incorrectly) being used.
3917 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3918 only the Cygwin environment.
3920 11. When return_path_on_delivery was set as a log selector, if different remote
3921 addresses in the same message used different return paths and parallel
3922 remote delivery occurred, the wrong values would sometimes be logged.
3923 (Whenever a remote delivery process finished, the return path value from
3924 the most recently started remote delivery process was logged.)
3926 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3927 lines when AUTH and/or TLS are in use. This is the "received protocol"
3928 field. Exim used to use "asmtp" for authenticated SMTP, without any
3929 indication (in the protocol name) for TLS use. Now it follows the RFC and
3930 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3931 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3932 names appear in log lines as well as in Received: header lines.
3934 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3935 extensions, and comparison tests, and to fix some bugs.
3937 14. Changes to the "personal" filter test:
3939 (1) The test was buggy in that it was just doing the equivalent of
3940 "contains" tests on header lines. For example, if a user's address was
3941 anne@some.where, the "personal" test would incorrectly be true for
3943 To: susanne@some.where
3945 This test is now done by extracting each address from the header in turn,
3946 and checking the entire address. Other tests that are part of "personal"
3947 are now done using regular expressions (for example, to check local parts
3948 of addresses in From: header lines).
3950 (2) The list of non-personal local parts in From: addresses has been
3951 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3952 taken from the Sieve specification recommendations.
3954 (3) If the message contains any header line starting with "List-" it is
3955 treated as non-personal.
3957 (4) The test for "circular" in the Subject: header line has been removed
3958 because it now seems ill-conceived.
3960 15. Minor typos in src/EDITME comments corrected.
3962 16. Installed latest exipick from John Jetmore.
3964 17. If headers_add on a router specified a text string that was too long for
3965 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3966 of string_sprintf() is now avoided.
3968 18. $message_body_size was not set (it was always zero) when running the DATA
3969 ACL and the local_scan() function.
3971 19. For the "mail" command in an Exim filter, no default was being set for
3972 the once_repeat time, causing a random time value to be used if "once" was
3973 specified. (If the value happened to be <= 0, no repeat happened.) The
3974 default is now 0s, meaning "never repeat". The "vacation" command was OK
3975 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3976 (I found it when inspecting the code).
3978 20. There is now an overall timeout for performing a callout verification. It
3979 defaults to 4 times the callout timeout, which applies to individual SMTP
3980 commands during the callout. The overall timeout applies when there is more
3981 than one host that can be tried. The timeout is checked before trying the
3982 next host. This prevents very long delays if there are a large number of
3983 hosts and all are timing out (e.g. when the network connections are timing
3984 out). The value of the overall timeout can be changed by specifying an
3985 additional sub-option for "callout", called "maxwait". For example:
3987 verify = sender/callout=5s,maxwait=20s
3989 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3990 to the end before writing, but this should make it even safer).
3992 22. Exim was forgetting that it had advertised PIPELINING for the second and
3993 subsequent messages on an SMTP connection. It was also not resetting its
3994 memory on STARTTLS and an internal HELO.
3996 23. When Exim logs an SMTP synchronization error within a session, it now
3997 records whether PIPELINING has been advertised or not.
3999 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4000 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4001 rather than long int.
4003 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4005 26. Added the never_mail option to autoreply.
4011 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4012 crash if the getsockname() call failed; this can happen if a connection is
4013 closed very soon after it is established. The problem was simply in the
4014 order in which certain operations were done, causing Exim to try to write
4015 to the SMTP stream before it had set up the file descriptor. The bug has
4016 been fixed by making things happen in the correct order.
4022 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4023 before the connection was closed, thus losing the rejection response.
4025 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4026 some early Solaris releases, but causes trouble in current releases where
4027 socklen_t is defined.
4029 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4032 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4035 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4036 that releasing the top of it at the end releases what was used for sub-
4037 expansions (unless the block got too big). However, discard this block if
4038 the first thing is a variable or header, so that we can use its block when
4039 it is dynamic (useful for very large $message_headers, for example).
4041 6. Lookups now cache *every* query, not just the most recent. A new, separate
4042 store pool is used for this. It can be recovered when all lookup caches are
4043 flushed. Lookups now release memory at the end of their result strings.
4044 This has involved some general refactoring of the lookup sources.
4046 7. Some code has been added to the store_xxx() functions to reduce the amount
4047 of flapping under certain conditions.
4049 8. log_incoming_interface used to affect only the <= reception log lines. Now
4050 it causes the local interface and port to be added to several more SMTP log
4051 lines, for example "SMTP connection from", and rejection lines.
4053 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4055 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4057 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4058 could be overwritten at the end of the current message (or the start of a
4059 new message if it was set in a HELO ACL). The value is now preserved for
4060 the duration of the SMTP connection.
4062 12. If a transport had a headers_rewrite setting, and a matching header line
4063 contained an unqualified address, that address was qualified, even if it
4064 did not match any rewriting rules. The underlying bug was that the values
4065 of the flags that permit the existence of unqualified sender and recipient
4066 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4067 non-local messages, and by -bnq for local messages) were not being
4068 preserved with the message after it was received.
4070 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4071 "next input=" as part of the text comprising the host identity instead of
4072 the correct text. The code was using the same buffer for two different
4073 strings. However, depending on which order the printing function evaluated
4074 its arguments, the bug did not always show up. Under Linux, for example, my
4075 test suite worked just fine.
4077 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4078 doesn't work with some older versions of Perl. It has been changed to "my",
4079 which in any case is probably the better facility to use.
4081 15. A really picky compiler found some instances of statements for creating
4082 error messages that either had too many or two few arguments for the format
4085 16. The size of the buffer for calls to the DNS resolver has been increased
4086 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4087 for addresses that have a lot of PTR records. This alleviates a problem; it
4088 does not fully solve it.
4090 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4091 buffer now truncates the list and logs the incident, which is the same
4092 action as happens when Exim is looking up a host name and its aliases.
4093 Previously in this situation something unpredictable would happen;
4094 sometimes it was "internal error: store_reset failed".
4096 18. If a server dropped the connection unexpectedly when an Exim client was
4097 using GnuTLS and trying to read a response, the client delivery process
4098 crashed while trying to generate an error log message.
4100 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4101 string, for example:
4103 warn message = H1: something\nH2: something
4105 the text was added as a single header line from Exim's point of view
4106 though it ended up OK in the delivered message. However, searching for the
4107 second and subsequent header lines using $h_h2: did not work. This has been
4108 fixed. Similarly, if a system filter added multiple headers in this way,
4109 the routers could not see them.
4111 20. Expanded the error message when iplsearch is called with an invalid key to
4112 suggest using net-iplsearch in a host list.
4114 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4115 ACLs are no longer actually imposed (and a message to that effect is
4118 22. If a "gecos" field in a passwd entry contained escaped characters, in
4119 particular, if it contained a \" sequence, Exim got it wrong when building
4120 a From: or a Sender: header from that name. A second bug also caused
4121 incorrect handling when an unquoted " was present following a character
4122 that needed quoting.
4124 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4125 was not being matched caselessly.
4127 24. Arranged for all hyphens in the exim.8 source to be escaped with
4130 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4131 properly. Recipient callout cache records were still being keyed to include
4132 the sender, even when use_sender was set false. This led to far more
4133 callouts that were necessary. The sender is no longer included in the key
4134 when use_sender is false.
4136 26. Added "control = submission" modifier to ACLs.
4138 27. Added the ${base62d: operator to decode base 62 numbers.
4140 28. dnsdb lookups can now access SRV records.
4142 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4143 the configuration file.
4145 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4146 (-v) mode. This makes the output for a verbose queue run more intelligible.
4148 31. Added a use_postmaster feature to recipient callouts.
4150 32. Added the $body_zerocount variable, containing the number of binary zero
4151 bytes in the message body.
4153 33. The time of last modification of the "new" subdirectory is now used as the
4154 "mailbox time last read" when there is a quota error for a maildir
4157 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4159 35. Added +ignore_unknown as a special item in host lists.
4161 36. Code for decoding IPv6 addresses in host lists is now included, even if
4162 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4163 address was recognized as an IP address, but was then not correctly decoded
4164 into binary, causing unexpected and incorrect effects when compared with
4171 1. Very minor rewording of debugging text in manualroute to say "list of
4172 hosts" instead of "hostlist".
4174 2. If verify=header_syntax was set, and a header line with an unqualified
4175 address (no domain) and a large number of spaces between the end of the
4176 name and the colon was received, the reception process suffered a buffer
4177 overflow, and (when I tested it) crashed. This was caused by some obsolete
4178 code that should have been removed. The fix is to remove it!
4180 3. When running in the test harness, delay a bit after writing a bounce
4181 message to get a bit more predictability in the log output.
4183 4. Added a call to search_tidyup() just before forking a reception process. In
4184 theory, someone could use a lookup in the expansion of smtp_accept_max_
4185 per_host which, without the tidyup, could leave open a database connection.
4187 5. Added the variables $recipient_data and $sender_data which get set from a
4188 lookup success in an ACL "recipients" or "senders" condition, or a router
4189 "senders" option, similar to $domain_data and $local_part_data.
4191 6. Moved the writing of debug_print from before to after the "senders" test
4194 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4195 problems for message scanning, either using a data ACL, or using
4196 local_scan() because the Received: header was not generated till after they
4197 were called (in order to set the time as the time of reception completion).
4198 I have revised the way this works. The header is now generated after the
4199 body is received, but before the ACL or local_scan() are called. After they
4200 are run, the timestamp in the header is updated.
4206 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4207 before starting a queue runner without re-exec. This happened only when
4208 deliver_drop_privilege was set or when the Exim user was set to root. The
4209 effect of the bug was that timeouts during subsequent deliveries caused
4210 crashes instead of being properly handled. The handler is now left at its
4211 default (and expected) setting.
4213 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4214 message, again when deliver_drop_privilege is set or Exim is run as root.
4215 The bug described in (1) was not present in this case, but the tidying up
4216 of the other signals was missing. I have made the two cases consistent.
4218 3. The ignore_target_hosts setting on a manualroute router was being ignored
4219 for hosts that were looked up using the /MX notation.
4221 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4224 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4225 operated on the sender address. After changing the $sender_address to <>
4226 for the sender address verify, Exim was re-instated it as the original
4227 (before rewriting) address, but remembering that it had rewritten it, so it
4228 wasn't rewriting it again. This bug also had the effect of breaking the
4229 sender address verification caching when the sender address was rewritten.
4231 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4232 This has been changed so that if the ip literal address matches
4233 ignore_target_hosts, the router declines.
4235 7. Added expansion conditions match_domain, match_address, and match_local_
4236 part (NOT match_host).
4238 8. The placeholder for the Received: header didn't have a length field set.
4240 9. Added code to Exim itself and to exim_lock to test for a specific race
4241 condition that could lead to file corruption when using MBX delivery. The
4242 issue is with the lockfile that is created in /tmp. If this file is removed
4243 after a process has opened it but before that process has acquired a lock,
4244 there is the potential for a second process to recreate the file and also
4245 acquire a lock. This could lead to two Exim processes writing to the file
4246 at the same time. The added code performs the same test as UW imapd; it
4247 checks after acquiring the lock that its file descriptor still refers to
4248 the same named file.
4250 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4251 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4254 11. Added the smtp_active_hostname option. If used, this will typically be made
4255 to depend on the incoming interface address. Because $interface_address is
4256 not set up until the daemon has forked a reception process, error responses
4257 that can happen earlier (such as "too many connections") no longer contain
4260 12. If an expansion in a condition on a "warn" statement fails because a lookup
4261 defers, the "warn" statement is abandoned, and the next ACL statement is
4262 processed. Previously this caused the whole ACL to be aborted.
4264 13. Added the iplsearch lookup type.
4266 14. Added ident_timeout as a log selector.
4268 15. Added tls_certificate_verified as a log selector.
4270 16. Added a global option tls_require_ciphers (compare the smtp transport
4271 option of the same name). This controls incoming TLS connections.
4273 17. I finally figured out how to make tls_require_ciphers do a similar thing
4274 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4275 before starting the TLS session.
4277 18. Tabs are now shown as \t in -bP output.
4279 19. If the log selector return_path_on_delivery was set, Exim crashed when
4280 bouncing a message because it had too many Received: header lines.
4282 20. If two routers both had headers_remove settings, and the first one included
4283 a superfluous trailing colon, the final name in the first list and the
4284 first name in the second list were incorrectly joined into one item (with a
4285 colon in the middle).
4291 1. Added -C and -D options to the exinext utility, mainly to make it easier
4292 to include in the automated testing, but these could be helpful when
4293 multiple configurations are in use.
4295 2. The exinext utility was not formatting the output nicely when there was
4296 an alternate port involved in the retry record key, nor when there was a
4297 message id as well (for retries that were specific to a specific message
4298 and a specific host). It was also confused by IPv6 addresses, because of
4299 the additional colons they contain. I have fixed the IPv4 problem, and
4300 patched it up to do a reasonable job for IPv6.
4302 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4303 delivery, the log line now contains "pipelined" if PIPELINING was used.
4305 4. An SMTP transport process used to panic and die if the bind() call to set
4306 an explicit outgoing interface failed. This has been changed; it is now
4307 treated in the same way as a connect() failure.
4309 5. A reference to $sender_host_name in the part of a conditional expansion
4310 that was being skipped was still causing a DNS lookup. This no longer
4313 6. The def: expansion condition was not recognizing references to header lines
4314 that used bh_ and bheader_.
4316 7. Added the _cache feature to named lists.
4318 8. The code for checking quota_filecount in the appendfile transport was
4319 allowing one more file than it should have been.
4321 9. For compatibility with Sendmail, the command line option
4329 and sets the incoming protocol and host name (for trusted callers). The
4330 host name and its colon can be omitted when only the protocol is to be set.
4331 Note the Exim already has two private options, -pd and -ps, that refer to
4332 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4333 "s", but I don't think that's a major issue.
4335 10. A number of refactoring changes to the code, none of which should affect
4338 (a) The number of logging options was getting close to filling up the
4339 32-bit word that was used as a bit map. I have split them into two classes:
4340 those that are passed in the argument to log_write(), and those that are
4341 only ever tested independently outside of that function. These are now in
4342 separate 32-bit words, so there is plenty of room for expansion again.
4343 There is no change in the user interface or the logging behaviour.
4345 (b) When building, for example, log lines, the code previously used a
4346 macro that called string_cat() twice, in order to add two strings. This is
4347 not really sufficiently general. Furthermore, there was one instance where
4348 it was actually wrong because one of the argument was used twice, and in
4349 one call a function was used. (As it happened, calling the function twice
4350 did not affect the overall behaviour.) The macro has been replaced by a
4351 function that can join an arbitrary number of extra strings onto a growing
4354 (c) The code for expansion conditions now uses a table and a binary chop
4355 instead of a serial search (which was left over from when there were very
4356 few conditions). Also, it now recognizes conditions like "pam" even when
4357 the relevant support is not compiled in: a suitably worded error message is
4358 given if an attempt is made to use such a condition.
4360 11. Added ${time_interval:xxxxx}.
4362 12. A bug was causing one of the ddress fields not to be passed back correctly
4363 from remote delivery subprocesses. The field in question was not being
4364 subsequently used, so this caused to problems in practice.
4366 13. Added new log selectors queue_time and deliver_time.
4368 14. Might have fixed a bug in maildirsizefile handling that threw up
4369 "unexpected character" debug warnings, and recalculated the data
4370 unnecessarily. In any case, I expanded the warning message to give more
4373 15. Added the message "Restricted characters in address" to the statements in
4374 the default ACL that block characters like @ and % in local parts.
4376 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4377 Three changes have been made:
4379 (a) There was a serious bug; a negative response to MAIL caused the whole
4380 recipient domain to be cached as invalid, thereby blocking all messages
4381 to all local parts at the same domain, from all senders. This bug has
4382 been fixed. The domain is no longer cached after a negative response to
4383 MAIL if the sender used is not empty.
4385 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4388 (c) A new callout option, "use_sender" has been added for people who want
4389 the modified behaviour.
4395 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4398 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4399 indeed breaks things for older releases.
4401 3. Added additional logging to the case where there is a problem reading data
4402 from a filter that is running in a subprocess using a pipe, in order to
4403 try to track down a specific problem.
4405 4. Testing facility fudge: when running in the test harness and attempting
4406 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4407 getting "No route to host". Convert this to a timeout.
4409 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4412 6. Some OS don't have socklen_t but use size_t instead. This affects the
4413 fifth argument of getsockopt() amongst other things. This is now
4414 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4415 can be set for individual OS. I have set it for SunOS5, OSF1, and
4416 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4417 some earlier ones do not.
4419 7. Change 4.30/15 was not doing the test caselessly.
4421 8. The standard form for an IPv6 address literal was being rejected by address
4422 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4423 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4424 this, as well as the form without the "IPv6" on the front (but only when
4425 address literals are enabled, of course).
4427 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4429 10. Exim crashed if a message with an empty sender address specified by -f
4430 encountered a router with an errors_to setting. This could be provoked only
4431 by a command such as
4435 where an empty string was supplied; "<>" did not hit this bug.
4437 11. Installed PCRE release 4.5.
4439 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4440 remained set. It is now erased.
4442 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4443 times from message ids (which are base 36 rather than the normal 62).
4445 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4446 were being counted as actual protocol errors, and logged if the log
4447 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4448 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4449 and DATA following a set of rejected RCPTs do not count as protocol errors.
4450 In other words, Exim assumes they were pipelined, though this may not
4451 actually be the case. Of course, in all cases the client gets an
4452 appropriate error code.
4454 15. If a lookup fails in an ACL condition, a message about the failure may
4455 be available; it is used if testing the ACL cannot continue, because most
4456 such messages specify what the cause of the deferral is. However, some
4457 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4458 that caused an old message to be retained and used if a later statement
4459 caused a defer, replacing the real cause of the deferral.
4461 16. If an IP address had so many PTR records that the DNS lookup buffer
4462 was not large enough to hold them, Exim could crash while trying to process
4463 the truncated data. It now detects and logs this case.
4465 17. Further to 4.21/58, another change has been made: if (and only if) the
4466 first line of a message (the first header line) ends with CRLF, a bare LF
4467 in a subsequent header line has a space inserted after it, so as not to
4468 terminate the header.
4470 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4471 unnecessarily, used atoi() instead of strtol(), and didn't check the
4472 termination when getting file sizes from file names by regex.
4474 19. Completely re-implemented the support for maildirsize files, in the light
4475 of a number of problems with the previous contributed implementation
4476 (4.30/29). In particular:
4478 . If the quota is zero, the maildirsize file is maintained, but no quota is
4481 . If the maildir directory does not exist, it is created before any attempt
4482 to write a maildirsize file.
4484 . The quota value in the file is just a cache; if the quota is changed in
4485 the transport, the new value overrides.
4487 . A regular expression is available for excluding directories from the
4490 20. The autoreply transport checks the characters in options that define the
4491 message's headers; it allows continued headers, but it was checking with
4492 isspace() after an embedded newline instead of explicitly looking for a
4495 21. If all the "regular" hosts to which an address was routed had passed their
4496 expiry times, and had not reached their retry times, the address was
4497 bounced, even if fallback hosts were defined. Now Exim should go on to try
4500 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4501 equivalent code in the SMTP transport. Some hosts send humungous responses
4502 to HELO/EHLO, more than 1024 it seems.
4504 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4505 gives compiler warnings in some environments. I've now done it "properly",
4508 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4509 (because of problems with the built-in one) was declared to return uschar *
4510 instead of char *, causing compiler failure.
4512 25. Fixed a file descriptor leak when processing alias/forward files.
4514 26. Fixed a minor format string issue in dbfn.c.
4516 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4518 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4519 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4522 29. When a sender address is verified, it is cached, to save repeating the test
4523 when there is more than one recipient in a message. However, when the
4524 verification involves a callout, it is possible for different callout
4525 options to be set for different recipients. It is too complicated to keep
4526 track of this in the cache, so now Exim always runs a verification when a
4527 callout is required, relying on the callout cache for the optimization.
4528 The overhead is duplication of the address routing, but this should not be
4531 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4532 to be verified with callout=postmaster, and the main callout worked but the
4533 postmaster check failed, the verification correctly failed. However, if a
4534 subsequent RCPT command asked for sender verification *without* the
4535 postmaster check, incorrect caching caused this verification also to fail,
4538 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4539 it was not caching the DNS options (qualify_single, search_parents) that
4540 were used when the lookup failed. A subsequent lookup with different
4541 options therefore always gave the same answer, though there were cases
4542 where it should not have. (Example: a "domains = !$mx_any" option on a
4543 dnslookup router: the "domains" option is always processed without any
4544 widening, but the router might have qualify_single set.) Now Exim uses the
4545 cached value only when the same options are set.
4547 32. Added John Jetmore's "exipick" utility to the distribution.
4549 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4550 than a timeout (e.g. a certificate is required, and is not provided), an
4551 Exim server now closes the connection immediately. Previously it waited for
4552 the client to close - but if the client is SSL, it seems that they each
4553 wait for each other, leading to a delay before one of them times out.
4555 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4556 maintained 0.8.x compatibility because I don't think many are using it, and
4557 it is clearly obsolete.
4559 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4562 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4563 client certificate was expired. A simple patch fixes this, though I don't
4564 understand the full logic of why the verify callback is called multiple
4567 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4568 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4569 which causes problems with some clients (such as the Certicom SSL Plus
4570 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4571 disables the coutermeasure allowing Eudora to connect."
4573 38. Exim was not checking that a write() to a log file succeeded. This could
4574 lead to Bad Things if a log got too big, in particular if it hit a file
4575 size limit. Exim now panics and dies if it cannot write to a log file, just
4576 as it does if it cannot open a log file.
4578 39. Modified OS/Makefile-Linux so that it now contains
4580 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4582 The two -D definitions ensure that Exim is compiled with large file
4583 support, which makes it possible to handle log files that are bigger than
4586 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4587 instance) a domain was checked against a named list that involved a lookup,
4588 causing $domain_data to be set, then another domain was checked against the
4589 same list, then the first domain was re-checked, the value of $domain_data
4590 after the final check could be wrong. In particular, if the second check
4591 failed, it could be set empty. This bug probably also applied to
4594 41. The strip_trailing_dot option was not being applied to the address given
4595 with the -f command-line option.
4597 42. The code for reading a message's header from the spool was incrementing
4598 $received_count, but never initializing it. This meant that the value was
4599 incorrect (doubled) while delivering a message in the same process in which
4600 it was received. In the most common configuration of Exim, this never
4601 happens - a fresh exec is done - but it can happen when
4602 deliver_drop_privilege is set.
4604 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4605 it now includes up to 150 characters of the unexpected data in the log
4608 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4609 and building data strings. The size of both of these buffers was 10 000
4610 bytes - far larger than anybody would *ever* want, thought I. Needless to
4611 say, somebody hit the limit. I have increased the maximum line length to
4612 20 000 and the maximum data length of concatenated lines to 100 000. I have
4613 also fixed two bugs, because there was no checking on these buffers. Tsk,
4614 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4615 buffer is too small.
4617 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4618 lsearch lookups. Now it does.
4620 46. When parsing a route_list item in a manualroute router, a fixed-length
4621 buffer was used for the list of hosts. I made this 1024 bytes long,
4622 thinking that nobody would ever have a list of hosts that long. Wrong.
4623 Somebody had a whole pile of complicated expansion conditions, and the
4624 string was silently truncated, leading to an expansion error. It turns out
4625 that it is easier to change to an unlimited length (owing to other changes
4626 that have happened since this code was originally written) than to build
4627 structure for giving a limitation error. The length of the item that
4628 expands into the list of hosts is now unlimited.
4630 47. The lsearch lookup could not handle data where the length of text line was
4631 more than 4095 characters. Such lines were truncated, leading to shortened
4632 data being returned. It should now handle lines of any length.
4634 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4635 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4638 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4639 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4640 no longer happens in comment lines. A list of replacements is now placed
4641 at the head of all of the source files, except those whose only change is
4642 to replace PERL_COMMAND in the very first #! line.
4644 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4645 messages on the queue, with a bottom-up merge sort, using code contributed
4646 by Michael Haardt. This should make operations like -bp somewhat faster on
4647 large queues. It won't affect queue runners, except when queue_run_in_order
4650 51. Installed eximstats 1.31 in the distribution.
4652 52. Added support for SRV lookups to the dnslookup router.
4654 53. If an ACL referred to $message_body or $message_body_end, the value was not
4655 reset for any messages that followed in the same SMTP session.
4657 54. The store-handling optimization for building very long strings was not
4658 differentiating between the different store pools. I don't think this
4659 actually made any difference in practice, but I've tidied it.
4661 55. While running the routers to verify a sender address, $sender_address
4662 was still set to the sender address. This is wrong, because when routing to
4663 send a bounce to the sender, it would be empty. Therefore, I have changed
4664 it so that, while verifying a sender address, $sender_address is set to <>.
4665 (There is no change to what happens when verifying a recipient address.)
4667 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4668 target A or AAAA records (if not already returned) without resetting the
4669 qualify_single or search_parents options of the DNS resolver. These are
4670 inappropriate in this case because the targets of MX and SRV records must
4671 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4672 target that, when qualified, matched something in the local domain. These
4673 two options are now turned off when doing these lookups.
4675 57. It seems that at least some releases of Reiserfs (which does not have the
4676 concept of a fixed number of inodes) returns zero and not -1 for the
4677 number of available inodes. This interacted badly with check_spool_inodes,
4678 which assumed that -1 was the "no such thing" setting. What I have done is
4679 to check that the total number of inodes is greater than zero before doing
4680 the test of how many are available.
4682 58. When a "warn" ACL statement has a log_message modifier, the message is
4683 remembered, and not repeated. This is to avoid a lot of repetition when a
4684 message has many recipients that cause the same warning to be written.
4685 Howewer, Exim was preserving the list of already written lines for an
4686 entire SMTP session, which doesn't seem right. The memory is now reset if a
4687 new message is started.
4689 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4690 debugging output unless log_rewrite was also set.
4692 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4693 of (int)(handle) when we know that handle contains (void *)(-1).
4695 61. The Exim daemon panic-logs an error return when it closes the incoming
4696 connection. However "connection reset by peer" seems to be common, and
4697 isn't really an error worthy of noting specially, so that particular error
4700 62. When Exim is trying to find all the local interfaces, it used to panic and
4701 die if the ioctl to get the interface flags failed. However, it seems that
4702 on at least one OS (Solaris 9) it is possible to have an interface that is
4703 included in the list of interfaces, but for which you get a failure error
4704 for this call. This happens when the interface is not "plumbed" into a
4705 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4706 failure of the "get flags" call assumes that the interface is down.
4708 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4709 makes life easier for people who are doing arithmetic on fields extracted
4710 from dates, where you often get leading zeros that should not be
4711 interpreted as octal.
4713 64. Added qualify_domain to the redirect router, to override the global
4716 65. If a pathologically long header line contained very many addresses (the
4717 report of this problem mentioned 10 000) and each of them was rewritten,
4718 Exim could use up a very large amount of memory. (It kept on making new
4719 copies of the header line as it rewrote, and never released the old ones.)
4720 At the expense of a bit more processing, the header rewriting function has
4721 been changed so that it no longer eats memory in this way.
4723 66. The generation of the Received: header has been moved from the time that a
4724 message starts to be received, to the time that it finishes. The timestamp
4725 in the Received: header should now be very close to that of the <= log
4726 line. There are two side-effects of this change:
4728 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4729 logged header lines no longer include the local Received: line, because
4730 it has not yet been created. The same applies to a copy of the message
4731 that is returned to a non-SMTP sender when a message is rejected.
4733 (b) When a filter file is tested using -bf, no additional Received: header
4734 is added to the test message. After some thought, I decided that this
4737 This change does not affect the value of $received_for. It is still set
4738 after address rewriting, but before local_scan() is called.
4740 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4742 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4743 gave an unhelpful panic error message, and a defer error. I have managed to
4744 change this behaviour so that it now rejects any supplied certificate,
4745 which seems right, as the list of acceptable certificates is empty.
4747 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4748 gave an unhelpful defer error. I have not managed to make this reject any
4749 supplied certificates, but the error message it gives is "no certificate
4750 supplied", which is not helpful.
4752 70. exigrep's output now also includes lines that are not associated with any
4753 message, but which match the given pattern. Implemented by a patch from
4754 Martin Sluka, which also tidied up the Perl a bit.
4756 71. Recipient callout verification, like sender verification, was using <> in
4757 the MAIL FROM command. This isn't really the right thing, since the actual
4758 sender may affect whether the remote host accepts the recipient or not. I
4759 have changed it to use the actual sender in the callout; this means that
4760 the cache record is now keyed on a recipient/sender pair, not just the
4761 recipient address. There doesn't seem to be a real danger of callout loops,
4762 since a callout by the remote host to check the sender would use <>.
4763 [SEE ABOVE: changed after hitting problems.]
4765 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4766 temporary errors. However, in the case of such a code being given after
4767 the end of a data transmission (i.e. after ".") Exim was failing to write
4768 a retry record for the message. (Yes, there was some broken host that was
4769 actually sending 8xx at this point.)
4771 73. An unknown lookup type in a host list could cause Exim to panic-die when
4772 the list was checked. (An example that provoked this was putting <; in the
4773 middle of a list instead of at the start.) If this happened during a DATA
4774 ACL check, a -D file could be left lying around. This kind of configuration
4775 error no longer causes Exim to die; instead it causes a defer errror. The
4776 incident is still logged to the main and panic logs.
4778 74. Buglet left over from Exim 3 conversion. The message "too many messages
4779 in one connection" was written to the rejectlog but not the mainlog, except
4780 when address rewriting (yes!) was being logged.
4782 75. Added write_rejectlog option.
4784 76. When a system filter was run not as root (that is, when system_filter_user
4785 was set), the values of the $n variables were not being returned to the
4786 main process; thus, they were not subsequently available in the $sn
4789 77. Added +return_path_on_delivery log selector.
4791 78. A connection timeout was being treated differently from recipients deferred
4792 when testing hosts_max_try with a message that was older than the host's
4793 retry timeout. (The host should not be counted, thus allowing all hosts to
4794 be tried at least once before bouncing.) This may have been the cause of an
4795 occasionally reported bug whereby a message would remain on the queue
4796 longer than the retry timeout, but would be bounced if a delivery was
4797 forced. I say "may" because I never totally pinned down the problem;
4798 setting up timeout/retry tests is difficult. See also the next item.
4800 79. The ultimate address timeout was not being applied to errors that involved
4801 a combination of host plus message (for example, a timeout on a MAIL
4802 command). When an address resolved to a number of possible hosts, and they
4803 were not all tried for each delivery (e.g. because of hosts_max_try), a
4804 message could remain on the queue longer than the retry timeout.
4806 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4809 81. Fixed an obscure SMTP outgoing bug which required at least the following
4810 conditions: (a) there was another message waiting for the same server;
4811 (b) the server returned 5xx to all RCPT commands in the first message so
4812 that the message was not completed; (c) the server dropped the connection
4813 or gave a negative response to the RSET that Exim sends to abort the
4814 transaction. The observed case was a dropped connection after DATA that had
4815 been sent in pipelining mode. That is, the server had advertised PIPELINING
4816 but was not implementing it correctly. The effect of the bug was incorrect
4817 behaviour, such as trying another host, and this could lead to a crash.
4823 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4824 and daemon.c were passed as pointers to ints; they should have been
4825 pointers to socklen_t variables (which are typically unsigned ints).
4827 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4830 3. Fixed a really odd bug that affected only the testing scheme; patching a
4831 certain fixed string in the binary changed the value of another string that
4832 happened to be identical to the end of the original first string.
4834 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4835 name", it returns that address as the IP address. On some operating
4836 systems (e.g. Solaris), it also passes back the IP address string as the
4837 "host name". However, on others (e.g. Linux), it passes back an empty
4838 string. Exim wasn't checking for this, and was changing the host name to an
4839 empty string, assuming it had been canonicized.
4841 5. Although rare, it is permitted to have more than one PTR record for a given
4842 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4843 all the names associated with an address, because they do in Solaris.
4844 However, it seems that they do not in Linux for data that comes from the
4845 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4846 I found this out when I moved to a new Linux workstation and tried to run
4847 the Exim test suite.
4849 To get round this problem I have changed the code so that it now does its
4850 own call to the DNS to look up PTR records when searching for a host name.
4851 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4852 addresses that are only in /etc/hosts are still found.
4854 This behaviour is, however, controlled by an option called host_lookup_
4855 order, which defaults to "bydns:byaddr". If people want to use the other
4856 order, or indeed, just use one or the other means of lookup, they can
4857 specify it in this variable.
4859 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4860 some operating systems, this comes back from gethostbyaddr() as an empty
4861 string, and this is what Exim used to test for. However, it seems that in
4862 other systems, "." is yielded. Exim now tests for this case too.
4864 7. The values of check_spool_space and check_log_space are now held internally
4865 as a number of kilobytes instead of an absolute number of bytes. If a
4866 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4867 kilobyte. This means that much larger values can be stored.
4869 8. Exim monitor: an attempt to get the action menu when not actually pointing
4870 at a message produces an empty menu entitled "No message selected". This
4871 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4872 no entries in it ("Shell widget menu has zero width and/or height"). So I
4873 have added a single, blank menu entry in this case.
4875 9. Added ${quote_local_part.
4877 10. MIME decoding is now applied to the contents of Subject: header lines when
4880 11. Now that a reference to $sender_host_address automatically causes a reverse
4881 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4882 host lookup before query-style lookups in lists that might use this
4883 variable. This has therefore been abolished, and the "net-" prefix is no
4884 longer necessary for query-style lookups.
4886 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4887 have been a typo for LFLAGS, so it has been changed.
4889 13. The install script calls Exim with "-C /dev/null" in order to find the
4890 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4891 to be output. Howeve, since Exim outputs its version number before the
4892 error, it didn't break the script. It just looked ugly. I fixed this by
4893 always allowing "-C /dev/null" if the caller is root.
4895 14. Ignore overlarge ACL variable number when reading spool file - insurance
4896 against a later release with more variables having written the file.
4898 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4899 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4900 this, as well as the form without the "IPv6" on the front.
4902 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4903 OS/Makefile-Darwin file.
4905 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4906 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4908 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4909 variable. However, some OS use ints for inodes. I've added cast to long int
4910 to get rid of the compiler warning.
4912 19. I had forgotten to lock out "/../" in configuration file names when
4913 ALT_CONFIG_PREFIX was set.
4915 20. Routers used for verification do not need to specify transports. However,
4916 if such a router generated a host list, and callout was configured, Exim
4917 crashed, because it could not find a port number from the (non-existent)
4918 transport. It now assumes port 25 in this circumstance.
4920 21. Added the -t option to exigrep.
4922 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4923 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4924 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4925 them is set, LOOKUP_LSEARCH is forced.
4927 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4929 24. Added sender and host information to the "rejected by local_scan()" log
4930 line; previously there was no indication of these.
4932 25. Added .include_if_exists.
4934 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4935 Linux. It turns out that not all file systems support this. Apparently some
4936 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4937 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4938 not supported on the file descriptor, is now ignored when Exim is trying to
4939 sync a directory. This applies only to Linux.
4941 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4943 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4944 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4947 29. Added support for maildirsize files from supplied patch (modified a bit).
4949 30. The use of :fail: followed by an empty string could lead Exim to respond to
4950 sender verification failures with (e.g.):
4952 550 Verification failed for <xxx>
4953 550 Sender verify failed
4955 where the first response line was missing the '-' that indicates it is not
4956 the final line of the response.
4958 31. The loop for finding the name of the user that called Exim had a hardwired
4959 limit of 10; it now uses the value of finduser_retries, which is used for
4960 all other user lookups.
4962 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4965 33. Exim was neglecting to zero errno before one call of strtol() when
4966 expanding a string and expecting an integer value. On some systems this
4967 resulted in spurious "integer overflow" errors. Also, it was casting the
4968 result into an int without checking.
4970 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4971 did not work. The code looks as if it has *never* worked, though it appears
4972 to have been documented since at least releast 1.62. I have made it work.
4974 35. The "timeout_DNS" error in retry rules, also documented since at least
4975 1.62, also never worked. As it isn't clear exactly what this means, and
4976 clearly it isn't a major issue, I have abolished the feature by treating it
4977 as "timeout", and writing a warning to the main and panic logs.
4979 36. The display of retry rules for -brt wasn't always showing the error code
4982 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4983 timeout_connect_A, timeout_connect_MX.
4985 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4986 to the empty sender.
4988 39. The daemon was not analysing the content of -oX till after it had closed
4989 stderr and disconnected from the controlling terminal. This meant that any
4990 syntax errors were only noted on the panic log, and the return code from
4991 the command was 0. By re-arranging the code a little, I've made the
4992 decoding happen first, so such errors now appear on stderr, and the return
4993 code is 1. However, the actual setting up of the sockets still happens in
4994 the disconnected process, so errors there are still only recorded on the
4997 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4998 connections (as happens on some IP stacks) was logged at start up time as
4999 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5000 it from "IPv6 and IPv4", which means that two separate sockets are being
5003 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5004 says whether AF_INET or AF_INET6 was passed as an argument.
5006 42. Exiwhat output was messed up when time zones were included in log
5009 43. Exiwhat now gives more information about the daemon's listening ports,
5010 and whether -tls-on-connect was used.
5012 44. The "port" option of the smtp transport is now expanded.
5014 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5015 silently ignored. Now an error message is written to the main and panic
5018 46. There's a new ACL modifier called "logwrite" which writes to a log file
5019 as soon as it is encountered.
5021 47. Added $local_user_uid and $local_user_gid at routing time.
5023 48. Exim crashed when trying to verify a sender address that was being
5026 49. Exim was recognizing only a space character after ".include". It now also
5027 recognizes a tab character.
5029 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5030 extracting the relevant information from the specification. The man page no
5031 longer contains scrambled data for the -d option, and I've added a section
5032 at the front about calling Exim under different names.
5034 51. Added "extra_headers" argument to the "mail" command in filter files.
5036 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5039 53. Installed eximstats 1.29.
5041 54. Added transport_filter_timeout as a generic transport option.
5043 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5044 Cc: header lines. This was required by RFC 822, but it not required by RFC
5047 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5048 incoming messages that did not have them. Now it does so only if the
5049 message originates locally, that is, if there is no associated remote host
5050 address. When Resent- header lines are present, this applies to the Resent-
5051 lines rather than the non-Resent- lines.
5053 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5054 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5056 58. Messages for configuration errors now include the name of the main
5057 configuration file - useful now that there may be more than one file in a
5058 list (.included file names were always shown).
5060 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5061 for those rare installations that do not start the daemon as root or run it
5062 setuid root. I've cut out the call to initgroups() if the daemon is not
5065 60. The Exim user and group can now be bound into the binary as text strings
5066 that are looked up at the start of Exim's processing.
5068 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5070 62. Added $mailstore_basename variable.
5072 63. Installed patch to sieve.c from Michael Haardt.
5074 64. When Exim failed to open the panic log after failing to open the main log,
5075 the original message it was trying to log was written to stderr and debug
5076 output, but if they were not available (the usual case in production), it
5077 was lost. Now it is written to syslog before the two lines that record the
5078 failures to open the logs.
5080 65. Users' Exim filters run in subprocesses under the user's uid. It is
5081 possible for a "deliver" command or an alias in a "personal" command to
5082 provoke an address rewrite. If logging of address rewriting is configured,
5083 this fails because the process is not running as root or exim. There may be
5084 a better way of dealing with this, but for the moment (because 4.30 needs
5085 to be released), I have disabled address rewrite logging when running a
5086 filter in a non-root, non-exim process.
5092 1. The buildconfig auxiliary program wasn't quoting the value set for
5093 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5094 not defined. This bug was masked in 4.22 by the effect that was fixed in
5097 2. Some messages that were rejected after a message id was allocated were
5098 shown as "incomplete" by exigrep. It no longer does this for messages that
5099 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5101 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5102 have allow_domain_literals set, the ID did not get logged in the <= line.
5103 Domain literals are now always recognized in Message-ID: header lines.
5105 4. The first argument for a ${extract expansion item is the key name or field
5106 number. Leading and trailing spaces in this item were not being ignored,
5107 causing some misleading effects.
5109 5. When deliver_drop_privilege was set, single queue runner processes started
5110 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5111 same command in the process it spins off) were not dropping privilege.
5113 6. When the daemon running as "exim" started a queue runner, it always
5114 re-executed Exim in the spun-off process. This is a waste of effort when
5115 deliver_drop_privilege is set. The new process now just calls the
5116 queue-runner function directly.
5122 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5125 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5126 ignored. Though the use of -oP was forcing the writing of a pid file, it
5127 was always written to the default place.
5129 3. If the message "no IP address found for host xxxx" is generated during
5130 incoming verification, it is now followed by identification of the incoming
5131 connection (so you can more easily find what provoked it).
5133 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5135 5. Added some features to "harden" Exim a bit more against certain attacks:
5137 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5138 be put in Local/Makefile. This is like the never_users runtime option,
5139 but it cannot be overridden. The default setting is "root".
5141 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5142 prefix string with which any file named in a -C command line option
5145 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5146 is retained for -C and -D only if the caller of Exim is root. Without
5147 it, the exim user may also use -C and -D and retain privilege.
5149 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5150 command line option is disabled.
5152 6. Macro names set by the -D option must start with an upper case letter, just
5153 like macro names defined in the configuration file.
5155 7. Added "dereference=" facility to LDAP.
5157 8. Two instances of the typo "uknown" in the source files are fixed.
5159 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5160 the Configure-Makefile script screwed up while processing it.
5162 10. Incorporated PCRE 4.4.
5164 11. The SMTP synchronization check was not operating right at the start of an
5165 SMTP session. For example, it could not catch a HELO sent before the client
5166 waited for the greeting. There is now a check for outstanding input at the
5167 point when the greeting is written. Because of the duplex, asynchronous
5168 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5169 way, but not yet received, when the check is performed.
5171 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5172 on TCP/IP sockets, because this apparently causes some broken clients to
5175 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5176 unchanged) from the Cygwin maintainer.
5178 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5179 was supported instead of testing for mbx. Effectively a typo.
5181 15. The spa authenticator server code was not checking that the input it
5182 received was valid base64.
5184 16. The debug output line for the "set" modifier in ACLs was not showing the
5185 name of the variable that was being set.
5187 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5189 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5190 lookup on its own. Something else was needed to trigger the lookup. For
5191 example, a match in host_lookup or the need for a host name in a host list.
5192 Now, if $sender_host_name is referenced and the host name has not yet been
5193 looked up, a lookup is performed. If the lookup fails, the variable remains
5194 empty, and $host_lookup_failed is set to "1".
5196 19. Added "eqi" as a case-independent comparison operator.
5198 20. The saslauthd authentication condition could segfault if neither service
5199 nor realm was specified.
5201 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5202 error message that was logged was misleading, and incoming SMTP
5203 connections were dropped. The message is now more accurate, and temporary
5204 errors are given to SMTP connections.
5206 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5207 (except RSET) with a 421 error, until QUIT is received. However, it was
5208 failing to send a response to QUIT.
5210 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5211 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5212 ACL is now run after the helo_try_verify_hosts code.
5214 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5215 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5216 case-independent, so other case variants are also recognized). Apparently
5217 some systems use these upper case variants.
5219 25. If more than two messages were waiting for the same host, and a transport
5220 filter was specified for the transport, Exim sent two messages over the
5221 same TCP/IP connection, and then failed with "socket operation on non-
5222 socket" when it tried to send the third.
5224 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5226 27. The extern definition of crypt16() in expand.c was not being excluded when
5227 the OS had its own crypt16() function.
5229 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5230 as a preferred synonym for return_size_limit, both as an option and as an
5233 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5235 30. Changed the default configuration ACL to relax the local part checking rule
5236 for addresses that are not in any local domains. For these addresses,
5237 slashes and pipe symbols are allowed within local parts, but the sequence
5238 /../ is explicitly forbidden.
5240 31. SPA server authentication was not clearing the challenge buffer before
5243 32. log_message in a "warn" ACL statement was writing to the reject log as
5244 well as to the main log, which contradicts the documentation and doesn't
5245 seem right (because no rejection is happening). So I have stopped it.
5247 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5248 However, I am unable to do any testing of this.
5250 34. Fixed an infelicity in the appendfile transport. When checking directories
5251 for a mailbox, to see if any needed to be created, it was accidentally
5252 using path names with one or more superfluous leading slashes; tracing
5253 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5255 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5256 contents are added to the log line that is written for every discarded
5257 recipient. (Previously a log_message setting was ignored.)
5259 36. The ${quote: operator now quotes the string if it is empty.
5261 37. The install script runs exim in order to find its version number. If for
5262 some reason other than non-existence or emptiness, which it checks, it
5263 could not run './exim', it was installing it with an empty version number,
5264 i.e. as "exim-". This error state is now caught, and the installation is
5267 38. An argument was missing from the function that creates an error message
5268 when Exim fails to connect to the socket for saslauthd authentication.
5269 This could cause Exim to crash, or give a corrupted message.
5271 39. Added isip, isip4, and isip6 to ${if conditions.
5273 40. The ACL variables $acl_xx are now saved with the message, and can be
5274 accessed later in routers, transports, and filters.
5276 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5277 strings in the file are not string-expanded.
5279 42. If a MAIL command specified a SIZE value that was too large to fit into an
5280 int variable, the check against message_size_limit failed. Such values are
5281 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5282 day this will have to be increased, but I don't think I want to be around
5283 when emails are that large.
5290 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5291 iconv() is not standard in FreeBSD.
5293 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5294 IPv6 enabled. The observed symptom was a segmentation fault on return from
5295 the function os_common_find_running_interfaces() in src/os.c.
5297 3. In the check_special_case() function in daemon.c I had used "errno" as an
5298 argument name, which causes warnings on some systems. This was basically a
5299 typo, since it was named "eno" in the comments!
5301 4. The code that waits for the clock to tick (at a resolution of some fraction
5302 of a second) so as to ensure message-id uniqueness was always waiting for
5303 at least one whole tick, when it could have waited for less. [This is
5304 almost certainly not relevant at current processor speeds, where it is
5305 unlikely to ever wait at all. But we try to future-proof.]
5307 5. The function that sleeps for a time interval that includes fractions of a
5308 second contained a race. It did not block SIGALRM between setting the
5309 timer, and suspending (a couple of lines later). If the interval was short
5310 and the sigsuspend() was delayed until after it had expired, the suspension
5311 never ended. On busy systems this could lead to processes getting stuck for
5314 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5315 process, before it forks any delivery processes. The open lookup caching
5316 mechanism meant that the open file or database connection was passed into
5317 the delivery process. The problem was that delivery processes always tidy
5318 up cached lookup data. This could cause a problem for the next delivery
5319 process started by the queue runner, because the external queue runner
5320 process does not know about the closure. So the next delivery process
5321 still has data in the lookup cache. In the case of a file lookup, there was
5322 no problem because closing a file descriptor in a subprocess doesn't affect
5323 the parent. However, if the lookup was caching a connection to a database,
5324 the connection was closed, and the second delivery process was likely to
5325 see errors such as "PGSQL: query failed: server closed the connection
5326 unexpectedly". The problem has been fixed by closing all cached lookups
5327 in a queue runner before running a delivery process.
5329 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5330 seem to have the "const" qualifier which it has on other OS. I've
5333 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5334 *of the same type* (client or server) with the same public name that an
5335 error should be diagnosed.
5337 9. When Exim looked up a host name for an IP address, but failed to find the
5338 original IP address when looking up the host name (a safety check), it
5339 output the message "<ip address> does not match any IP for NULL", which was
5340 confusing, to say the least. The bug was that the host name should have
5341 appeared instead of "NULL".
5343 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5344 user that is built into the binary, and the -C or -D options is used, root
5345 privilege is dropped before the configuration file is read. In addition,
5346 logging is switched to stderr instead of the normal log files. If the
5347 configuration then re-defines the Exim user, the unprivileged environment
5348 is probably not what is expected, so Exim logs a panic warning message (but
5351 However, if deliver_drop_privilege is set, the unprivileged state may well
5352 be exactly what is intended, so the warning has been cut out in that case,
5353 and Exim is allowed to try to write to its normal log files.
5359 1. smtp_return_error_details was not giving details for temporary sender
5360 or receiver verification errors.
5362 2. Diagnose a configuration error if two authenticators have the same public
5365 3. Exim used not to create the message log file for a message until the first
5366 delivery attempt. This could be confusing when incoming messages were held
5367 for policy or load reasons. The message log file is now created at the time
5368 the message is received, and an initial "Received" line is written to it.
5370 4. The automatically generated man page for command line options had a minor
5371 bug that caused no ill effects; however, a more serious problem was that
5372 the procedure for building the man page automatically didn't always
5373 operate. Consequently, release 4.20 contains an out-of-date version. This
5374 shouldn't happen again.
5376 5. When building Exim with embedded Perl support, the script that builds the
5377 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5378 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5380 6. The freeze_tell option was not being used for messages that were frozen on
5381 arrival, either by an ACL or by local_scan().
5383 7. Added the smtp_incomplete_transaction log selector.
5385 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5386 was accepting AUTH without a new EHLO.
5388 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5389 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5392 10. Logging of TCP/IP connections (when configured) now happens in the main
5393 daemon process instead of the child process, so that the TCP/IP connection
5394 count is more accurate (but it can never be perfect).
5396 11. The use of "drop" in a nested ACL was not being handled correctly in the
5397 outer ACL. Now, if condition failure induced by the nested "drop" causes
5398 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5399 or "require"), the connection is dropped.
5401 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5402 that yield "discard" can now be used with an "accept" or a "discard" verb,
5403 but an error is generated for any others (because I can't see a useful way
5404 to define what should happen).
5406 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5407 are now processed in the same way as lines in the Exim configuration file.
5408 In particular, continuation lines are supported.
5410 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5412 15. Added -ti meaning -t -i.
5414 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5415 domains, and warn by logging if others are found.
5417 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5418 returned from GIFCONF when Exim is trying to find the list of interfaces on
5419 a host. The code in os.c has been modified to copy each ifreq to an aligned
5420 structure in all cases.
5422 Also, in some cases, the returned ifreq's were being copied to a 'struct
5423 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5424 means the last couple of bytes of an IPv6 address could be chopped if the
5425 ifreq contained only a normal sockaddr (14 bytes storage).
5427 18. Named domain lists were not supported in the hosts_treat_as_local option.
5428 An entry such as +xxxx was not recognized, and was treated as a literal
5431 19. Ensure that header lines added by a DATA ACL are included in the reject log
5432 if the ACL subsequently rejects the message.
5434 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5435 MD5 (which is deprecated).
5437 21. When testing a filter file using -bf, Exim was writing a message when it
5438 took the sender from a "From " line in the message, but it was not doing so
5439 when it took $return_path from a Return-Path: header line. It now does.
5441 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5442 with a valid header line field name (a series of printing characters
5443 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5445 23. Changed "disc" in the source to "disk" to conform to the documentation and
5446 the book and for uniformity.
5448 24. Ignore Sendmail's -Ooption=value command line item.
5450 25. When execve() failed while trying to run a command in a pipe transport,
5451 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5452 could be confused with a return value of 69 from the command itself. This
5453 has been changed to 127, the value the shell returns if it is asked to run
5454 a non-existent command. The wording for the related log line suggests a
5455 non-existent command as the problem.
5457 26. If received_header_text expands to an empty string, do not add a Received:
5458 header line to the message. (Well, it adds a token one on the spool, but
5459 marks it "old" so that it doesn't get used or transmitted.)
5461 27. Installed eximstats 1.28 (addition of -nt option).
5463 28. There was no check for failure on the call to getsockname() in the daemon
5464 code. This can fail if there is a shortage of resources on the system, with
5465 ENOMEM, for example. A temporary error is now given on failure.
5467 29. Contrary to the C standard, it seems that in some environments, the
5468 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5469 program. Exim now does this explicitly; it affects the formatting of
5470 timestamps using strftime().
5472 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5473 complaints. I've now initialized all the variables, to avoid this.
5475 32. Header lines added by a system filter were not being "seen" during
5476 transport-time rewrites.
5478 33. The info_callback() function passed to OpenSSL is set up with type void
5479 (*)(SSL *, int, int), as described somewhere. However, when calling the
5480 function (actually a macro) that sets it up, the type void(*)() is
5481 expected. I've put in a cast to prevent warnings from picky compilers.
5483 34. If a DNS black list lookup found a CNAME record, but there were no A
5484 records associated with the domain it pointed at, Exim crashed.
5486 35. If a DNS black list lookup returned more than one A record, Exim ignored
5487 all but the first. It now scans all returned addresses if a particular IP
5488 value is being sought. In this situation, the contents of the
5489 $dnslist_value variable are a list of all the addresses, separated by a
5492 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5493 to accept a host name and all its aliases if the forward lookup for any of
5494 them yielded the IP address of the incoming connection. Now it accepts only
5495 those names whose forward lookup yields the correct IP address. Any other
5496 names are discarded. This closes a loophole whereby a rogue DNS
5497 administrator could create reverse DNS records to break through a
5498 wildcarded host restriction in an ACL.
5500 37. If a user filter or a system filter that ran in a subprocess used any of
5501 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5502 the wrong values were passed to the pipe command ($thisaddress had the
5503 value of $0, $0 had the value of $1, etc). This bug was introduced by
5504 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5506 38. Improved the line breaking for long SMTP error messages from ACLs.
5507 Previously, if there was no break point between 40 and 75 characters, Exim
5508 left the rest of the message alone. Two changes have been made: (a) I've
5509 reduced the minimum length to 35 characters; (b) if it can't find a break
5510 point between 35 and 75 characters, it looks ahead and uses the first one
5511 that it finds. This may give the occasional overlong line, but at least the
5512 remaining text gets split now.
5514 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5515 file descriptors might be low, and that setting 1000 would always raise it.
5516 It turns out that in some environments, the limit is already over 1000 and
5517 that lowering it causes trouble. So now Exim takes care not to decrease it.
5519 40. When delivering a message, the value of $return_path is set to $sender_
5520 address at the start of routing (routers may change the value). By an
5521 oversight, this default was not being set up when an address was tested by
5522 -bt or -bv, which affected the outcome if any router or filter referred to
5525 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5526 log only when "message" or "log_message" are set. However, if one of the
5527 conditions was an address verification, or a call to a nested ACL, the
5528 messages generated by the underlying test were being passed through. This
5529 no longer happens. The underlying message is available in $acl_verify_
5530 message for both "message" and "log_message" expansions, so it can be
5531 passed through if needed.
5533 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5534 new expansion $bh_ to give the encoded byte string without charset
5535 translation. Translation happens only if iconv() is available; HAVE_ICONV
5536 indicates this at build time. HEADERS_CHARSET gives the charset to
5537 translate to; headers_charset can change it in the configuration, and
5538 "headers charset" can change it in an individual filter file.
5540 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5541 that creates RFC 2047 encoded "words" labels them as that charset instead
5542 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5543 expansion operator; (ii) when Exim creates a From: line for a local
5544 message; (iii) when a header line is rewritten to include a "phrase" part.
5546 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5547 buggy, causing it to skip the first lines of messages whose message ID
5548 ended in 'D'. This would not have bitten before Exim release 4.14, because
5549 message IDs were unlikely to end in 'D' before then. The effect was to have
5550 incorrect size information for certain domains.
5552 45. #include "config.h" was missing at the start of the crypt16.c module. This
5553 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5556 46. If there was a timeout during a "random" callout check, Exim treated it as
5557 a failure of the random address, and carried on sending RSET and the real
5558 address. If the delay was just some slowness somewhere, the response to the
5559 original RCPT would be taken as a response to RSET and so on, causing
5560 mayhem of various kinds.
5562 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5563 when I implemented it. It didn't allow for the fact that some option values
5564 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5565 the right test for positive values.
5567 48. Domain names in DNS records are case-independent. Exim always looks them up
5568 in lower case. Some resolvers return domain names in exactly the case they
5569 appear in the zone file, that is, they may contain uppercase letters. Not
5570 all resolvers do this - some return always lower case. Exim was treating a
5571 change of case by a resolver as a change of domain, similar to a widening
5572 of a domain abbreviation. This triggered its re-routing code and so it was
5573 trying to route what was effectively the same domain again. This normally
5574 caused routing to fail (because the router wouldn't handle the domain
5575 twice). Now Exim checks for this case specially, and just changes the
5576 casing of the domain that it ultimately uses when it transmits the message
5579 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5582 50. If a filter generated a file delivery with a non-absolute name (possible if
5583 no home directory exists for the router), the forbid_file option was not
5586 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5587 the existing equality matching.
5589 52. Exim was using ints instead of ino_t variables in some places where it was
5590 dealing with inode numbers.
5592 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5593 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5594 called TMPDIR, and if it finds it is different, it changes its value.
5596 54. The smtp_printf() function is now made available to local_scan() so
5597 additional output lines can be written before returning. There is also an
5598 smtp_fflush() function to enable the detection of a dropped connection.
5599 The variables smtp_input and smtp_batched_input are exported to
5602 55. Changed the default runtime configuration: the message "Unknown user"
5603 has been removed from the ACL, and instead placed on the localuser router,
5604 using the cannot_route_message feature. This means that any verification
5605 failures that generate their own messages won't get overridden. Similarly,
5606 the "Unrouteable address" message that was in the ACL for unverifiable
5607 relay addresses has also been removed.
5609 56. Added hosts_avoid_esmtp to the smtp transport.
5611 57. The exicyclog script was not checking for the esoteric option
5612 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5613 will work only if exicyclog is run under the appropriate euid.
5615 58. Following a discussion on the list, the rules by which Exim recognises line
5616 endings on incoming messages have been changed. The -dropcr and drop_cr
5617 options are now no-ops, retained only for backwards compatibility. The
5618 following line terminators are recognized: LF CRLF CR. However, special
5619 processing applies to CR:
5621 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5622 nor a local message in the state where . is a terminator.
5624 (ii) If a bare CR is encountered in a header line, an extra space is added
5625 after the line terminator so as not to end the header. The reasoning
5626 behind this is that bare CRs in header lines are most likely either
5627 to be mistakes, or people trying to play silly games.
5629 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5630 was being incorrectly given as 18 bytes larger than it should have been.
5631 This is a VOB (very old bug).
5633 60. This may never have affected anything current, but just in case it has:
5634 When the local host is found other than at the start of a list of hosts,
5635 the local host, those with the same MX, and any that follow, are discarded.
5636 When the list in question was part of a longer list of hosts, the following
5637 hosts (not currently being processed) were also being discarded. This no
5638 longer happens. I'm not sure if this situation could ever has previously
5641 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5644 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5645 header. This is something that is recommended in a new Internet Draft, and
5646 is something that is documented as being done by Sendmail. There are two
5647 possible values. For messages generated by the autoreply transport, Exim
5650 Auto-Submitted: auto-replied
5652 whereas for all other generated messages (e.g. bounces) it adds
5654 Auto-Submitted: auto-generated
5656 63. The "personal" condition in filters now includes a test for the
5657 Auto-Submitted: header. If it contains the string "auto-" the message it
5658 not considered personal.
5660 64. Added rcpt_include_affixes as a generic transport option.
5662 65. Added queue_only_override (default true).
5664 66. Added the syslog_duplication option.
5666 67. If what should have been the first header line of a message consisted of
5667 a space followed by a colon, Exim was mis-interpreting it as a header line.
5668 It isn't of course - it is syntactically invalid and should therefore be
5669 treated as the start of the message body. The misbehaviour could have
5670 caused a number of strange effects, including loss of data in subsequent
5671 header lines, and spool format errors.
5673 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5674 client host had authenticated. This control can now be exercised by an ACL
5675 for more flexibility.
5677 69. By default, callouts do not happen when testing with -bh. There is now a
5678 variant, -bhc, which does actually run the callout code, including
5679 consulting and updating the callout cache.
5681 70. Added support for saslauthd authentication, courtesy of Alexander
5684 71. If statvfs() failed on the spool or log directories while checking their
5685 size for availability, Exim confusingly gave the error "space shortage".
5686 Furthermore, in debugging mode it crashed with a floating point exception.
5687 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5688 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5689 serious problem, Exim now writes to the main and panic logs when this
5690 happens, with details of the failure. It then refuses to accept the
5691 incoming message, giving the message "spool directory problem" or "log
5692 directory problem" with a 421 code for SMTP messages.
5694 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5695 0, 1, and 2 exist, because some OS complain for execs without them (see
5696 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5697 descriptors. However, the code omitted to check that the open succeeded,
5698 causing mysterious errors if for some reason the permissions on /dev/null
5699 got screwed. Now Exim writes a message to the main and panic logs, and
5700 bombs out if it can't open /dev/null.
5702 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5703 interact so that it is all more flexible. It is supposed to remain
5704 backwards compatible. Also added extra_local_interfaces.
5706 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5707 to bomb out with an assertion failure - to the client this appears as a
5708 connection drop. This problem occurs in the part of the code that was taken
5709 from the Samba project. Fortunately, the assertion is in a very simple
5710 function, so I have fixed this by reproducing the function inline in the
5711 one place where it is called, and arranging for authentication to fail
5712 instead of killing the process with assert().
5714 75. The SPA client code was not working when the server requested OEM rather
5715 than Unicode encoding.
5717 76. Added code to make require_files with a specific uid setting more usable in
5718 the case where statting the file as root fails - usually a non-root-mounted
5719 NFS file system. When this happens and the failure is EACCES, Exim now
5720 forks a subprocess and does the per-uid checking as the relevant uid.
5722 77. Added process_log_path.
5724 78. If log_file_path was not explicitly set, a setting of check_log_space or
5725 check_log_inodes was ignored.
5727 79. If a space check for the spool or log partitions fails, the incident is now
5728 logged. Of course, in the latter case the data may get lost...
5730 80. Added the %p formatting code to string_format() so that it can be used to
5731 print addresses in debug_print(). Adjusted all the address printing in the
5732 debugging in store.c to use %p rather than %d.
5734 81. There was a concern that a line of code in smtp_in.c could overflow a
5735 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5736 initially expressed, the concern was not well-founded, because trailing
5737 spaces are removed early. However, if the trailing spaces were followed by
5738 a NULL, they did not get removed, so the overflow was possible. Two fixes
5741 (a) I re-wrote the offending code in a cleaner fashion.
5742 (b) If an incoming SMTP command contains a NULL character, it is rejected
5745 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5746 runs initgroups(), so that if the Exim user is in any additional groups,
5747 they will be used during message reception.
5753 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob