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.
98 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
100 This, or similar/improved, will also be change PP/11 of 4.82.
106 PP/01 Handle short writes when writing local log-files.
107 In practice, only affects FreeBSD (8 onwards).
108 Bugzilla 1053, with thanks to Dmitry Isaikin.
110 NM/01 Bugzilla 949 - Documentation tweak
112 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
115 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
117 PP/02 Implemented gsasl authenticator.
119 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
121 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
122 `pkg-config foo` for cflags/libs.
124 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
125 with rest of GSASL and with heimdal_gssapi.
127 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
128 `pkg-config foo` for cflags/libs for the TLS implementation.
130 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
131 properties get this fed in as external SSF. A number of robustness
132 and debugging improvements to the cyrus_sasl authenticator.
134 PP/08 cyrus_sasl server now expands the server_realm option.
136 PP/09 Bugzilla 1214 - Log authentication information in reject log.
137 Patch by Jeremy Harris.
139 PP/10 Added dbmjz lookup type.
141 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
143 PP/12 MAIL args handles TAB as well as SP, for better interop with
144 non-compliant senders.
145 Analysis and variant patch by Todd Lyons.
147 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
148 Bug report from Lars Müller <lars@samba.org> (via SUSE),
149 Patch from Dirk Mueller <dmueller@suse.com>
151 PP/13 tls_peerdn now print-escaped for spool files.
152 Observed some $tls_peerdn in wild which contained \n, which resulted
153 in spool file corruption.
155 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
156 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
157 or write after TLS renegotiation, which otherwise led to messages
160 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
161 as a tracking header (ie: a signed header comes before the signature).
162 Patch from Wolfgang Breyha.
164 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
165 comma-sep list; embedded commas doubled.
167 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
169 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
171 Report and patch from Dmitry Banschikov.
173 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
174 Removed SSL_clear() after SSL_new() which led to protocol negotiation
175 failures. We appear to now support TLS1.1+ with Exim.
177 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
178 lets Exim select keys and certificates based upon TLS SNI from client.
179 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
180 before an outbound SMTP session. New log_selector, +tls_sni.
182 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
183 NULL dereference. Report and patch from Alun Jones.
185 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
186 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
187 Not seeing resolver debug output on NetBSD, but suspect this is a
188 resolver implementation change.
190 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
191 Left warnings. Added "eximon gdb" invocation mode.
193 PP/21 Defaulting "accept_8bitmime" to true, not false.
195 PP/22 Added -bw for inetd wait mode support.
197 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
198 locate the relevant includes and libraries. Made this the default.
200 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
201 Bugzilla 1246, report and most of solution from Tomasz Kusy.
203 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
204 This may cause build issues on older platforms.
206 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
207 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
208 gnutls_require_mac & gnutls_require_protocols (no longer supported).
209 Added SNI support via GnuTLS too.
210 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
212 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
214 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
215 Applied second patch from Janne, implementing suggestion to default
216 multiple-strings-in-record handling to match SPF spec.
218 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
220 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
221 read-only, out of scope).
222 Patch from Wolfgang Breyha, report from Stuart Northfield.
224 PP/29 Fix three issues highlighted by clang analyser static analysis.
225 Only crash-plausible issue would require the Cambridge-specific
226 iplookup router and a misconfiguration.
227 Report from Marcin Mirosław.
229 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
231 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
232 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
233 As part of this, removing so much warning spew let me fix some minor
234 real issues in debug logging.
236 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
237 assignment on my part. Fixed.
239 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
240 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
241 Janne Snabb (who went above and beyond: thank you).
243 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
244 string otherwise requires a connection and a bunch more work and it's
245 relatively easy to get wrong. Should also expose TLS library linkage
248 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
249 64-bit ${eval} (JH/03).
251 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
252 GNU libc to support some of the 64-bit stuff, should not lead to
253 conflicts. Defined before os.h is pulled in, so if a given platform
254 needs to override this, it can.
256 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
257 protection layer was required, which is not implemented.
258 Bugzilla 1254, patch from Wolfgang Breyha.
260 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
261 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
262 tls_dhparam take prime identifiers. Also unbreak combination of
263 OpenSSL+DH_params+TLSSNI.
265 PP/39 Disable SSLv2 by default in OpenSSL support.
271 PP/01 Solaris build fix for Oracle's LDAP libraries.
272 Bugzilla 1109, patch from Stephen Usher.
274 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
276 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
279 TF/02 Fix a couple more cases where we did not log the error message
280 when unlink() failed. See also change 4.74-TF/03.
282 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
283 lock up or crash if it happened to be inside a call to libc when it
284 got a SIGUSR1 from exiwhat.
286 The SIGUSR1 handler appends the current process status to the process
287 log which is later printed by exiwhat. It used to use the general
288 purpose logging code to do this, but several functions it calls are
289 not safe for signals.
291 The new output code in the SIGUSR1 handler is specific to the process
292 log, and simple enough that it's easy to inspect for signal safety.
293 Removing some special cases also simplifies the general logging code.
294 Removing the spurious timestamps from the process log simplifies
297 TF/04 Improved ratelimit ACL condition.
299 The /noupdate option has been deprecated in favour of /readonly which
300 has clearer semantics. The /leaky, /strict, and /readonly update modes
301 are mutually exclusive. The update mode is no longer included in the
302 database key; it just determines when the database is updated. (This
303 means that when you upgrde Exim will forget old rate measurements.)
305 Exim now checks that the per_* options are used with an update mode that
306 makes sense for the current ACL. For example, when Exim is processing a
307 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
308 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
309 must specify per_mail/readonly. If you omit the update mode it defaults to
310 /leaky where that makes sense (as before) or /readonly where required.
312 The /noupdate option is now undocumented but still supported for
313 backwards compatibility. It is equivalent to /readonly except that in
314 ACLs where /readonly is required you may specify /leaky/noupdate or
315 /strict/noupdate which are treated the same as /readonly.
317 A useful new feature is the /count= option. This is a generalization
318 of the per_byte option, so that you can measure the throughput of other
319 aggregate values. For example, the per_byte option is now equivalent
320 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
322 The per_rcpt option has been generalized using the /count= mechanism
323 (though it's more complicated than the per_byte equivalence). When it is
324 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
325 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
326 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
327 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
328 Note that using per_rcpt with a non-readonly update mode in more than
329 one ACL will cause the recipients to be double-counted. (The per_mail
330 and per_byte options don't have this problem.)
332 The handling of very low rates has changed slightly. If the computed rate
333 is less than the event's count (usually one) then this event is the first
334 after a long gap. In this case the rate is set to the same as this event's
335 count, so that the first message of a spam run is counted properly.
337 The major new feature is a mechanism for counting the rate of unique
338 events. The new per_addr option counts the number of different
339 recipients that someone has sent messages to in the last time period. It
340 behaves like per_rcpt if all the recipient addresses are different, but
341 duplicate recipient addresses do not increase the measured rate. Like
342 the /count= option this is a general mechanism, so the per_addr option
343 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
344 example, measure the rate that a client uses different sender addresses
345 with the options per_mail/unique=$sender_address. There are further
346 details in the main documentation.
348 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
350 TF/06 Removed a few PCRE remnants.
352 TF/07 Automatically extract Exim's version number from tags in the git
353 repository when doing development or release builds.
355 PP/02 Raise smtp_cmd_buffer_size to 16kB.
356 Bugzilla 879. Patch from Paul Fisher.
358 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
359 Heavily based on revision 40f9a89a from Simon Arlott's tree.
362 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
364 PP/05 Variable $av_failed, true if the AV scanner deferred.
365 Bugzilla 1078. Patch from John Horne.
367 PP/06 Stop make process more reliably on build failure.
368 Bugzilla 1087. Patch from Heiko Schlittermann.
370 PP/07 Make maildir_use_size_file an _expandable_ boolean.
371 Bugzilla 1089. Patch from Heiko Schlittermann.
373 PP/08 Handle ${run} returning more data than OS pipe buffer size.
374 Bugzilla 1131. Patch from Holger Weiß.
376 PP/09 Handle IPv6 addresses with SPF.
377 Bugzilla 860. Patch from Wolfgang Breyha.
379 PP/10 GnuTLS: support TLS 1.2 & 1.1.
381 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
384 PP/11 match_* no longer expand right-hand-side by default.
385 New compile-time build option, EXPAND_LISTMATCH_RHS.
386 New expansion conditions, "inlist", "inlisti".
388 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
390 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
392 PP/14 fix log_write() format string regression from TF/03.
393 Bugzilla 1152. Patch from Dmitry Isaikin.
399 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
401 PP/02 Harmonised TLS library version reporting; only show if debugging.
402 Layout now matches that introduced for other libraries in 4.74 PP/03.
404 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
406 PP/04 New "dns_use_edns0" global option.
408 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
411 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
412 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
414 TK/01 Updated PolarSSL code to 0.14.2.
415 Bugzilla 1097. Patch from Andreas Metzler.
417 PP/07 Catch divide-by-zero in ${eval:...}.
420 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
423 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
424 format-string attack -- SECURITY: remote arbitrary code execution.
426 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
427 time unintentionally subject to list matching rules, letting the header
428 cause arbitrary Exim lookups (of items which can occur in lists, *not*
429 arbitrary string expansion). This allowed for information disclosure.
431 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
432 INT_MIN/-1 -- value coerced to INT_MAX.
438 NM/01 Workround for PCRE version dependency in version reporting
441 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
442 This fixes portability to compilers other than gcc, notably
443 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
445 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
446 makefiles for portability to HP-UX and POSIX correctness.
448 PP/01 Permit LOOKUP_foo enabling on the make command-line.
449 Also via indented variable definition in the Makefile.
450 (Debugging by Oliver Heesakkers).
452 PP/02 Restore caching of spamd results with expanded spamd_address.
453 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
455 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
456 Improves build reliability. Fix from: Frank Elsner
458 NM/02 Fix wide character breakage in the rfc2047 coding
459 Fixes bug 1064. Patch from Andrey N. Oktyabrski
461 NM/03 Allow underscore in dnslist lookups
462 Fixes bug 1026. Patch from Graeme Fowler
464 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
465 Code patches from Adam Ciarcinski of NetBSD.
467 NM/04 Fixed exiqgrep to cope with mailq missing size issue
470 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
471 is logged, to avoid truncation. Patch from John Horne.
473 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
474 Patch from Jakob Hirsch.
476 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
477 SQL string expansion failure details.
478 Patch from Andrey Oktyabrski.
480 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
481 Patch from Simon Arlott.
483 PP/09 New lookups functionality failed to compile on old gcc which rejects
484 extern declarations in function scope.
485 Patch from Oliver Fleischmann
487 PP/10 Use sig_atomic_t for flags set from signal handlers.
488 Check getgroups() return and improve debugging.
489 Fixed developed for diagnosis in bug 927 (which turned out to be
492 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
493 Patch from Mark Zealey.
495 PP/12 Bugzilla 1056: Improved spamd server selection.
496 Patch from Mark Zealey.
498 PP/13 Bugzilla 1086: Deal with maildir quota file races.
499 Based on patch from Heiko Schlittermann.
501 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
502 Patch from Uwe Doering, sign-off by Michael Haardt.
504 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
505 variable declaration deep within a block. Bug and patch from
508 PP/15 lookups-Makefile IRIX compatibilty coercion.
510 PP/16 Make DISABLE_DKIM build knob functional.
512 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
513 Patch by Simon Arlott
515 TF/03 Fix valgrind.h portability to C89 compilers that do not support
516 variable argument macros. Our copy now differs from upstream.
522 TF/01 Failure to get a lock on a hints database can have serious
523 consequences so log it to the panic log.
525 TF/02 Log LMTP confirmation messages in the same way as SMTP,
526 controlled using the smtp_confirmation log selector.
528 TF/03 Include the error message when we fail to unlink a spool file.
530 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
531 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
532 for maintaining out-of-tree patches for some time.
534 PP/01 Bugzilla 139: Documentation and portability issues.
535 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
536 Handle per-OS dynamic-module compilation flags.
538 PP/02 Let /dev/null have normal permissions.
539 The 4.73 fixes were a little too stringent and complained about the
540 permissions on /dev/null. Exempt it from some checks.
541 Reported by Andreas M. Kirchwitz.
543 PP/03 Report version information for many libraries, including
544 Exim version information for dynamically loaded libraries. Created
545 version.h, now support a version extension string for distributors
546 who patch heavily. Dynamic module ABI change.
548 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
549 privilege escalation vulnerability whereby the Exim run-time user
550 can cause root to append content of the attacker's choosing to
553 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
556 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
557 If dropping privileges for untrusted macros, we disabled normal logging
558 on the basis that it would fail; for the Exim run-time user, this is not
559 the case, and it resulted in successful deliveries going unlogged.
560 Fixed. Reported by Andreas Metzler.
566 PP/01 Date: & Message-Id: revert to normally being appended to a message,
567 only prepend for the Resent-* case. Fixes regression introduced in
568 Exim 4.70 by NM/22 for Bugzilla 607.
570 PP/02 Include check_rfc2047_length in configure.default because we're seeing
571 increasing numbers of administrators be bitten by this.
573 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
575 PP/03 Bugzilla 994: added openssl_options main configuration option.
577 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
579 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
581 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
583 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
584 without a peer certificate, leading to a segfault because of an
585 assumption that peers always have certificates. Be a little more
586 paranoid. Problem reported by Martin Tscholak.
588 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
589 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
590 NB: ClamAV planning to remove STREAM in "middle of 2010".
591 CL also introduces -bmalware, various -d+acl logging additions and
592 more caution in buffer sizes.
594 PP/09 Implemented reverse_ip expansion operator.
596 PP/10 Bugzilla 937: provide a "debug" ACL control.
598 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
600 PP/12 Bugzilla 973: Implement --version.
602 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
604 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
606 PP/15 Bugzilla 816: support multiple condition rules on Routers.
608 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
609 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
610 ignore trailing whitespace.
612 JJ/02 prevent non-panic DKIM error from being sent to paniclog
614 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
617 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
618 Notification from Dr Andrew Aitchison.
620 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
621 ExtendedDetectionInfo response format.
622 Notification from John Horne.
624 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
627 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
628 XSL and documented dependency on system catalogs, with examples of how
631 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
634 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
635 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
636 configuration file which is writeable by the Exim user or group.
638 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
639 of configuration files to cover files specified with the -C option if
640 they are going to be used with root privileges, not just the default
643 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
644 option (effectively making it always true).
646 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
647 files to be used while preserving root privileges.
649 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
650 that rogue child processes cannot use them.
652 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
653 run-time user, instead of root.
655 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
656 Exim run-time user without dropping privileges.
658 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
659 result string, instead of calling string_vformat() twice with the same
662 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
663 for other users. Others should always drop root privileges if they use
664 -C on the command line, even for a whitelisted configure file.
666 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
668 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
674 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
675 $data_path, and $header_path variables; fixed documentation bugs and
678 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
679 exipick to access non-standard spools, including the "frozen" queue
682 NM/01 Bugzilla 965: Support mysql stored procedures.
683 Patch from Alain Williams
685 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
687 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
688 Patch from Andreas Metzler
690 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
691 Patch from Kirill Miazine
693 NM/05 Bugzilla 671: Added umask to procmail example.
695 JJ/03 installed exipick 20100323.0, fixing doc bug
697 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
698 directory. Notification and patch from Dan Rosenberg.
700 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
702 TK/02 Improve log output when DKIM signing operation fails.
704 MH/01 Treat the transport option dkim_domain as a colon separated
705 list, not as a single string, and sign the message with each element,
706 omitting multiple occurences of the same signer.
708 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
709 Bugzilla 985, 986. Patch by Simon Arlott
711 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
712 Patch by Simon Arlott
714 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
715 MBX locking. Notification from Dan Rosenberg.
721 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
723 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
725 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
727 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
729 NM/04 Bugzilla 915: Flag broken perl installation during build.
735 TK/01 Added patch by Johannes Berg that expands the main option
736 "spamd_address" if it starts with a dollar sign.
738 TK/02 Write list of recipients to X-Envelope-Sender header when building
739 the mbox-format spool file for content scanning (suggested by Jakob
742 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
743 (http://www.dcc-servers.net/) support via dccifd. Activated by
744 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
746 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
747 by Mark Daniel Reidel <mr@df.eu>.
749 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
750 When building exim an external PCRE library is now needed -
751 PCRE is a system library on the majority of modern systems.
752 See entry on PCRE_LIBS in EDITME file.
754 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
755 conversation. Added nologin parameter to request.
756 Patch contributed by Kirill Miazine.
758 TF/01 Do not log submission mode rewrites if they do not change the address.
760 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
762 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
763 log files in place. Contributed by Roberto Lima.
765 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
767 TF/03 Bugzilla 615: When checking the local_parts router precondition
768 after a local_part_suffix or local_part_prefix option, Exim now
769 does not use the address's named list lookup cache, since this
770 contains cached lookups for the whole local part.
772 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
773 Robert Millan. Documentation is in experimental-spec.txt.
775 TF/04 Bugzilla 668: Fix parallel build (make -j).
777 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
779 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
780 Patch provided by Jan Srzednicki.
782 TF/05 Leading white space used to be stripped from $spam_report which
783 wrecked the formatting. Now it is preserved.
785 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
786 that they are available at delivery time.
788 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
790 TF/08 TLS error reporting now respects the incoming_interface and
791 incoming_port log selectors.
793 TF/09 Produce a more useful error message if an SMTP transport's hosts
794 setting expands to an empty string.
796 NM/06 Bugzilla 744: EXPN did not work under TLS.
797 Patch provided by Phil Pennock.
799 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
800 Patch provided by Richard Godbee.
802 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
803 acl_smtp_notquit, added index entry.
805 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
806 Patch provided by Eugene Bujak.
808 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
809 accept(). Patch provided by Maxim Dounin.
811 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
812 Patch provided by Phil Pennock.
814 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
816 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
817 Patch provided by Brad "anomie" Jorsch.
819 NM/14 Bugzilla 622: Added timeout setting to transport filter.
820 Patch provided by Dean Brooks.
822 TK/05 Add native DKIM support (does not depend on external libraries).
824 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
825 Patch provided by Graeme Fowler.
827 NM/16 Bugzilla 851: Documentation example syntax fix.
829 NM/17 Changed NOTICE file to remove references to embedded PCRE.
831 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
834 NM/19 Bugzilla 745: TLS version reporting.
835 Patch provided by Phil Pennock.
837 NM/20 Bugzilla 167: bool: condition support.
838 Patch provided by Phil Pennock.
840 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
841 clients. Patch provided by Phil Pennock.
843 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
844 Patch provided by Brad "anomie" Jorsch.
846 NM/23 Bugzilla 687: Fix misparses in eximstats.
847 Patch provided by Heiko Schlittermann.
849 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
850 Patch provided by Heiko Schlittermann.
852 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
853 plus update to original patch.
855 NM/26 Bugzilla 799: Documentation correction for ratelimit.
857 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
858 Patch provided by David Brownlee.
860 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
862 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
864 NM/30 Bugzilla 888: TLS documentation bugfixes.
866 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
868 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
869 Unlike the original bugzilla I have changed all shell scripts in src tree.
871 NM/33 Bugzilla 898: Transport filter timeout fix.
872 Patch by Todd Rinaldo.
874 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
875 Patch by Serge Demonchaux.
877 NM/35 Bugzilla 39: Base64 decode bug fixes.
878 Patch by Jakob Hirsch.
880 NM/36 Bugzilla 909: Correct connect() call in dcc code.
882 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
884 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
886 NM/39 Bugzilla 911: Fixed MakeLinks build script.
892 TK/01 Add preliminary DKIM support. Currently requires a forked version of
893 ALT-N's libdkim that I have put here:
894 http://duncanthrax.net/exim-experimental/
896 Note to Michael Haardt: I had to rename some vars in sieve.c. They
897 were called 'true' and it seems that C99 defines that as a reserved
898 keyword to be used with 'bool' variable types. That means you could
899 not include C99-style headers which use bools without triggering
900 build errors in sieve.c.
902 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
903 as mailq or other aliases. Changed the --help handling significantly
904 to do whats expected. exim_usage() emits usage/help information.
906 SC/01 Added the -bylocaldomain option to eximstats.
908 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
910 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
912 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
918 PH/01 Another patch from the Sieve maintainer.
920 PH/02 When an IPv6 address is converted to a string for single-key lookup
921 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
922 dots are used instead of colons so that keys in lsearch files need not
923 contain colons. This was done some time before quoting was made available
924 in lsearch files. However, iplsearch files do require colons in IPv6 keys
925 (notated using the quote facility) so as to distinguish them from IPv4
926 keys. This meant that lookups for IP addresses in host lists did not work
927 for iplsearch lookups.
929 This has been fixed by arranging for IPv6 addresses to be expressed with
930 colons if the lookup type is iplsearch. This is not incompatible, because
931 previously such lookups could never work.
933 The situation is now rather anomolous, since one *can* have colons in
934 ordinary lsearch keys. However, making the change in all cases is
935 incompatible and would probably break a number of configurations.
937 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
940 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
941 conversion specification without a maximum field width, thereby enabling
942 a rogue spamd server to cause a buffer overflow. While nobody in their
943 right mind would setup Exim to query an untrusted spamd server, an
944 attacker that gains access to a server running spamd could potentially
945 exploit this vulnerability to run arbitrary code as the Exim user.
947 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
948 $primary_hostname instead of what libspf2 thinks the hosts name is.
950 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
951 a directory entry by the name of the lookup key. Previously, if a
952 symlink pointed to a non-existing file or a file in a directory that
953 Exim lacked permissions to read, a lookup for a key matching that
954 symlink would fail. Now it is enough that a matching directory entry
955 exists, symlink or not. (Bugzilla 503.)
957 PH/03 The body_linecount and body_zerocount variables are now exported in the
960 PH/04 Added the $dnslist_matched variable.
962 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
963 This means they are set thereafter only if the connection becomes
966 PH/06 Added the client_condition to authenticators so that some can be skipped
967 by clients under certain conditions.
969 PH/07 The error message for a badly-placed control=no_multiline_responses left
970 "_responses" off the end of the name.
972 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
974 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
975 (without spaces) instead of just copying the configuration text.
977 PH/10 Added the /noupdate option to the ratelimit ACL condition.
979 PH/11 Added $max_received_linelength.
981 PH/12 Added +ignore_defer and +include_defer to host lists.
983 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
984 way in which PCRE > 7.0 is built.
986 PH/14 Implemented queue_only_load_latch.
988 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
989 MAIL command. The effect was to mangle the value on 64-bit systems.
991 PH/16 Another patch from the Sieve maintainer.
993 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
995 PH/18 If a system quota error occurred while trying to create the file for
996 a maildir delivery, the message "Mailbox is full" was not appended to the
997 bounce if the delivery eventually timed out. Change 4.67/27 below applied
998 only to a quota excession during the actual writing of the file.
1000 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1001 characters?) which causes problems in log lines. The DN values are now
1002 passed through string_printing() before being added to log lines.
1004 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1005 and InterBase are left for another time.)
1007 PH/21 Added message_body_newlines option.
1009 PH/22 Guard against possible overflow in moan_check_errorcopy().
1011 PH/23 POSIX allows open() to be a macro; guard against that.
1013 PH/24 If the recipient of an error message contained an @ in the local part
1014 (suitably quoted, of course), incorrect values were put in $domain and
1015 $local_part during the evaluation of errors_copy.
1021 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1022 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1025 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1026 issue a MAIL command.
1028 PH/02 In an ACL statement such as
1030 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1032 if a client was not listed at all, or was listed with a value other than
1033 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1034 the condition was not true (as it should be), so access was not denied.
1035 The bug was that the ! inversion was incorrectly passed on to the second
1036 item. This has been fixed.
1038 PH/03 Added additional dnslists conditions == and =& which are different from
1039 = and & when the dns lookup returns more than one IP address.
1041 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1042 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1044 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1045 FSYNC, which compiles an option called disable_fsync that allows for
1046 bypassing fsync(). The documentation is heavily laced with warnings.
1048 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1050 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1051 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1052 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1053 including adding "make clean"; (3) Added -fPIC when compiling the test
1054 dynamically loaded module, to get rid of a warning.
1056 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1057 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1058 = 0s. The bug is otherwise harmless.
1060 PH/07 There was a bug in the dovecot authenticator such that the value of
1061 $auth1 could be overwritten, and so not correctly preserved, after a
1062 successful authentication. This usually meant that the value preserved by
1063 the server_setid option was incorrect.
1065 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1067 PH/09 Installed PCRE release 7.0.
1069 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1070 run for batched SMTP input. It is now run at the start of every message
1071 in the batch. While fixing this I discovered that the process information
1072 (output by running exiwhat) was not always getting set for -bs and -bS
1073 input. This is fixed, and it now also says "batched" for BSMTP.
1075 PH/11 Added control=no_pipelining.
1077 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1078 patch, slightly modified), and move the expansion of helo_data till after
1079 the connection is made in the smtp transport (so it can use these
1082 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1084 PH/14 Added log_selector = +pid.
1086 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1088 PH/16 Add ${if forany and ${if forall.
1090 PH/17 Added dsn_from option to vary the From: line in DSNs.
1092 PH/18 Flush SMTP output before performing a callout, unless control =
1093 no_callout_flush is set.
1095 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1096 was true (the default) a successful delivery failed to delete the retry
1097 item, thus causing premature timeout of the address. The bug is now
1100 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1102 PH/21 Long custom messages for fakedefer and fakereject are now split up
1103 into multiline reponses in the same way that messages for "deny" and
1104 other ACL rejections are.
1106 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1107 with slight modification.
1109 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1110 draft, changing the syntax and factoring some duplicate code".
1112 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1113 for deliveries of the second and subsequent messages over the same SMTP
1116 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1117 ${reduce, with only minor "tidies".
1119 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1121 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1122 expansion side effects.
1124 PH/27 When a message times out after an over-quota error from an Exim-imposed
1125 quota, the bounce message says "mailbox is full". This message was not
1126 being given when it was a system quota that was exceeded. It now should
1129 MH/03 Made $recipients available in local_scan(). local_scan() already has
1130 better access to the recipient list through recipients_list[], but
1131 $recipients can be useful in postmaster-provided expansion strings.
1133 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1134 in the case of a MAIL command with additional options following the
1135 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1136 were accidentally chopped off.
1138 PH/29 SMTP synchronization checks are implemented when a command is read -
1139 there is a check that no more input is waiting when there shouldn't be
1140 any. However, for some commands, a delay in an ACL can mean that it is
1141 some time before the response is written. In this time, more input might
1142 arrive, invalidly. So now there are extra checks after an ACL has run for
1143 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1144 pipelining has not been advertised.
1146 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1148 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1149 returned in the SMTP response when smtp_return_error_details was false.
1150 This has been fixed.
1152 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1153 instead of the C I/O that was originally supplied, because problems were
1154 reported on Solaris.
1156 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1157 Exim which did not show up earlier: it was assuming that a call to
1158 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1159 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1160 was a macro that became an assignment, so it seemed to work. This has
1161 changed to a proper function call with a void return, hence the compile
1162 error. Exim's code has been fixed.
1164 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1167 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1169 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1171 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1172 because it turns out to be too big or there is a timeout) there is no
1173 "Completed" line in the log. When some messages of this type were
1174 selected by exigrep, they were listed as "not completed". Others were
1175 picked up by some special patterns. I have improved the selection
1176 criteria to be more general.
1178 PH/38 The host_find_failed option in the manualroute router can now be set
1179 to "ignore", to completely ignore a host whose IP address cannot be
1180 found. If all hosts are ignored, the behaviour is controlled by the new
1181 host_all_ignored option.
1183 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1184 homing or because of multiple MX records with /mx) generated more than
1185 one IP address, and the following item turned out to be the local host,
1186 all the secondary addresses of the first item were incorrectly removed
1187 from the list, along with the local host and any following hosts (which
1188 is what is supposed to happen).
1190 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1191 whoever called Exim into the -H file. In the case of the daemon it was
1192 behaving confusingly. When first started, it used values for whoever
1193 started the daemon, but after a SIGHUP it used the Exim user (because it
1194 calls itself on a restart). I have changed the code so that it now always
1197 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1198 message are rejected with the same error (e.g. no authentication or bad
1199 sender address), and a DATA command is nevertheless sent (as can happen
1200 with PIPELINING or a stupid MUA), the error message that was given to the
1201 RCPT commands is included in the rejection of the DATA command. This is
1202 intended to be helpful for MUAs that show only the final error to their
1205 PH/42 Another patch from the Sieve maintainer.
1207 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1208 Eximstats - Fixed some broken HTML links and added missing column headers
1210 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1211 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1213 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1215 PH/43 Yet another patch from the Sieve maintainer.
1217 PH/44 I found a way to check for a TCP/IP connection going away before sending
1218 the response to the final '.' that terminates a message, but only in the
1219 case where the client has not sent further data following the '.'
1220 (unfortunately, this is allowed). However, in many cases there won't be
1221 any further data because there won't be any more messages to send. A call
1222 to select() can be used: if it shows that the input is "ready", there is
1223 either input waiting, or the socket has been closed. An attempt to read
1224 the next input character can distinguish the two cases. Previously, Exim
1225 would have sent an OK response which the client would never have see.
1226 This could lead to message repetition. This fix should cure that, at
1227 least in a lot of common cases.
1229 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1230 advertised in response to EHLO.
1236 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1237 fixed by 4.65/MH/01 (is this a record?) are fixed:
1239 (i) An empty string was always treated as zero by the numeric comparison
1240 operators. This behaviour has been restored.
1242 (ii) It is documented that the numeric comparison operators always treat
1243 their arguments as decimal numbers. This was broken in that numbers
1244 starting with 0 were being interpreted as octal.
1246 While fixing these problems I realized that there was another issue that
1247 hadn't been noticed. Values of message_size_limit (both the global option
1248 and the transport option) were treated as octal if they started with 0.
1249 The documentation was vague. These values are now always treated as
1250 decimal, and I will make that clear in the documentation.
1256 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1257 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1260 MH/01 Don't check that the operands of numeric comparison operators are
1261 integers when their expansion is in "skipping" mode (fixes bug
1262 introduced by 4.64-PH/07).
1264 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1265 child addresses, Exim now panics and dies. Previously, because the count
1266 is held in a short int, deliveries were likely to be lost. As such a
1267 large number of recipients for a single message is ridiculous
1268 (performance will be very, very poor), I have chosen to impose a limit
1269 rather than extend the field.
1275 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1276 leftover -K file (the existence of which was triggered by #402).
1277 While we were at it, introduced process PID as part of the -K
1278 filename. This should rule out race conditions when creating
1281 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1282 processing considerably. Previous code took too long for large mails,
1283 triggering a timeout which in turn triggers #401.
1285 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1286 in the DK code in transports.c. sendfile() is not really portable,
1287 hence the _LINUX specificness.
1289 TF/01 In the add_headers option to the mail command in an Exim filter,
1290 there was a bug that Exim would claim a syntax error in any
1291 header after the first one which had an odd number of characters
1294 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1295 callout verification, Exim cached a "reject" for the entire domain. This
1296 is correct for most verifications, but it is not correct for a recipient
1297 verification with use_sender or use_postmaster set, because in that case
1298 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1299 case of MAIL FROM:<> rejection from other early rejections (e.g.
1300 rejection of HELO). When verifying a recipient using a non-null MAIL
1301 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1302 Whatever the result of the callout, the value of the domain cache is
1303 left unchanged (for any other kind of callout, getting as far as trying
1304 RCPT means that the domain itself is ok).
1306 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1309 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1310 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1311 session, the daemon ignores ECONNECT errors and logs others; it now
1312 ignores EPIPE as well.
1314 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1315 (quoted-printable decoding).
1317 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1318 later the small subsequent patch to fix an introduced bug.
1320 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1322 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1324 PH/08 An error is now given if message_size_limit is specified negative.
1326 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1327 to be given (somewhat) arbitrary names.
1329 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1332 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1333 miscellaneous code fixes
1335 PH/10 Added the log_reject_target ACL modifier to specify where to log
1338 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1339 hostname. This is wrong, because it relates to the incoming message (and
1340 probably the interface on which it is arriving) and not to the outgoing
1341 callout (which could be using a different interface). This has been
1342 changed to use the value of the helo_data option from the smtp transport
1343 instead - this is what is used when a message is actually being sent. If
1344 there is no remote transport (possible with a router that sets up host
1345 addresses), $smtp_active_hostname is used.
1347 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1348 tweaks were necessary in order to get it to work (see also 21 below):
1349 (a) The code assumed that strncpy() returns a negative number on buffer
1350 overflow, which isn't the case. Replaced with Exim's string_format()
1352 (b) There were several signed/unsigned issues. I just did the minimum
1353 hacking in of casts. There is scope for a larger refactoring.
1354 (c) The code used strcasecmp() which is not a standard C function.
1355 Replaced with Exim's strcmpic() function.
1356 (d) The code set only $1; it now sets $auth1 as well.
1357 (e) A simple test gave the error "authentication client didn't specify
1358 service in request". It would seem that Dovecot has changed its
1359 interface. Fortunately there's a specification; I followed it and
1360 changed what the client sends and it appears to be working now.
1362 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1365 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1366 address A is aliased to B and C, where B exists and C does not. Without
1367 -v the output is "A verified" because verification stops after a
1368 successful redirection if more than one address is generated. However,
1369 with -v the child addresses are also verified. Exim was outputting "A
1370 failed to verify" and then showing the successful verification for C,
1371 with its parentage. It now outputs "B failed to verify", showing B's
1372 parentage before showing the successful verification of C.
1374 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1375 look up a TXT record in a specific list after matching in a combined
1378 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1379 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1380 they consult the DNS. I had assumed they would set it the way they
1381 wanted; and indeed my experiments on Linux seem to show that in some
1382 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1383 To be on the safe side, however, I have now made the interface to
1384 host_find_byname() similar to host_find_bydns(), with an argument
1385 containing the DNS resolver options. The host_find_byname() function now
1386 sets these options at its start, just as host_find_bydns() does. The smtp
1387 transport options dns_qualify_single and dns_search_parents are passed to
1388 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1389 of host_find_byname() use the default settings of RES_DEFNAMES
1390 (qualify_single) but not RES_DNSRCH (search_parents).
1392 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1393 spool_read_header() do less string testing, by means of a preliminary
1394 switch on the second character of optional "-foo" lines. (This is
1395 overdue, caused by the large number of possibilities that now exist.
1396 Originally there were few.) While I was there, I also converted the
1397 str(n)cmp tests so they don't re-test the leading "-" and the first
1398 character, in the hope this might squeeze out yet more improvement.
1400 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1401 flag allowing group syntax was set by the header_syntax check but not
1402 turned off, possible causing trouble later; (2) The flag was not being
1403 set at all for the header_verify test, causing "group"-style headers to
1404 be rejected. I have now set it in this case, and also caused header_
1405 verify to ignore an empty address taken from a group. While doing this, I
1406 came across some other cases where the code for allowing group syntax
1407 while scanning a header line wasn't quite right (mostly, not resetting
1408 the flag correctly in the right place). These bugs could have caused
1409 trouble for malformed header lines. I hope it is now all correct.
1411 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1412 with the "reply" argument non-NULL. The code, however (which originally
1413 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1414 but it didn't always do it. This confused somebody who was copying the
1415 code for some other use. I have removed all the tests.
1417 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1418 feature that was used to support insecure browsers during the U.S. crypto
1419 embargo. It requires special client support, and Exim is probably the
1420 only MTA that supported it -- and would never use it because real RSA is
1421 always available. This code has been removed, because it had the bad
1422 effect of slowing Exim down by computing (never used) parameters for the
1423 RSA_EXPORT functionality.
1425 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1426 authenticator to fail if there's a tab character in the incoming data
1427 (there should never be unless someone is messing about, as it's supposed
1428 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1429 if the connection is using TLS or if the remote IP is the same as the
1430 local IP, and the "valid-client-cert option" if a client certificate has
1433 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1434 authenticators. This can be used for authorization after authentication
1435 succeeds. (In the case of plaintext, it servers for both authentication
1438 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1439 if any retry times were supplied.
1441 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1442 connection, where there is no client IP address to check. In this
1443 situation, the verify now always succeeds.
1445 PH/25 Applied John Jetmore's -Mset patch.
1447 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1449 PH/27 In a string expansion for a processed (not raw) header when multiple
1450 headers of the same name were present, leading whitespace was being
1451 removed from all of them, but trailing whitespace was being removed only
1452 from the last one. Now trailing whitespace is removed from each header
1453 before concatenation. Completely empty headers in a concatenation (as
1454 before) are ignored.
1456 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1457 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1459 PH/29 [Removed. This was a change that I later backed out, and forgot to
1460 correct the ChangeLog entry (that I had efficiently created) before
1461 committing the later change.]
1463 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1464 address errors (4xx response to RCPT) over the same connection as other
1465 messages routed to the same hosts. Such deliveries are always "forced",
1466 so retry times are not inspected. This resulted in far too many retries
1467 for the affected addresses. The effect occurred only when there were more
1468 hosts than the hosts_max_try setting in the smtp transport when it had
1469 the 4xx errors. Those hosts that it had tried were not added to the list
1470 of hosts for which the message was waiting, so if all were tried, there
1471 was no problem. Two fixes have been applied:
1473 (i) If there are any address or message errors in an SMTP delivery, none
1474 of the hosts (tried or untried) are now added to the list of hosts
1475 for which the message is waiting, so the message should not be a
1476 candidate for sending over the same connection that was used for a
1477 successful delivery of some other message. This seems entirely
1478 reasonable: after all the message is NOT "waiting for some host".
1479 This is so "obvious" that I'm not sure why it wasn't done
1480 previously. Hope I haven't missed anything, but it can't do any
1481 harm, as the worst effect is to miss an optimization.
1483 (ii) If, despite (i), such a delivery is accidentally attempted, the
1484 routing retry time is respected, so at least it doesn't keep
1485 hammering the server.
1487 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1488 in ${readsocket because some servers need this prod.
1490 PH/32 Added some extra debug output when updating a wait-xxx database.
1492 PH/33 The hint "could be header name not terminated by colon", which has been
1493 given for certain expansion errors for a long time, was not being given
1494 for the ${if def:h_colon_omitted{... case.
1496 PH/34 The spec says: "With one important exception, whenever a domain list is
1497 being scanned, $domain contains the subject domain." There was at least
1498 one case where this was not true.
1500 PH/35 The error "getsockname() failed: connection reset by peer" was being
1501 written to the panic log as well as the main log, but it isn't really
1502 panic-worthy as it just means the connection died rather early on. I have
1503 removed the panic log writing for the ECONNRESET error when getsockname()
1506 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1507 runs only) independently of the message's sender address. This meant
1508 that, if the 4xx error was in fact related to the sender, a different
1509 message to the same recipient with a different sender could confuse
1510 things. In particualar, this can happen when sending to a greylisting
1511 server, but other circumstances could also provoke similar problems.
1512 I have changed the default so that the retry time for these errors is now
1513 based a combination of the sender and recipient addresses. This change
1514 can be overridden by setting address_retry_include_sender=false in the
1517 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1518 remote server are returned as part of bounce messages. This was not
1519 happening for LMTP over a pipe (the lmtp transport), but now it is the
1520 same for both kinds of LMTP.
1522 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1523 in header lines that were in fact CNAMEs. This is no longer the case.
1525 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1526 and queue runs started by the daemon processed all messages. This has
1527 been fixed so that -R and -S can now usefully be given with -q<time>.
1529 PH/40 Import PCRE release 6.7 (fixes some bugs).
1531 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1533 PH/42 Give an error if -q is specified more than once.
1535 PH/43 Renamed the variables $interface_address and $interface_port as
1536 $received_ip_address and $received_port, to make it clear that these
1537 values apply to message reception, and not to the outgoing interface when
1538 a message is delivered. (The old names remain recognized, of course.)
1540 PH/44 There was no timeout on the connect() call when using a Unix domain
1541 socket in the ${readsocket expansion. There now is.
1543 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1544 be meaningful with "accept".
1546 SC/01 Eximstats V1.43
1547 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1549 SC/02 Eximstats V1.44
1550 Use a glob alias rather than an array ref in the generated
1551 parser. This improves both readability and performance.
1553 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1554 Collect SpamAssassin and rejection statistics.
1555 Don't display local sender or destination tables unless
1556 there is data to show.
1557 Added average volumes into the top table text output.
1559 SC/04 Eximstats V1.46
1560 Collect data on the number of addresses (recipients)
1561 as well as the number of messages.
1563 SC/05 Eximstats V1.47
1564 Added 'Message too big' to the list of mail rejection
1565 reasons (thanks to Marco Gaiarin).
1567 SC/06 Eximstats V1.48
1568 Mainlog lines which have GMT offsets and are too short to
1569 have a flag are now skipped.
1571 SC/07 Eximstats V1.49 (Alain Williams)
1572 Added the -emptyok flag.
1574 SC/08 Eximstats V1.50
1575 Fixes for obtaining the IP address from reject messages.
1577 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1578 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1579 whitesspace changes from 4.64-PH/27
1581 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1584 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1585 are found, allow negative numbers in numeric criteria)
1587 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1589 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1590 to match changes made in 4.64-PH/43
1592 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1594 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1595 "domain" is actually an IP address, and force a failure. This locks out
1596 those revolvers/nameservers that support "A-for-A" lookups, in
1597 contravention of the specifications.
1599 PH/48 When a host name was looked up from an IP address, and the subsequent
1600 forward lookup of the name timed out, the host name was left in
1601 $sender_host_name, contrary to the specification.
1603 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1604 restricted to single-key lookups, Exim was not diagnosing an error if
1605 * or *@ was used with a query-style lookup.
1607 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1609 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1610 long ago, but noone interested enough thought of it. Let's just say that
1611 the "1.1" means that there are some new functions that weren't there at
1612 some point in the past.
1614 PH/51 Error processing for expansion failure of helo_data from an smtp
1615 transport during callout processing was broken.
1617 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1618 tested/used via the -bh/-bhc/-bs options.
1620 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1621 bug, fixed in subsequent PCRE releases).
1623 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1624 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1626 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1632 SC/01 Use a glob alias rather than an array ref in eximstats generated
1633 parser. This improves both readability and performance.
1635 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1636 Don't display local sender or destination tables in eximstats unless
1637 there is data to show.
1638 Added average volumes into the eximstats top table text output.
1640 SC/03 Collect data on the number of addresses (recipients) as well
1641 as the number of messages in eximstats.
1643 TF/01 Correct an error in the documentation for the redirect router. Exim
1644 does (usually) call initgroups() when daemonizing.
1646 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1647 with consistent privilege compared to when running as a daemon.
1649 TF/03 Note in the spec that $authenticated_id is not set for local
1650 submissions from trusted users.
1652 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1653 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1655 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1656 by adding some example configuration directives to the default
1657 configuration file. A little bit of work is required to uncomment the
1658 directives and define how usernames and passwords are checked, but
1659 there is now a framework to start from.
1661 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1662 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1663 without this. I don't know how relevant this is to other LDAP libraries.
1665 PH/02 Add the verb name to the "unknown ACL verb" error.
1667 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1669 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1671 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1672 directory not expanded when it should be if an expanded home directory
1673 was set for the address (which is overridden by the transport).
1675 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1678 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1679 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1680 because it is too late at that time, and has no effect.
1682 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1683 security issue with \' (bugzilla #107). I could not use the
1684 PQescapeStringConn() function, because it needs a PGconn value as one of
1687 PH/08 When testing addresses using -bt, indicate those final addresses that
1688 are duplicates that would not cause an additional delivery. At least one
1689 person was confused, thinking that -bt output corresponded to deliveries.
1690 (Suppressing duplicates isn't a good idea as you lose the information
1691 about possibly different redirections that led to the duplicates.)
1693 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1694 systems where poll() doesn't work, in particular OS X.
1696 PH/10 Added more information to debugging output for retry time not reached.
1698 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1699 operations in malware.c.
1701 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1704 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1705 syslog_duplication set false, log lines that would normally be written
1706 both the the main log and to the reject log were not written to syslog at
1709 PH/14 In the default configuration, change the use of "message" in ACL warn
1710 statements to "add_header".
1712 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1713 not followed by a command (e.g. "seen endif").
1715 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1716 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1719 PH/17 Added extra conditions to the default value of delay_warning_condition
1723 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1724 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1725 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1728 The Auto-Submitted: and various List- headers are standardised, whereas I
1729 don't think Precedence: ever was.
1731 PH/18 Refactored debugging code in route_finduser() to show more information,
1732 in particular, the error code if getpwnam() issues one.
1734 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1735 This is apparently needed in addition to the PH/07 change above to avoid
1736 any possible encoding problems.
1738 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1739 but not after initializing Perl.
1741 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1742 output them only if debugging. By default they are written stderr,
1743 apparently, which is not desirable.
1745 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1748 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1751 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1753 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1754 authenticated or an ident call has been made. Suppress the default
1755 values for $authenticated_id and $authenticated_sender (but permit -oMai
1756 and -oMas) when testing with -bh.
1758 PH/24 Re-jigged the order of the tests in the default configuration so that the
1759 tests for valid domains and recipients precede the DNS black list and CSA
1760 tests, on the grounds that those ones are more expensive.
1762 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1763 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1764 This bug exists in every version of Exim that I still have, right back to
1767 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1768 However, an attempt to turn on case-sensitivity in a regex key by
1769 including (?-i) didn't work because the subject string was already
1770 lowercased, and the effects were non-intuitive. It turns out that a
1771 one-line patch can be used to allow (?-i) to work as expected.
1777 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1778 other effects) broke the use of negated acl sub-conditions.
1780 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1783 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1784 "Deny" causes Exim to reject the incoming connection with a 554 error.
1785 Unfortunately, if there is a major crisis, such as a disk failure,
1786 tcp-wrappers gives "deny", whereas what one would like would be some
1787 kind of temporary error. A kludge has been added to help with this.
1788 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1789 554 error is used if errno is still zero or contains ENOENT (which occurs
1790 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1793 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1795 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1796 errors. Otherwise a message that provokes a temporary error (when other
1797 messages do not) can cause a whole host to time out.
1799 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1800 addresses were routed directly to files or pipes from a redirect router.
1801 File deliveries just didn't batch; pipe deliveries might have suffered
1804 PH/06 A failure to get a lock for a hints database would erroneously always say
1805 "Failed to get write lock", even when it was really a read lock.
1807 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1808 of 0600. This has been changed to use the value of the lockfile_mode
1809 option (which defaults to 0600).
1811 PH/08 Applied small patch from the Sieve maintainer.
1813 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1814 folder from quota calculations, a direct delivery into this folder messed
1815 up the contents of the maildirsize file. This was because the regex was
1816 used only to exclude .Trash (or whatever) when the size of the mailbox
1817 was calculated. There was no check that a delivery was happening into an
1818 excluded directory. This bug has been fixed by ignoring all quota
1819 processing for deliveries into excluded directories.
1821 PH/10 Added the maildirfolder_create_regex option to appendfile.
1827 PH/01 The code for finding all the local interface addresses on a FreeBSD
1828 system running IPv6 was broken. This may well have applied to all BSD
1829 systems, as well as to others that have similar system calls. The broken
1830 code found IPv4 interfaces correctly, but gave incorrect values for the
1831 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1832 that it would not match correctly against @[] and not recognize the IPv6
1835 PH/02 The ipliteral router was not recognizing addresses of the form user@
1836 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1838 PH/03 Added disable_ipv6.
1840 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1841 decoded form, because it is most often used to construct To: headers
1842 lines in autoreplies, and the decoded form may well be syntactically
1843 invalid. However, $reply_address has leading white space removed, and all
1844 newlines turned into spaces so that the autoreply transport does not
1847 PH/05 If group was specified without a user on a router, and no group or user
1848 was specified on a transport, the group from the router was ignored.
1850 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1851 for visible compile-time settings that can be used to change these
1852 numbers, for those that want even more. Backwards compatibility with old
1853 spool files has been maintained. However, going back to a previous Exim
1854 release will lost any variables that are in spool files.
1856 PH/07 Two small changes when running in the test harness: increase delay when
1857 passing a TCP/IP connection to a new process, in case the original
1858 process has to generate a bounce, and remove special handling of
1859 127.0.0.2 (sic), which is no longer necessary.
1861 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1862 be the same on different OS.
1864 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1867 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1868 whether --show-vars was specified or not
1870 JJ/02 exipick: Added support for new ACL variable spool format introduced
1873 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1874 syntactically invalid From: or Reply-to: line, and a filter used this to
1875 generate an autoreply, and therefore failed to obtain an address for the
1876 autoreply, Exim could try to deliver to a non-existent relative file
1877 name, causing unrelated and misleading errors. What now happens is that
1878 it logs this as a hard delivery error, but does not attempt to create a
1881 PH/11 The exinext utility has a -C option for testing purposes, but although
1882 the given file was scanned by exinext itself; it wasn't being passed on
1883 when Exim was called.
1885 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1886 an end-of-file indication when reading a command response.
1888 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1889 compiled. In many other places in Exim, IPv6 addresses are always
1890 recognized, so I have changed this. It also means that IPv4 domain
1891 literals of the form [IPV4:n.n.n.n] are now always recognized.
1893 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1894 used if the router is not running as root, for example, when verifying at
1895 ACL time, or when using -bh. The debugging output from this situation was
1896 non-existent - all you got was a failure to exec. I have made two
1899 (a) Failures to set uid/gid, the current directory, or a process leader
1900 in a subprocess such as that created by queryprogram now generate
1901 suitable debugging ouput when -d is set.
1903 (b) The queryprogram router detects when it is not running as root,
1904 outputs suitable debugging information if -d is set, and then runs
1905 the subprocess without attempting to change uid/gid.
1907 PH/15 Minor change to Makefile for building test_host (undocumented testing
1910 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1911 additional section of a DNS packet that returns MX or SRV records.
1912 Instead, it always explicitly searches for A/AAAA records. This avoids
1913 major problems that occur when a DNS server includes only records of one
1914 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1915 fixed another bug: if SRV records were looked up and the corresponding
1916 address records were *not* found in the additional section, the port
1917 values from the SRV records were lost.
1919 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1920 using the correct key (the original address) when searching the retry
1921 rules in order to find which one to use for generating the retry hint.
1923 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1924 adding the default one. Similarly, if it contains a Reply-To: header, the
1925 errors_reply_to option, if set, is not used.
1927 PH/19 When calculating a retry time, Exim used to measure the "time since
1928 failure" by looking at the "first failed" field in the retry record. Now
1929 it does not use this if it is later than than the arrival time of the
1930 message. Instead it uses the arrival time. This makes for better
1931 behaviour in cases where some deliveries succeed, thus re-setting the
1932 "first failed" field. An example is a quota failure for a huge message
1933 when small messages continue to be delivered. Without this change, the
1934 "time since failure" will always be short, possible causing more frequent
1935 delivery attempts for the huge message than are intended.
1936 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1938 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1939 $1, $2, $3) because the numerical variables can be reset during some
1940 expansion items (e.g. "match"), thereby losing the authentication data.
1942 PH/21 Make -bV show the size of off_t variables so that the test suite can
1943 decide whether to run tests for quotas > 2G.
1945 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1946 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1947 filecount value is greater than 2G or if a quota value is greater than 2G
1948 on a system where the size of off_t is not greater than 4, a panic error
1951 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1952 never match. The debug and -bh output now contains an explicit error
1953 message indicating a malformed IPv4 address or mask.
1955 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1956 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1957 PH/23 above applies.
1959 PH/25 Do not write to syslog when running in the test harness. The only
1960 occasion when this arises is a failure to open the main or panic logs
1961 (for which there is an explicit test).
1963 PH/26 Added the /no_tell option to "control=freeze".
1965 PH/27 If a host name lookup failed very early in a connection, for example, if
1966 the IP address matched host_lookup and the reverse lookup yielded a name
1967 that did not have a forward lookup, an error message of the form "no IP
1968 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1969 could be logged. Now it outputs the IP address instead of "NULL".
1971 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1972 allows the sender and the authenticated sender to be set when
1973 submitting a message from within Exim. Since child_open_exim() is
1974 documented for local_scan(), the new function should be too.
1976 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1977 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1978 results in an empty string is now treated as unset.
1980 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1982 PH/31 Added sender_verify_fail logging option.
1984 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1985 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1986 tidied the source and removed it altogether.
1988 PH/33 When a queue run was abandoned because the load average was too high, a
1989 log line was always written; now it is written only if the queue_run log
1990 selector is set. In addition, the log line for abandonment now contains
1991 information about the queue run such as the pid. This is always present
1992 in "start" and "stop" lines but was omitted from the "abandon" line.
1994 PH/34 Omit spaces between a header name and the colon in the error message that
1995 is given when verify = headers_syntax fails (if there are lots of them,
1996 the message gets confusing).
1998 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1999 names, as there are now some PTR records that contain slashes. This check
2000 is only to protect against broken name servers that fall over on strange
2001 characters, so the fact that it applies to all lookups doesn't matter.
2003 PH/36 Now that the new test suite is complete, we can remove some of the
2004 special code in Exim that was needed for the old test suite. For example,
2005 sorting DNS records because real resolvers return them in an arbitrary
2006 order. The new test suite's fake resolver always returns records in the
2009 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2010 bounces) except when queue_only is set, to avoid logging races between
2011 the different processes.
2013 PH/38 Panic-die if .include specifies a non-absolute path.
2015 PH/39 A tweak to the "H" retry rule from its user.
2017 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2018 a label. They prevented compilation on older perls.
2020 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2021 a warning to be raised on newish perls.
2023 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2024 on queue. Changes to match documented behaviour of showing count of
2025 messages matching specified criteria.
2027 PH/40 Changed the default ident timeout from 30s to 5s.
2029 PH/41 Added support for the use of login_cap features, on those BSD systems
2030 that have them, for controlling the resources used by pipe deliveries.
2032 PH/42 The content-scanning code uses fopen() to create files in which to put
2033 message data. Previously it was not paying any attention to the mode of
2034 the files. Exim runs with umask(0) because the rest of the code creates
2035 files with open(), and sets the required mode explicitly. Thus, these
2036 files were ending up world-writeable. This was not a big issue, because,
2037 being within the spool directory, they were not world-accessible. I have
2038 created a function called modefopen, which takes an additional mode
2039 argument. It sets umask(777), creates the file, chmods it to the required
2040 mode, then resets the umask. All the relevant calls to fopen() in the
2041 content scanning code have been changed to use this function.
2043 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2044 to 24 hours. This avoids potential overflow problems when processing G
2045 and H retry rules. I suspect nobody ever tinkers with this value.
2047 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2049 PH/45 When the plaintext authenticator is running as a client, the server's
2050 challenges are checked to ensure they are valid base64 strings. By
2051 default, the authentication attempt is cancelled if an invalid string is
2052 received. Setting client_ignore_invalid_base64 true ignores these errors.
2053 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2054 they are received. Thus, the responses can be made to depend on the
2055 challenges. If an invalid string is ignored, an empty string is placed in
2058 PH/46 Messages that are created by the autoreply transport now contains a
2059 References: header, in accordance with RFCs 2822 and 3834.
2061 PH/47 Added authenticated_sender_force to the smtp transport.
2063 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2065 PH/49 Installed latest patch from the Sieve maintainer.
2067 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2068 was also set, the appendfile transport was unnecessarily scanning a
2069 directory of message files (e.g. for maildir delivery) to find the count
2070 of files (along with the size), even though it did not need this
2071 information. It now does the scan only if it needs to find either the
2072 size of the count of files.
2074 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2076 PH/52 Two bugs concerned with error handling when the smtp transport is
2079 (i) Exim was not creating retry information for temporary errors given
2080 for individual recipients after the DATA command when the smtp transport
2081 was used in LMTP mode. This meant that they could be retried too
2082 frequently, and not timed out correctly.
2084 (ii) Exim was setting the flag that allows error details to be returned
2085 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2086 recipients that were returned after the DATA command.
2088 PH/53 This is related to PH/52, but is more general: for any failing address,
2089 when detailed error information was permitted to be returned to the
2090 sender, but the error was temporary, then after the final timeout, only
2091 "retry timeout exceeded" was returned. Now it returns the full error as
2092 well as "retry timeout exceeded".
2094 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2095 do this, and (what is worse) MTAs that accept it.
2097 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2098 will now be deprecated.
2100 PH/56 New os.c-cygwin from the Cygwin maintainer.
2102 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2103 formats (previously only available in exim formats via -bpr, -bpru,
2104 and -bpra. Now also available in native and exiqgrep formats)
2106 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2107 with very large, slow to parse queues
2109 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2111 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2113 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2114 responses to authentication challenges, though it was showing the
2115 challenges; (ii) I've removed the CR characters from the debug output for
2118 PH/58 Allow for the insertion of a newline as well as a space when a string
2119 is turned into more than one encoded-word during RFC 2047 encoding. The
2120 Sieve code now uses this.
2122 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2123 data_4xx, lost_connection, tls_required.
2125 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2126 message was being sent as an SMTP response.
2128 PH/61 Add -l and -k options to exicyclog.
2130 PH/62 When verifying, if an address was redirected to one new address, so that
2131 verification continued, and the new address failed or deferred after
2132 having set something in $address_data, the value of $address_data was not
2133 passed back to the ACL. This was different to the case when no
2134 redirection occurred. The value is now passed back in both cases.
2136 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2137 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2138 use login_cap.h, so on its own it isn't the distinguishing feature. The
2139 new name refers directly to the setclassresources() function.
2141 PH/65 Added configuration files for NetBSD3.
2143 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2145 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2146 is preferred over IPv4.
2148 PH/68 The bounce_return_message and bounce_return_body options were not being
2149 honoured for bounces generated during the reception of non-SMTP messages.
2150 In particular, this applied to messages rejected by the ACL. This bug has
2151 been fixed. However, if bounce_return_message is true and bounce_return_
2152 body is false, the headers that are returned for a non-SMTP message
2153 include only those that have been read before the error was detected.
2154 (In the case of an ACL rejection, they have all been read.)
2156 PH/69 The HTML version of the specification is now built in a directory called
2157 spec_html instead of spec.html, because the latter looks like a path with
2158 a MIME-type, and this confuses some software.
2160 PH/70 Catch two compiler warnings in sieve.c.
2162 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2163 function verify_get_ident() calls ip_connect() to connect a socket, but
2164 if the "connect()" function timed out, ip_connect() used to close the
2165 socket. However, verify_get_ident() also closes the socket later, and in
2166 between Exim writes to the log, which may get opened at this point. When
2167 the socket was closed in ip_connect(), the log could get the same file
2168 descriptor number as the socket. This naturally causes chaos. The fix is
2169 not to close the socket in ip_connect(); the socket should be closed by
2170 the function that creates it. There was only one place in the code where
2171 this was missing, in the iplookup router, which I don't think anybody now
2172 uses, but I've fixed it anyway.
2174 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2175 well as to direct DNS lookups. Otherwise the handling of names in host
2176 lists is inconsistent and therefore confusing.
2182 PH/01 Two changes to the default runtime configuration:
2184 (1) Move the checks for relay_from_hosts and authenticated clients from
2185 after to before the (commented out) DNS black list checks.
2187 (2) Add control=submission to the relay_from_hosts and authenticated
2188 clients checks, on the grounds that messages accepted by these
2189 statements are most likely to be submissions.
2191 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2193 (1) Generate an error if the third argument for the ${prvs expansion is
2196 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2199 (3) Reset the variables that are obtained from the first argument of
2200 ${prvscheck and used in the second argument before leaving the code,
2201 because their memory is reclaimed, so using them afterwards may do
2204 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2205 one (it's much easier than Tom thought :-).
2207 (5) Because of (4), we can now allow for the use of $prvscheck_result
2208 inside the third argument.
2210 PH/03 For some reason, the default setting of PATH when running a command from
2211 a pipe transport was just "/usr/bin". I have changed it to
2214 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2215 anything to be listed in the output from -bV.
2217 PH/05 When a filter generated an autoreply, the entire To: header line was
2218 quoted in the delivery log line, like this:
2220 => >A.N.Other <ano@some.domain> <original@ddress> ...
2222 This has been changed so that it extracts the operative address. There
2223 may be more than one such address. If so, they are comma-separated, like
2226 => >ano@some.domain,ona@other.domain <original@ddress> ...
2228 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2229 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2230 not being looked up in the rDNS to get a host name, Exim was showing the
2231 IP address twice in Received: lines, even though the IP addresses were
2232 identical. For example:
2234 Received: from [1.2.3.4] (helo=[1.2.3.4])
2236 However, if the real host name was known, it was omitting the HELO data
2237 if it matched the actual IP address. This has been tidied up so that it
2238 doesn't show the same IP address twice.
2240 PH/07 When both +timestamp and +memory debugging was on, the value given by
2241 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2242 called by the debug printing, thereby overwriting the timestamp buffer.
2243 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2245 PH/08 When the original message was included in an autoreply transport, it
2246 always said "this is a copy of the message, including all the headers",
2247 even if body_only or headers_only was set. It now gives an appropriate
2250 PH/09 Applied a patch from the Sieve maintainer which:
2252 o fixes some comments
2253 o adds the (disabled) notify extension core
2254 o adds some debug output for the result of if/elsif tests
2255 o points to the current vacation draft in the documentation
2256 and documents the missing references header update
2260 o fixes a bug in processing the envelope test (when testing
2261 multiple envelope elements, the last element determinted the
2264 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2265 Electronic Mail") by including:
2267 Auto-submitted: auto-generated
2269 in the messages that it generates (bounce messages and others, such as
2270 warnings). In the case of bounce messages for non-SMTP mesages, there was
2271 also a typo: it was using "Auto_submitted" (underscore instead of
2272 hyphen). Since every message generated by Exim is necessarily in response
2273 to another message, thes have all been changed to:
2275 Auto-Submitted: auto-replied
2277 in accordance with these statements in the RFC:
2279 The auto-replied keyword:
2281 - SHOULD be used on messages sent in direct response to another
2282 message by an automatic process,
2284 - MUST NOT be used on manually-generated messages,
2286 - MAY be used on Delivery Status Notifications (DSNs) and Message
2287 Disposition Notifications (MDNs),
2289 - MUST NOT be used on messages generated by automatic or periodic
2290 processes, except for messages which are automatic responses to
2293 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2294 to the default Received: header definition.
2296 PH/12 Added log selector acl_warn_skipped (default on).
2298 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2299 variables because (a) they are in the wrong storage pool and (b) even if
2300 they were copied, it wouldn't work properly because of the caching.
2302 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2303 checking when decoding. Apparently there are clients that generate
2304 overlong encoded strings. Why am I not surprised?
2306 PH/15 If the first argument of "${if match_address" was not empty, but did not
2307 contain an "@" character, Exim crashed. Now it writes a panic log message
2308 and treats the condition as false.
2310 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2312 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2313 extension "envelope-auth". The code is finished and in agreement with
2314 other implementations, but there is no documentation so far and in fact,
2315 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2316 not changing the active code.
2318 Print executed "if" and "elsif" statements when debugging is used. This
2319 helps a great deal to understand what a filter does.
2321 Document more things not specified clearly in RFC3028. I had all this
2322 sorted out, when out of a sudden new issues came to my mind. Oops."
2324 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2327 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2328 canonical form (as documented). However, after a host name lookup from
2329 the IP address, check_host() was doing a simple string comparison with
2330 addresses acquired from the DNS when checking that the found name did
2331 have the original IP as one of its addresses. Since any found IPv6
2332 addresses are likely to be in abbreviated form, the comparison could
2333 fail. Luckily, there already exists a function for doing the comparison
2334 by converting both addresses to binary, so now that is used instead of
2335 the text comparison.
2337 PH/20 There was another similar case to PH/19, when a complete host name was
2338 given in a host list; looking up its IP address could give an abbreviated
2339 form, whereas the current host's name might or might not be abbreviated.
2340 The same fix has been applied.
2346 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2347 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2350 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2351 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2353 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2355 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2356 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2357 Also, for "Auto-Submitted", treat anything other than "no" as
2358 non-personal, in accordance with RFC 3834. (Previously it treated
2359 anything starting "auto-" as non-personal.)
2361 TF/01 The control=submission/name=... option had a problem with syntax
2362 errors if the name included a slash character. The /name= option
2363 now slurps the rest of the string, so it can include any characters
2364 but it must come last in the list of options (after /sender_retain
2367 PH/05 Some modifications to the interface to the fake nameserver for the new
2375 TK/01 Added the "success_on_redirect" address verification option. See
2376 NewStuff for rationale and an example.
2378 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2380 PH/02 Patch to exigrep to allow it to work on syslog lines.
2382 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2383 fread() to skip over the body file's header line, because in Cygwin the
2384 header line is locked and is inaccessible.
2386 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2387 co-exist for some time) to make it clear that it is the Exim ID that is
2388 referenced, not the Message-ID: header line.
2390 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2391 string_format() function, because snprintf() does not exist on all
2394 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2395 ${stat: expansion item.
2397 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2398 protocol synchronization error", to keep the pedants happy.
2400 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2401 well as for IRIX systems, when gcc is being used. See the host.c source
2404 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2406 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2409 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2410 transport and to the smtp transport in LMTP mode.
2412 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2414 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2415 run a filter in a subprocess. This could lead to confusion in subsequent
2416 lookups in the parent process. There should also be a search_tidyup() at
2417 the end of the subprocess.
2419 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2420 only if the host matched helo_try_verify_hosts, which caused the
2421 verification to occur when the EHLO/HELO command was issued. The ACL just
2422 tested the remembered result. Now, if a previous verification attempt has
2423 not happened, "verify = helo" does it there and then.
2425 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2427 TK/03 Fix log output including CR from clamd.
2429 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2430 exist provoked a memory error which could cause a segfault.
2432 PH/15 Installed PCRE 6.2
2434 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2436 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2437 of the problem. Specifically, suggested +O2 rather than +O1 for the
2440 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2442 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2443 record, the port was not correctly logged when the outgoing_port log
2444 selector was set (it logged the transort's default port).
2446 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2447 fallback_hosts, and "hosts" in the smtp transport.
2449 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2450 host errors such as "Connection refused".
2452 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2453 authentication with radiusclient 0.4.9:
2455 - Error returned from rc_read_config was caught wrongly
2456 - Username/password not passed on to radius server due to wrong length.
2458 The presumption is that some radiusclient API changes for 4.51/PH/17
2459 were not taken care of correctly. The code is still untested by me (my
2460 Linux distribution still has 0.3.2 of radiusclient), but it was
2461 contributed by a Radius user.
2463 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2464 expanding the "port" option of the smtp transport.
2466 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2467 while reading a MIME header. Thanks to Tom Hughes for a patch.
2469 PH/24 Include config.h inside local_scan.h so that configuration settings are
2472 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2473 that in an ACL for RCPT (for example), you can examine exactly what was
2476 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2477 commands, but it was not correctly comparing the address with the actual
2478 client host address. Thus, it would show the EHLO address in Received:
2479 header lines when this was not necessary.
2481 PH/27 Added the % operator to ${eval:}.
2483 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2484 it should be ignoring failures (because with -C, for example, it has lost
2485 privilege). It wasn't ignoring creation failures other than "already
2488 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2491 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2492 because an input line was too long, either on its own, or by virtue of
2493 too many continuations, the temporary file was not being removed, and the
2494 return code was incorrect.
2496 PH/31 Missing "BOOL" in function definition in filtertest.c.
2498 PH/32 Applied Sieve patches from the maintainer.
2500 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2502 PH/33 Added "verify = not_blind".
2504 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2505 Local/Makefile (with some defaults set). These are used in built scripts
2506 such as exicyclog, but they have never been used in the exim_install
2507 script (though there are many overriding facilities there). I have
2508 arranged that the exim_install script now takes note of these two
2511 PH/35 Installed configuration files for Dragonfly.
2513 PH/36 When a locally submitted message by a trusted user did not contain a
2514 From: header, and the sender address was obtained from -f or from an SMTP
2515 MAIL command, and the trusted user did not use -F to supply a sender
2516 name, $originator_name was incorrectly used when constructing a From:
2517 header. Furthermore, $originator_name was used for submission mode
2518 messages from external hosts without From: headers in a similar way,
2519 which is clearly wrong.
2521 PH/37 Added control=suppress_local_fixups.
2523 PH/38 When log_selector = +received_sender was set, and the addition of the
2524 sender made the log line's construction buffer exactly full, or one byte
2525 less than full, an overflow happened when the terminating "\n" was
2528 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2529 when the result of a list match is failure because a DNS lookup failed.
2531 PH/40 RM_COMMAND is now used in the building process.
2533 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2534 the "build-* directories that it finds.
2536 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2537 address in a domain literal was a prefix of an interface address.
2539 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2540 when verifying a sender address, unless rewrite_headers is false.
2542 PH/44 Wrote a long comment about why errors_to addresses are verified as
2543 recipients, not senders.
2545 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2546 the ratelimit ACL was added.
2548 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2550 PH/46 Added extra information about PostgreSQL errors to the error string.
2552 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2553 avoiding the need to install special zones in a real server. This is
2554 backwards compatible; if it can't find the fake resolver, it drops back.
2555 Thus, both old and new test suites can be run.
2557 TF/02 Added util/ratelimit.pl
2559 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2560 clock is set back in time.
2562 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2563 Candler <B.Candler@pobox.com>.
2565 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2566 OK for addresses that are the result of redirections.
2568 PH/48 A number of further additions for the benefit of the new test suite,
2569 including a fake gethostbyname() that interfaces to the fake DNS resolver
2572 TF/06 The fix for widen_domains has also been applied to qualify_single and
2573 search_parents which are the other dnslookup options that can cause
2576 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2579 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2581 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2582 service" if it encountered an MX record with an empty target hostname.
2583 The message is now "an MX or SRV record indicated no SMTP service".
2585 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2586 if the DNS of the sending site is misconfigured. This is quite a
2587 common situation. This change restores the behaviour of treating a
2588 helo verification defer as a failure.
2590 PH/51 If self=fail was set on a router, the bounce message did not include the
2591 actual error message.
2597 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2599 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2600 command itself ended in error, the underlying message about the transport
2601 filter timeout was being overwritten with the pipe command error. Now the
2602 underlying error message should be appended to the second error message.
2604 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2606 PH/02 Reduce the amount of output that "make" produces by default. Full output
2607 can still be requested.
2609 PH/03 The warning log line about a condition test deferring for a "warn" verb
2610 was being output only once per connection, rather than after each
2611 occurrence (because it was using the same function as for successful
2612 "warn" verbs). This seems wrong, so I have changed it.
2614 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2615 it should not have, which might have caused a crash in the right
2616 circumstances, but probably never did.
2618 PH/04 Installed a modified version of Tony Finch's patch to make submission
2619 mode fix the return path as well as the Sender: header line, and to
2620 add a /name= option so that you can make the user's friendly name appear
2623 TF/03 Added the control = fakedefer ACL modifier.
2625 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2626 Mark Lowes for thorough testing.
2628 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2630 TK/03 Merged latest SRS patch from Miles Wilton.
2632 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2633 with the definition in sysexits.h (which is #included earlier).
2634 Fortunately, Exim does not actually use EX_OK. The code used to try to
2635 preserve the sysexits.h value, by assumimg that macro definitions were
2636 scanned for macro replacements. I have been disabused of this notion,
2637 so now the code just undefines EX_OK before #including unistd.h.
2639 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2640 in the smtp transport. When a block could not be written in a single
2641 write() function, the timeout was being re-applied to each part-write.
2642 This seems wrong - if the receiver was accepting one byte at a time it
2643 would take for ever. The timeout is now adjusted when this happens. It
2644 doesn't have to be particularly precise.
2646 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2647 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2649 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2650 without a domain if the check to <postmaster@domain> fails.
2652 SC/01 Eximstats: added -xls and the ability to specify output files
2653 (patch written by Frank Heydlauf).
2655 SC/02 Eximstats: use FileHandles for outputing results.
2657 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2659 SC/04 Eximstats: fixed display of large numbers with -nvr option
2661 SC/05 Eximstats: fixed merging of reports with empty tables.
2663 SC/06 Eximstats: added the -include_original_destination flag
2665 SC/07 Eximstats: removed tabs and trailing whitespace.
2667 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2669 TK/06 MBOX spool code: Add real "From " MBOX separator line
2670 so the .eml file is really in mbox format (even though
2671 most programs do not really care). Patch from Alex Miller.
2673 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2674 The latter is generated from $received_to and is only set if the
2675 message has one envelope recipient. SA can use these headers,
2676 obviously out-of-the-box. Patch from Alex Miller.
2678 PH/08 The ${def test on a variable was returning false if the variable's
2679 value was "0", contrary to what the specification has always said!
2680 The result should be true unless the variable is empty.
2682 PH/09 The syntax error of a character other than { following "${if
2683 def:variable_name" (after optional whitespace) was not being diagnosed.
2684 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2685 accidental colon was present, for example, could give incorrect results.
2687 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2688 result is used (not including appendfile, where other changes are about
2691 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2692 This involved changing a lot of size variables from int to off_t. It
2693 should work with maildirs and everything.
2695 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2696 spamd dying while we are connected to it.
2698 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2701 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2702 function for BDB 4.3.
2704 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2706 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2707 the sources for PCRE have been re-arranged and more files are now
2710 PH/15 The code I had for printing potentially long long variables in PH/11
2711 above was not the best (it lost precision). The length of off_t variables
2712 is now inspected at build time, and an appropriate printing format (%ld
2713 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2714 to be "long long int" or "long int". This is needed for the internal
2715 formatting function string_vformat().
2717 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2718 the configuration file to be ":syslog", then the script "guesses" where
2719 the logs files are, rather than using the compiled in default. In our
2720 case the guess is not the same as the compiled default, so the script
2721 suddenly stopped working when I started to use syslog. The patch checks
2722 to see if log_file_path is "". If so, it attempts to read it from exim
2723 with no configuration file to get the compiled in version, before it
2724 falls back to the previous guessing code."
2726 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2727 implementing BATV in an Exim configuration. See NewStuff for the gory
2730 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2731 Makefile that are specific to HP-UX.
2733 PH/18 If the "use_postmaster" option was set for a recipient callout together
2734 with the "random" option, the postmaster address was used as the MAIL
2735 FROM address for the random test, but not for the subsequent recipient
2736 test. It is now used for both.
2738 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2739 patch removes a few documentation additions to RFC 3028, because the
2740 latest draft now contains them. It adds the new en;ascii-case comparator
2741 and a new error check for 8bit text in MIME parts. Comparator and
2742 require names are now matched exactly. I enabled the subaddress
2743 extension, but it is not well tested yet (read: it works for me)."
2745 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2746 rework some of the code of TK/09 above to avoid the hardwired use of
2747 "%lld" and "long long". Replaced the call to snprintf() with a call to
2750 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2751 records point to non-existent hosts", "retry timeout exceeded", and
2752 "retry time not reached for any host after a long failure period".
2754 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2755 experimental DomainKeys support:
2757 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2758 (2) On an error such as an illegally used "control", the wrong name for
2759 the control was given.
2761 These problems did NOT occur unless DomainKeys support was compiled.
2763 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2765 PH/24 Added ${if match_ip condition.
2767 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2768 databases so that it will be absolutely obvious if a crash occurs in the
2769 DB library. This is a regular occurrence (often caused by mis-matched
2772 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2773 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2774 picked up on a user's system that detects such things. There doesn't seem
2775 to be a gcc warning option for this - only an attribute that has to be
2776 put on the function's prototype. It seems that in Fedora Core 4 they have
2777 set this on a number of new functions. No doubt there will be more in due
2780 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2781 specify a transport. However, if an address that was verified by such a
2782 router was the subject of a callout, Exim crashed because it tried to
2783 read the rcpt_include_affixes from the non-existent transport. Now it
2784 just assumes that the setting of that option is false. This bug was
2785 introduced by 4.51/PH/31.
2787 PH/28 Changed -d+all to exclude +memory, because that information is very
2788 rarely of interest, but it makes the output a lot bigger. People tend to
2789 do -d+all out of habit.
2791 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2792 code in os-type was giving problems when libc.so lives in lib64, like on
2795 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2796 aren't the modern standard, and it seems that some systems' include files
2797 don't always have them. Exim was already checking for some of the newer
2798 ones like T_AAAA, and defining it itself. I've added checks for all the
2799 record types that Exim uses.
2801 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2802 not automatically generating a new one, as it is supposed to. This
2803 prevented TLS from working. If the file did exist, but contained invalid
2804 data, a new version was generated, as expected. It was only the case of a
2805 non-existent file that was broken.
2807 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2808 with a change in libdomainkeys > 0.64.
2810 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2811 from DNS. If the selector record carries the flag, it now has
2812 precedence over the domain-wide flag.
2814 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2816 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2817 the use of a port name that isn't defined in /etc/services) occurred, the
2818 message was deferred as in a normal delivery, and thus remained on the
2819 spool, instead of being failed because of the mua_wrapper setting. This
2820 is now fixed, and I tidied up some of the mua_wrapper messages at the
2823 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2824 the messages in a hash of arrays rather than using individual hashes.
2825 This is a bit cleaner and results in dramatic memory savings, albeit
2826 at a slight CPU cost.
2828 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2829 as requested by Marc Sherman.
2831 SC/10 Eximstats: added histograms for user specified patterns as requested
2834 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2836 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2837 fopen() in the content-scanning modules that did not already have it.
2843 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2844 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2846 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2848 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2850 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2851 to test Sieve filters that use "vacation".
2853 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2854 that changes the way the GnuTLS parameters are stored in the cache file.
2855 The new format can be generated externally. For backward compatibility,
2856 if the data in the cache doesn't make sense, Exim assumes it has read an
2857 old-format file, and it generates new data and writes a new file. This
2858 means that you can't go back to an older release without removing the
2861 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2862 work if there are any delivery delays because "one_time" forces the
2863 parent to be marked "delivered", so its unseen clone is never tried
2864 again. For this reason, Exim now forbids the simultaneous setting of
2867 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2868 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2869 entry if you want to know the details. The fix, however, neglected to
2870 consider the case when local delivery batching is involved. The test for
2871 "previously delivered" was not happening when checking to see if an
2872 address could be batched with a previous (undelivered) one; under
2873 certain circumstances this could lead to multiple deliveries to the same
2876 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2877 in its include files, and this causes problems building Exim.
2879 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2880 header_syntax) but Exim was just ignoring anything given after a slash.
2881 In particular, this caused confusion with an attempt to use "verify =
2882 reverse_host_lookup/defer_ok". An error is now given when options are
2883 supplied for verify items that do not have them. (Maybe reverse_host_
2884 lookup should have a defer_ok option, but that's a different point.)
2886 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2887 defined by RFC 821) to 2048, because there were problems with some AUTH
2888 commands, and RFC 1869 says the size should be increased for extended
2889 SMTP commands that take arguments.
2891 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2894 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2895 "unknown" error; now it says that the functionality isn't in the binary.
2897 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2898 an address' error message when a string expansion fails (syntax or
2899 whatever). Otherwise the password may appear in the log. Following change
2900 PH/42 below, there is no longer a chance of it appearing in a bounce
2903 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2905 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2906 of its addresses was ever tried. (Bugzilla bug #2.)
2908 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2909 the result incorrectly in the debug output. (It correctly added a newline
2910 to what was transported.)
2912 TF/01 Added $received_time.
2914 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2915 commented out examples of how to interface to a virus scanner and to
2916 SpamAssassin. Also added commented examples of av_scanner and
2917 spamd_address settings.
2919 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2920 and controls are allowed in which ACLs. There were a couple of minor
2921 errors. Some of the entries in the conditions table (which is a table of
2922 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2923 negation of where the condition IS allowed.
2925 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2927 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2928 header file does not have a version number, so I've had to invent a new
2929 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2930 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2931 radiusclient), but it was contributed by a Radius user.
2933 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2934 files or directories, for OpenSSL.
2936 PH/19 When an Exim process that is running as root has to create an Exim log
2937 file, it does so in a subprocess that runs as exim:exim so as to get the
2938 ownership right at creation (otherwise, other Exim processes might see
2939 the file with the wrong ownership). There was no test for failure of this
2940 fork() call, which would lead to the process getting stuck as it waited
2941 for a non-existent subprocess. Forks do occasionally fail when resources
2942 run out. I reviewed all the other calls to fork(); they all seem to check
2945 PH/20 When checking for unexpected SMTP input at connect time (before writing
2946 the banner), Exim was not dealing correctly with a non-positive return
2947 from the read() function. If the client had disconnected by this time,
2948 the result was a log entry for a synchronization error with an empty
2949 string after "input=" when read() returned zero. If read() returned -1
2950 (an event I could not check), uninitialized data bytes were printed.
2951 There were reports of junk text (parts of files, etc) appearing after
2954 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2956 PH/22 Added support for macro redefinition, and (re)definition in between
2957 driver and ACL definitions.
2959 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2960 forgetting to use the resulting value; it was using the unexpanded value.
2962 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2963 hadn't been configured. The fix is from Juergen Kreileder, who
2964 understands it better than I do:
2966 "Here's what I see happening with three configured cyrus_sasl
2967 authenticators configured (plain, login, cram-md5):
2969 On startup auth_cyrus_sasl_init() gets called for each of these.
2970 This means three calls to sasl_listmech() without a specified mech_list.
2971 => SASL tests which mechs of all available mechs actually work
2972 => three warnings about OTP not working
2973 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2975 With the patch, sasl_listmech() also gets called three times. But now
2976 SASL's mech_list option is set to the server_mech specified in the the
2977 authenticator. Or in other words, the answer from sasl_listmech()
2978 gets limited to just the mech you're testing for (which is different
2980 => the return list contains just 'plain' or 'login', 'cram-md5' or
2981 nothing depending on the value of ob->server_mech.
2983 I've just tested the patch: Authentication still works fine,
2984 unavailable mechs specified in the exim configuration are still
2985 caught, and the auth.log warnings about OTP are gone."
2987 PH/25 When debugging is enabled, the contents of the command line are added
2988 to the debugging output, even when log_selector=+arguments is not
2991 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2992 answer is "GNU", and only if the return is "GNU/something" is the answer
2995 PH/27 $acl_verify_message is now set immediately after the failure of a
2996 verification in an ACL, and so is available in subsequent modifiers. In
2997 particular, the message can be preserved by coding like this:
2999 warn !verify = sender
3000 set acl_m0 = $acl_verify_message
3002 Previously, $acl_verify_message was set only while expanding "message"
3003 and "log_message" when a very denied access.
3005 PH/28 Modified OS/os.c-Linux with
3007 -#ifndef OS_LOAD_AVERAGE
3008 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3010 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3011 nomenclature these days.)
3013 PH/29 Installed patch from the Sieve maintainer that adds the options
3014 sieve_useraddress and sieve_subaddress to the redirect router.
3016 PH/30 In these circumstances:
3017 . Two addresses routed to the same list of hosts;
3018 . First host does not offer TLS;
3019 . First host accepts first address;
3020 . First host gives temporary error to second address;
3021 . Second host offers TLS and a TLS session is established;
3022 . Second host accepts second address.
3023 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3024 and peerdn, if requested) that were in fact used only for the second
3027 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3028 attention to any local part prefix or suffix that was matched by the
3029 router that accepted the address. It now behaves in the same way as it
3030 does for delivery: the affixes are removed from the local part unless
3031 rcpt_include_affixes is set on the transport.
3033 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3034 timeout during the DATA phase of an incoming message.
3036 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3037 applied a patch sanctioned by the Sieve maintainer.
3039 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3040 the uid or gid is negative. A case of a negative gid caused this to be
3041 noticed. The fix allows for either to be negative.
3043 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3044 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3047 PH/36 The change PH/12 above was broken. Fixed it.
3049 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3050 the grounds that routing the same address twice would always produce the
3051 same answer. This might have been true once, but it is certainly no
3052 longer true now. Routing a child address may depend on the previous
3053 routing that produced that child. Some complicated redirection strategies
3054 went wrong when messages had multiple recipients, and made Exim's
3055 behaviour dependent on the order in which the addresses were given.
3057 I have moved the duplicate checking until after the routing is complete.
3058 Exim scans the addresses that are assigned to local and remote
3059 transports, and removes any duplicates. This means that more work will be
3060 done, as duplicates will always all be routed, but duplicates are
3061 presumably rare, so I don't expect this is of any significance.
3063 For deliveries to pipes, files, and autoreplies, the duplicate checking
3064 still happens during the routing process, since they are not going to be
3067 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3068 It corrects a timeout issue with spamd. This is Ian's comment: "The
3069 background is that sometimes spamd either never reads data from a
3070 connection it has accepted, or it never writes response data. The exiscan
3071 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3072 blindly assumes that writes won't block so it may never time out."
3074 PH/39 Allow G after quota size as well as K and M.
3076 PH/40 The value set for $authenticated_id in an authenticator may not contain
3077 binary zeroes or newlines because the value is written to log lines and
3078 to spool files. There was no check on this. Now the value is run through
3079 the string_printing() function so that such characters are converted to
3080 printable escape sequences.
3082 PH/41 $message_linecount is a new variable that contains the total number of
3083 lines in the message. Compare $body_linecount, which is the count for the
3086 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3087 bounce and delay warning messages, except in certain special cases, which
3090 (a) An SMTP error message from a remote host;
3091 (b) A message specified in a :fail: redirection;
3092 (c) A message specified in a "fail" command in a system filter;
3093 (d) A message specified in a FAIL return from the queryprogram router;
3094 (e) A message specified by the cannot_route_message router option.
3096 In these cases only, Exim does include the error details in bounce and
3097 warning messages. There are also a few cases where bland messages such
3098 as "unrouteable address" or "local delivery error" are given.
3100 PH/43 $value is now also set for the "else" part of a ${run expansion.
3102 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3103 being worked on, but at least Exim now implements the latest version to
3106 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3107 process to complete was treated as a delivery failure, a timeout while
3108 writing the message to the pipe was logged, but erroneously treated as a
3109 successful delivery. Such timeouts include transport filter timeouts. For
3110 consistency with the overall process timeout, these timeouts are now
3111 treated as errors, giving rise to delivery failures by default. However,
3112 there is now a new Boolean option for the pipe transport called
3113 timeout_defer, which, if set TRUE, converts the failures into defers for
3114 both kinds of timeout. A transport filter timeout is now identified in
3117 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3118 systems where "make" and "gmake" are different, calling "gmake" at top
3119 level broke things. I've arranged for the value of $(MAKE) to be passed
3120 from the Makefile to this script so that it can call the same version of
3124 A note about Exim versions 4.44 and 4.50
3125 ----------------------------------------
3127 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3128 changes of various kinds. As a consequence, a big documentation update was
3129 needed. This delayed the release for rather longer than seemed good, especially
3130 in the light of a couple of (minor) security issues. Therefore, the changes
3131 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3132 release. So 4.44 and 4.50 are in effect two different branches that both start
3135 I have left the 4.50 change log unchanged; it contains all the changes since
3136 4.43. The change log for 4.44 is below; many of its items are identical to
3137 those for 4.50. This seems to be the most sensible way to preserve the
3138 historical information.
3144 1. Minor wording change to the doc/README.SIEVE file.
3146 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3147 computation of the current number of files was incorrect.
3149 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3150 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3153 4. Give more explanation in the error message when the command for a transport
3154 filter fails to execute.
3156 5. There are several places where Exim runs a non-Exim command in a
3157 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3158 was being done only for the command run by the queryprogram router. It is
3159 now done for all such subprocesses. The other cases are: ${run, transport
3160 filters, and the commands run by the lmtp and pipe transports.
3162 6. Added CONFIGURE_GROUP build-time option.
3164 7. Some older OS have a limit of 256 on the maximum number of file
3165 descriptors. Exim was using setrlimit() to set 1000 as a large value
3166 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3167 systems. I've change it so that if it can't get 1000, it tries for 256.
3169 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3170 was an oversight, and furthermore, ever since the addition of extra
3171 controls (e.g. 4.43/32), the checks on when to allow different forms of
3172 "control" were broken. There should now be diagnostics for all cases when a
3173 control that does not make sense is encountered.
3175 9. Added the /retain_sender option to "control=submission".
3177 10. $recipients is now available in the predata ACL (oversight).
3179 11. Tidy the search cache before the fork to do a delivery from a message
3180 received from the command line. Otherwise the child will trigger a lookup
3181 failure and thereby defer the delivery if it tries to use (for example) a
3182 cached ldap connection that the parent has called unbind on.
3184 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3185 of $address_data from the recipient verification was clobbered by the
3186 sender verification.
3188 13. The value of address_data from a sender verification is now available in
3189 $sender_address_data in subsequent conditions in the ACL statement.
3191 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3193 15. Added a new option "connect=<time>" to callout options, to set a different
3196 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3197 was its contents. (It was OK if the option was not defined at all.)
3199 17. A "Completed" log line is now written for messages that are removed from
3200 the spool by the -Mrm option.
3202 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3203 information about exactly what failed.
3205 19. Added -dd to debug only the daemon process.
3207 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3208 handles timeouts, both on the server side and network timeouts. Renamed the
3209 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3211 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3212 It is now set to "smtps".
3214 22. $host_address is now set to the target address during the checking of
3215 ignore_target_hosts.
3217 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3218 being passed; this would have caused $sender_host_name to have been used if
3219 matching the list had actually called for a host name (not very likely,
3220 since this list is usually IP addresses). A host name is now passed as
3223 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3224 code that specifies a non-restarting handler (typically sigaction() in
3225 modern systems) in an attempt to fix a rare and obscure crash bug.
3227 25. Narrowed the window for a race in the daemon that could cause it to ignore
3228 SIGCHLD signals. This is not a major problem, because they are used only to
3229 wake it up if nothing else does.
3231 26. A malformed maildirsize file could cause Exim to calculate negative values
3232 for the mailbox size or file count. Odd effects could occur as a result.
3233 The maildirsize information is now recalculated if the size or filecount
3236 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3237 support for a long time. Removed HAVE_SYS_VFS_H.
3239 28. Installed the latest version of exipick from John Jetmore.
3241 29. In an address list, if the pattern was not a regular expression, an empty
3242 subject address (from a bounce message) matched only if the pattern was an
3243 empty string. Non-empty patterns were not even tested. This was the wrong
3244 because it is perfectly reasonable to use an empty address as part of a
3245 database query. An empty address is now tested by patterns that are
3246 lookups. However, all the other forms of pattern expect the subject to
3247 contain a local part and a domain, and therefore, for them, an empty
3248 address still always fails if the pattern is not itself empty.
3250 30. Exim went into a mad DNS loop when attempting to do a callout where the
3251 host was specified on an smtp transport, and looking it up yielded more
3252 than one IP address.
3254 31. Re-factored the code for checking spool and log partition space into a
3255 function that finds that data and another that does the check. The former
3256 is then used to implement four new variables: $spool_space, $log_space,
3257 $spool_inodes, and $log_inodes.
3259 32. The RFC2047 encoding function was originally intended for short strings
3260 such as real names; it was not keeping to the 75-character limit for
3261 encoded words that the RFC imposes. It now respects the limit, and
3262 generates multiple encoded words if necessary. To be on the safe side, I
3263 have increased the buffer size for the ${rfc2047: expansion operator from
3266 33. It is now permitted to omit both strings after an "if" condition; if the
3267 condition is true, the result is "true". As before, when the second string
3268 is omitted, a false condition yields an empty string. This makes it less
3269 cumbersome to write custom ACL and router conditions.
3271 34. Failure to deliver a bounce message always caused it to be frozen, even if
3272 there was an errors_to setting on the router. The errors_to setting is now
3275 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3276 canonical form (fully expanded) before being placed in
3277 $sender_host_address.
3279 36. The table in the code that translates DNS record types into text (T_A to
3280 "A" for instance) was missing entries for NS and CNAME. It is just possible
3281 that this could have caused confusion if both these types were looked up
3282 for the same domain, because the text type is used as part of Exim's
3283 per-process caching. But the chance of anyone hitting this buglet seems
3286 37. The dnsdb lookup has been extended in a number of ways.
3288 (1) There is a new type, "zns", which walks up the domain tree until it
3289 finds some nameserver records. It should be used with care.
3291 (2) There is a new type, "mxh", which is like "mx" except that it returns
3292 just the host names, not the priorities.
3294 (3) It is now possible to give a list of domains (or IP addresses) to be
3295 looked up. The behaviour when one of the lookups defers can be
3296 controlled by a keyword.
3298 (4) It is now possible to specify the separator character for use when
3299 multiple records are returned.
3301 38. The dnslists ACL condition has been extended: it is now possible to supply
3302 a list of IP addresses and/or domains to be looked up in a particular DNS
3305 39. Added log_selector=+queue_time_overall.
3307 40. When running the queue in the test harness, wait just a tad after forking a
3308 delivery process, to get repeatability of debugging output.
3310 41. Include certificate and key file names in error message when GnuTLS fails
3311 to set them up, because the GnuTLS error message doesn't include the name
3312 of the failing file when there is a problem reading it.
3314 42. Allow both -bf and -bF in the same test run.
3316 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3318 44. The "Exiscan patch" is now merged into the mainline Exim source.
3320 45. Sometimes the final signoff response after QUIT could fail to get
3321 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3322 < 0 before doing a fflush(). This bug looks as though it goes back to the
3323 introduction of TLS in release 3.20, but "sometimes" must have been rare
3324 because the tests only now provoked it.
3326 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3327 (this can affect the format of dates).
3329 47. exim_tidydb, when checking for the continued existence of a message for
3330 which it has found a message-specific retry record, was not finding
3331 messages that were in split spool directories. Consequently, it was
3332 deleting retry records that should have stayed in existence.
3334 48. Steve fixed some bugs in eximstats.
3336 49. The SPA authentication driver was not abandoning authentication and moving
3337 on to the next authenticator when an expansion was forced to fail,
3338 contradicting the general specification for all authenticators. Instead it
3339 was generating a temporary error. It now behaves as specified.
3341 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3342 (the order specifies the preference for clients). The order is now AES256,
3343 AES128, 3DES, ARCFOUR128.
3345 51. Small patch to Sieve code - explicitly set From: when generating an
3348 52. Exim crashed if a remote delivery caused a very long error message to be
3349 recorded - for instance if somebody sent an entire SpamAssassin report back
3350 as a large number of 550 error lines. This bug was coincidentally fixed by
3351 increasing the size of one of Exim's internal buffers (big_buffer) that
3352 happened as part of the Exiscan merge. However, to be on the safe side, I
3353 have made the code more robust (and fixed the comments that describe what
3356 53. Now that there can be additional text after "Completed" in log lines (if
3357 the queue_time_overall log selector is set), a one-byte patch to exigrep
3358 was needed to allow it to recognize "Completed" as not the last thing in
3361 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3362 patch that reportedly fixes this has been added. I am not expert enough to
3363 create a test for it. This is what the patch creator wrote:
3365 "I found a little strange behaviour of ldap code when working with
3366 Windows 2003 AD Domain, where users was placed in more than one
3367 Organization Units. When I tried to give exim partial DN, the exit code
3368 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3369 But simultaneously result of request was absolutely normal ldap result,
3370 so I produce this patch..."
3372 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3373 REFERENCE, so I have modified the code to exclude the patch when that macro
3376 55. Some experimental protocols are using DNS PTR records for new purposes. The
3377 keys for these records are domain names, not reversed IP addresses. The
3378 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3379 leaves it alone. Component reversal etc. now happens only for IP addresses.
3382 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3384 57. Double the size of the debug message buffer (to 2048) so that more of very
3385 long debug lines gets shown.
3387 58. The exicyclog utility now does better if the number of log files to keep
3388 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3390 59. Two changes related to the smtp_active_hostname option:
3392 (1) $smtp_active_hostname is now available as a variable.
3393 (2) The default for smtp_banner uses $smtp_active_hostname instead
3394 of $primary_hostname.
3396 60. The host_aton() function is supposed to be passed a string that is known
3397 to be a valid IP address. However, in the case of IPv6 addresses, it was
3398 not checking this. This is a hostage to fortune. Exim now panics and dies
3399 if the condition is not met. A case was found where this could be provoked
3400 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3401 components; fortuitously, this particular loophole had already been fixed
3402 by change 4.50/55 above.
3404 If there are any other similar loopholes, the new check in host_aton()
3405 itself should stop them being exploited. The report I received stated that
3406 data on the command line could provoke the exploit when Exim was running as
3407 exim, but did not say which command line option was involved. All I could
3408 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3409 running as the user.
3412 61. There was a buffer overflow vulnerability in the SPA authentication code
3413 (which came originally from the Samba project). I have added a test to the
3414 spa_base64_to_bits() function which I hope fixes it.
3417 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3418 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3420 63. The daemon start-up calls getloadavg() while still root for those OS that
3421 need the first call to be done as root, but it missed one case: when
3422 deliver_queue_load_max is set with deliver_drop_privilege. This is
3423 necessary for the benefit of the queue runner, because there is no re-exec
3424 when deliver_drop_privilege is set.
3426 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3427 This has been fixed.
3429 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3430 was in use, was not putting the data itself into the right store pool;
3431 consequently, it could be overwritten for a subsequent message in the same
3432 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3435 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3437 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3438 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3439 the function were treating the return as a boolean value, which happened to
3440 work because 0=false and not-0=true, but is not correct code.
3442 68. The host_aton() function was not handling scoped IPv6 addresses (those
3443 with, for example, "%eth0" on the end) correctly.
3445 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3446 negated items (that is, ~something) in unsigned ints. Some compilers
3447 apparently mutter when there is no cast.
3449 70. If an address verification called from an ACL failed, and did not produce a
3450 user-specific message (i.e. there was only a "system" message), nothing was
3451 put in $acl_verify_message. In this situation, it now puts the system
3454 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3455 session; change 4.31/43 added the unwanted input to the log line - except
3456 that it did not do this in the start of session case. It now does.
3458 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3459 This is wrong and can cause the other end to generate a synchronization
3460 error if it is another Exim or anything else that does the synchronization
3461 check. A QUIT command is no longer sent after a timeout.
3463 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3464 during host lookups.
3466 74. The defer_ok option of callout verification was not working if it was used
3467 when verifying addresses in header lines, that is, for this case:
3469 verify = header_sender/callout=defer_ok
3471 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3472 those file descriptors could be used for SMTP connections. If anything
3473 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3474 could be sent to the SMTP client, causing chaos. The daemon now opens
3475 stdin, stdout, and stderr to /dev/null when it puts itself into the
3478 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3479 log by default. The user can override this with suitable Perl magic.
3481 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3482 the log message when discard triggers, was not working for the DATA ACL or
3483 for the non-SMTP ACL.
3485 78. Error message wording change in sieve.c.
3487 79. If smtp_accept_max_per_host was set, the number of connections could be
3488 restricted to fewer than expected, because the daemon was trying to set up
3489 a new connection before checking whether the processes handling previous
3490 connections had finished. The check for completed processes is now done
3491 earlier. On busy systems, this bug wouldn't be noticed because something
3492 else would have woken the daemon, and it would have reaped the completed
3495 80. If a message was submitted locally by a user whose login name contained one
3496 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3497 It caused a spool format error. I have fixed the spool reading code. A
3498 related problem was that the "from" clause in the Received: line became
3499 illegal because of the space(s). It is now covered by ${quote_local_part.
3501 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3504 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3506 83. Patch from Sieve maintainer to fix unterminated string problem in
3507 "vacation" handling.
3509 84. Some minor changes to the Linux configuration files to help with other
3510 OS variants using glibc.
3512 85. One more patch for Sieve to update vacation handling to latest spec.
3515 ----------------------------------------------------
3516 See the note above about the 4.44 and 4.50 releases.
3517 ----------------------------------------------------
3523 1. Change 4.43/35 introduced a bug that caused file counts to be
3524 incorrectly computed when quota_filecount was set in an appendfile
3527 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3528 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3531 3. Give more explanation in the error message when the command for a transport
3532 filter fails to execute.
3534 4. There are several places where Exim runs a non-Exim command in a
3535 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3536 was being done only for the command run by the queryprogram router. It is
3537 now done for all such subprocesses. The other cases are: ${run, transport
3538 filters, and the commands run by the lmtp and pipe transports.
3540 5. Some older OS have a limit of 256 on the maximum number of file
3541 descriptors. Exim was using setrlimit() to set 1000 as a large value
3542 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3543 systems. I've change it so that if it can't get 1000, it tries for 256.
3545 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3546 was an oversight, and furthermore, ever since the addition of extra
3547 controls (e.g. 4.43/32), the checks on when to allow different forms of
3548 "control" were broken. There should now be diagnostics for all cases when a
3549 control that does not make sense is encountered.
3551 7. $recipients is now available in the predata ACL (oversight).
3553 8. Tidy the search cache before the fork to do a delivery from a message
3554 received from the command line. Otherwise the child will trigger a lookup
3555 failure and thereby defer the delivery if it tries to use (for example) a
3556 cached ldap connection that the parent has called unbind on.
3558 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3559 of $address_data from the recipient verification was clobbered by the
3560 sender verification.
3562 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3563 was its contents. (It was OK if the option was not defined at all.)
3565 11. A "Completed" log line is now written for messages that are removed from
3566 the spool by the -Mrm option.
3568 12. $host_address is now set to the target address during the checking of
3569 ignore_target_hosts.
3571 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3572 being passed; this would have caused $sender_host_name to have been used if
3573 matching the list had actually called for a host name (not very likely,
3574 since this list is usually IP addresses). A host name is now passed as
3577 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3578 code that specifies a non-restarting handler (typically sigaction() in
3579 modern systems) in an attempt to fix a rare and obscure crash bug.
3581 15. Narrowed the window for a race in the daemon that could cause it to ignore
3582 SIGCHLD signals. This is not a major problem, because they are used only to
3583 wake it up if nothing else does.
3585 16. A malformed maildirsize file could cause Exim to calculate negative values
3586 for the mailbox size or file count. Odd effects could occur as a result.
3587 The maildirsize information is now recalculated if the size or filecount
3590 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3591 support for a long time. Removed HAVE_SYS_VFS_H.
3593 18. Updated exipick to current release from John Jetmore.
3595 19. Allow an empty sender to be matched against a lookup in an address list.
3596 Previously the only cases considered were a regular expression, or an
3599 20. Exim went into a mad DNS lookup loop when doing a callout where the
3600 host was specified on the transport, if the DNS lookup yielded more than
3603 21. The RFC2047 encoding function was originally intended for short strings
3604 such as real names; it was not keeping to the 75-character limit for
3605 encoded words that the RFC imposes. It now respects the limit, and
3606 generates multiple encoded words if necessary. To be on the safe side, I
3607 have increased the buffer size for the ${rfc2047: expansion operator from
3610 22. Failure to deliver a bounce message always caused it to be frozen, even if
3611 there was an errors_to setting on the router. The errors_to setting is now
3614 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3615 canonical form (fully expanded) before being placed in
3616 $sender_host_address.
3618 24. Updated eximstats to version 1.33
3620 25. Include certificate and key file names in error message when GnuTLS fails
3621 to set them up, because the GnuTLS error message doesn't include the name
3622 of the failing file when there is a problem reading it.
3624 26. Expand error message when OpenSSL has problems setting up cert/key files.
3627 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3628 (this can affect the format of dates).
3630 28. exim_tidydb, when checking for the continued existence of a message for
3631 which it has found a message-specific retry record, was not finding
3632 messages that were in split spool directories. Consequently, it was
3633 deleting retry records that should have stayed in existence.
3635 29. eximstats updated to version 1.35
3636 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3637 1.35 - bugfix such that pie charts by volume are generated correctly
3639 30. The SPA authentication driver was not abandoning authentication and moving
3640 on to the next authenticator when an expansion was forced to fail,
3641 contradicting the general specification for all authenticators. Instead it
3642 was generating a temporary error. It now behaves as specified.
3644 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3645 (the order specifies the preference for clients). The order is now AES256,
3646 AES128, 3DES, ARCFOUR128.
3648 31. Small patch to Sieve code - explicitly set From: when generating an
3651 32. Exim crashed if a remote delivery caused a very long error message to be
3652 recorded - for instance if somebody sent an entire SpamAssassin report back
3653 as a large number of 550 error lines. This bug was coincidentally fixed by
3654 increasing the size of one of Exim's internal buffers (big_buffer) that
3655 happened as part of the Exiscan merge. However, to be on the safe side, I
3656 have made the code more robust (and fixed the comments that describe what
3659 33. Some experimental protocols are using DNS PTR records for new purposes. The
3660 keys for these records are domain names, not reversed IP addresses. The
3661 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3662 leaves it alone. Component reversal etc. now happens only for IP addresses.
3665 34. The host_aton() function is supposed to be passed a string that is known
3666 to be a valid IP address. However, in the case of IPv6 addresses, it was
3667 not checking this. This is a hostage to fortune. Exim now panics and dies
3668 if the condition is not met. A case was found where this could be provoked
3669 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3670 components; fortuitously, this particular loophole had already been fixed
3671 by change 4.50/55 or 4.44/33 above.
3673 If there are any other similar loopholes, the new check in host_aton()
3674 itself should stop them being exploited. The report I received stated that
3675 data on the command line could provoke the exploit when Exim was running as
3676 exim, but did not say which command line option was involved. All I could
3677 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3678 running as the user.
3681 35. There was a buffer overflow vulnerability in the SPA authentication code
3682 (which came originally from the Samba project). I have added a test to the
3683 spa_base64_to_bits() function which I hope fixes it.
3686 36. The daemon start-up calls getloadavg() while still root for those OS that
3687 need the first call to be done as root, but it missed one case: when
3688 deliver_queue_load_max is set with deliver_drop_privilege. This is
3689 necessary for the benefit of the queue runner, because there is no re-exec
3690 when deliver_drop_privilege is set.
3692 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3693 was in use, was not putting the data itself into the right store pool;
3694 consequently, it could be overwritten for a subsequent message in the same
3695 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3698 38. Sometimes the final signoff response after QUIT could fail to get
3699 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3700 < 0 before doing a fflush(). This bug looks as though it goes back to the
3701 introduction of TLS in release 3.20, but "sometimes" must have been rare
3702 because the tests only now provoked it.
3708 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3709 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3710 Now it returns an integer. A number of calls were still expecting a T/F
3711 return. Fortuitously, in all cases, the tests worked in OK situations,
3712 which is the norm. However, things would have gone wrong on any write
3713 failures on the smtp file descriptor. This function is used when sending
3714 messages over SMTP and also when doing verify callouts.
3716 2. When Exim is called to do synchronous delivery of a locally submitted
3717 message (the -odf or -odi options), it no longer closes stderr before doing
3720 3. Implemented the mua_wrapper option.
3722 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3724 5. Implemented the functions header_remove(), header_testname(),
3725 header_add_at_position(), and receive_remove_recipient(), and exported them
3728 6. If an ACL "warn" statement specified the addition of headers, Exim already
3729 inserted X-ACL-Warn: at the start if there was no header name. However, it
3730 was not making this test for the second and subsequent header lines if
3731 there were newlines in the string. This meant that an invalid header could
3732 be inserted if Exim was badly configured.
3734 7. Allow an ACL "warn" statement to add header lines at the start or after all
3735 the Received: headers, as well as at the end.
3737 8. Added the rcpt_4xx retry error code.
3739 9. Added postmaster_mailfrom=xxx to callout verification option.
3741 10. Added mailfrom=xxxx to the callout verification option, for verify=
3744 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3745 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3747 12. Inserted some casts to stop certain compilers warning when using pointer
3748 differences as field lengths or precisions in printf-type calls (mostly
3749 affecting debugging statements).
3751 13. Added optional readline() support for -be (dynamically loaded).
3753 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3754 same clock tick as a message's arrival, so that its received time was the
3755 same as the "first fail" time on the retry record, and that message
3756 remained on the queue past the ultimate address timeout, every queue runner
3757 would try a delivery (because it was past the ultimate address timeout) but
3758 after another failure, the ultimate address timeout, which should have then
3759 bounced the address, did not kick in. This was a "< instead of <=" error;
3760 in most cases the first failure would have been in the next clock tick
3761 after the received time, and all would be well.
3763 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3764 being recognized when the domain list was tested by the match_domain
3765 condition in an expansion string.
3767 16. Added the ${str2b64: operator.
3769 17. Exim was always calling setrlimit() to set a large limit for the number of
3770 processes, without checking whether the existing limit was already
3771 adequate. (It did check for the limit on file descriptors.) Furthermore,
3772 errors from getrlimit() and setrlimit() were being ignored. Now they are
3773 logged to the main and panic logs, but Exim does carry on, to try to do its
3774 job under whatever limits there are.
3776 18. Imported PCRE 5.0.
3778 19. Trivial typo in log message " temporarily refused connection" (the leading
3781 20. If the log selector return_path_on_delivery was set and an address was
3782 redirected to /dev/null, the delivery process crashed because it assumed
3783 that a return path would always be set for a "successful" delivery. In this
3784 case, the whole delivery is bypassed as an optimization, and therefore no
3787 21. Internal re-arrangement: the function for sending a challenge and reading
3788 a response while authentication was assuming a zero-terminated challenge
3789 string. It's now changed to take a pointer and a length, to allow for
3790 binary data in such strings.
3792 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3794 23. Exim was not respecting finduser_retries when seeking the login of the
3795 uid under which it was called; it was always trying 10 times. (The default
3796 setting of finduser_retries is zero.) Also, it was sleeping after the final
3797 failure, which is pointless.
3799 24. Implemented tls_on_connect_ports.
3801 25. Implemented acl_smtp_predata.
3803 26. If the domain in control=submission is set empty, Exim assumes that the
3804 authenticated id is a complete email address when it generates From: or
3805 Sender: header lines.
3807 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3808 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3809 chown and chgrp in /bin and hostname in /usr/bin.
3811 28. Exim was keeping the "process log" file open after each use, just as it
3812 does for the main log. This opens the possibility of it remaining open for
3813 long periods when the USR1 signal hits a daemon. Occasional processlog
3814 errors were reported, that could have been caused by this. Anyway, it seems
3815 much more sensible not to leave this file open at all, so that is what now
3818 29. The long-running daemon process does not normally write to the log once it
3819 has entered its main loop, and it closes the log before doing so. This is
3820 so that log files can straightforwardly be renamed and moved. However,
3821 there are a couple of unusual error situations where the daemon does write
3822 log entries, and I had neglected to close the log afterwards.
3824 30. The text of an SMTP error response that was received during a remote
3825 delivery was being truncated at 512 bytes. This is too short for some of
3826 the long messages that one sometimes sees. I've increased the limit to
3829 31. It is now possible to make retry rules that apply only when a message has a
3830 specific sender, in particular, an empty sender.
3832 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3833 it possible to be selective about when SMTP synchronization is enforced.
3835 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3837 32. Implemented hosts_connection_nolog.
3839 33. Added an ACL for QUIT.
3841 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3844 35. Added mailbox_size and mailbox_filecount to appendfile.
3846 36. Added control = no_multiline_responses to ACLs.
3848 37. There was a bug in the logic of the code that waits for the clock to tick
3849 in the case where the clock went backwards by a substantial amount such
3850 that the microsecond fraction of "now" was more than the microsecond
3851 fraction of "then" (but the whole seconds number was less).
3853 38. Added support for the libradius Radius client library this is found on
3854 FreeBSD (previously only the radiusclient library was supported).
3860 1. When certain lookups returned multiple values in the form name=value, the
3861 quoting of the values was not always being done properly. Specifically:
3862 (a) If the value started with a double quote, but contained no whitespace,
3864 (b) If the value contained whitespace other than a space character (i.e.
3865 tabs or newlines or carriage returns) it was not quoted.
3866 This fix has been applied to the mysql and pgsql lookups by writing a
3867 separate quoting function and calling it from the lookup code. The fix
3868 should probably also be applied to nisplus, ibase and oracle lookups, but
3869 since I cannot test any of those, I have not disturbed their existing code.
3871 2. A hit in the callout cache for a specific address caused a log line with no
3872 reason for rejecting RCPT. Now it says "Previous (cached) callout
3873 verification failure".
3875 3. There was an off-by-one bug in the queryprogram router. An over-long
3876 return line was truncated at 256 instead of 255 characters, thereby
3877 overflowing its buffer with the terminating zero. As well as fixing this, I
3878 have increased the buffer size to 1024 (and made a note to document this).
3880 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3881 when Exim is waiting for an SMTP response from a remote server, Exim
3882 restarts its select() call on the socket, thereby resetting its timeout.
3883 This is not a problem when such interrupts are rare. Somebody set up a cron
3884 job to run exiwhat every 2 minutes, which is less than the normal select()
3885 timeout (5 or 10 minutes). This meant that the select() timeout never
3886 kicked in because it was always reset. I have fixed this by comparing the
3887 time when an interrupt arrives with the time at the start of the first call
3888 to select(). If more time than the timeout has elapsed, the interrupt is
3889 treated as a timeout.
3891 5. Some internal re-factoring in preparation for the addition of Sieve
3892 extensions (by MH). In particular, the "personal" test is moved to a
3893 separate function, and given an option for scanning Cc: and Bcc: (which is
3894 not set for Exim filters).
3896 6. When Exim created an email address using the login of the caller as the
3897 local part (e.g. when creating a From: or Sender: header line), it was not
3898 quoting the local part when it contained special characters such as @.
3900 7. Installed new OpenBSD configuration files.
3902 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3903 try to make them clearer.
3905 9. Callout options, other than the timeout value, were being ignored when
3906 verifying sender addresses in header lines. For example, when using
3908 verify = header_sender/callout=no_cache
3910 the cache was (incorrectly) being used.
3912 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3913 only the Cygwin environment.
3915 11. When return_path_on_delivery was set as a log selector, if different remote
3916 addresses in the same message used different return paths and parallel
3917 remote delivery occurred, the wrong values would sometimes be logged.
3918 (Whenever a remote delivery process finished, the return path value from
3919 the most recently started remote delivery process was logged.)
3921 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3922 lines when AUTH and/or TLS are in use. This is the "received protocol"
3923 field. Exim used to use "asmtp" for authenticated SMTP, without any
3924 indication (in the protocol name) for TLS use. Now it follows the RFC and
3925 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3926 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3927 names appear in log lines as well as in Received: header lines.
3929 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3930 extensions, and comparison tests, and to fix some bugs.
3932 14. Changes to the "personal" filter test:
3934 (1) The test was buggy in that it was just doing the equivalent of
3935 "contains" tests on header lines. For example, if a user's address was
3936 anne@some.where, the "personal" test would incorrectly be true for
3938 To: susanne@some.where
3940 This test is now done by extracting each address from the header in turn,
3941 and checking the entire address. Other tests that are part of "personal"
3942 are now done using regular expressions (for example, to check local parts
3943 of addresses in From: header lines).
3945 (2) The list of non-personal local parts in From: addresses has been
3946 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3947 taken from the Sieve specification recommendations.
3949 (3) If the message contains any header line starting with "List-" it is
3950 treated as non-personal.
3952 (4) The test for "circular" in the Subject: header line has been removed
3953 because it now seems ill-conceived.
3955 15. Minor typos in src/EDITME comments corrected.
3957 16. Installed latest exipick from John Jetmore.
3959 17. If headers_add on a router specified a text string that was too long for
3960 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3961 of string_sprintf() is now avoided.
3963 18. $message_body_size was not set (it was always zero) when running the DATA
3964 ACL and the local_scan() function.
3966 19. For the "mail" command in an Exim filter, no default was being set for
3967 the once_repeat time, causing a random time value to be used if "once" was
3968 specified. (If the value happened to be <= 0, no repeat happened.) The
3969 default is now 0s, meaning "never repeat". The "vacation" command was OK
3970 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3971 (I found it when inspecting the code).
3973 20. There is now an overall timeout for performing a callout verification. It
3974 defaults to 4 times the callout timeout, which applies to individual SMTP
3975 commands during the callout. The overall timeout applies when there is more
3976 than one host that can be tried. The timeout is checked before trying the
3977 next host. This prevents very long delays if there are a large number of
3978 hosts and all are timing out (e.g. when the network connections are timing
3979 out). The value of the overall timeout can be changed by specifying an
3980 additional sub-option for "callout", called "maxwait". For example:
3982 verify = sender/callout=5s,maxwait=20s
3984 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3985 to the end before writing, but this should make it even safer).
3987 22. Exim was forgetting that it had advertised PIPELINING for the second and
3988 subsequent messages on an SMTP connection. It was also not resetting its
3989 memory on STARTTLS and an internal HELO.
3991 23. When Exim logs an SMTP synchronization error within a session, it now
3992 records whether PIPELINING has been advertised or not.
3994 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3995 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3996 rather than long int.
3998 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4000 26. Added the never_mail option to autoreply.
4006 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4007 crash if the getsockname() call failed; this can happen if a connection is
4008 closed very soon after it is established. The problem was simply in the
4009 order in which certain operations were done, causing Exim to try to write
4010 to the SMTP stream before it had set up the file descriptor. The bug has
4011 been fixed by making things happen in the correct order.
4017 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4018 before the connection was closed, thus losing the rejection response.
4020 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4021 some early Solaris releases, but causes trouble in current releases where
4022 socklen_t is defined.
4024 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4027 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4030 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4031 that releasing the top of it at the end releases what was used for sub-
4032 expansions (unless the block got too big). However, discard this block if
4033 the first thing is a variable or header, so that we can use its block when
4034 it is dynamic (useful for very large $message_headers, for example).
4036 6. Lookups now cache *every* query, not just the most recent. A new, separate
4037 store pool is used for this. It can be recovered when all lookup caches are
4038 flushed. Lookups now release memory at the end of their result strings.
4039 This has involved some general refactoring of the lookup sources.
4041 7. Some code has been added to the store_xxx() functions to reduce the amount
4042 of flapping under certain conditions.
4044 8. log_incoming_interface used to affect only the <= reception log lines. Now
4045 it causes the local interface and port to be added to several more SMTP log
4046 lines, for example "SMTP connection from", and rejection lines.
4048 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4050 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4052 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4053 could be overwritten at the end of the current message (or the start of a
4054 new message if it was set in a HELO ACL). The value is now preserved for
4055 the duration of the SMTP connection.
4057 12. If a transport had a headers_rewrite setting, and a matching header line
4058 contained an unqualified address, that address was qualified, even if it
4059 did not match any rewriting rules. The underlying bug was that the values
4060 of the flags that permit the existence of unqualified sender and recipient
4061 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4062 non-local messages, and by -bnq for local messages) were not being
4063 preserved with the message after it was received.
4065 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4066 "next input=" as part of the text comprising the host identity instead of
4067 the correct text. The code was using the same buffer for two different
4068 strings. However, depending on which order the printing function evaluated
4069 its arguments, the bug did not always show up. Under Linux, for example, my
4070 test suite worked just fine.
4072 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4073 doesn't work with some older versions of Perl. It has been changed to "my",
4074 which in any case is probably the better facility to use.
4076 15. A really picky compiler found some instances of statements for creating
4077 error messages that either had too many or two few arguments for the format
4080 16. The size of the buffer for calls to the DNS resolver has been increased
4081 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4082 for addresses that have a lot of PTR records. This alleviates a problem; it
4083 does not fully solve it.
4085 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4086 buffer now truncates the list and logs the incident, which is the same
4087 action as happens when Exim is looking up a host name and its aliases.
4088 Previously in this situation something unpredictable would happen;
4089 sometimes it was "internal error: store_reset failed".
4091 18. If a server dropped the connection unexpectedly when an Exim client was
4092 using GnuTLS and trying to read a response, the client delivery process
4093 crashed while trying to generate an error log message.
4095 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4096 string, for example:
4098 warn message = H1: something\nH2: something
4100 the text was added as a single header line from Exim's point of view
4101 though it ended up OK in the delivered message. However, searching for the
4102 second and subsequent header lines using $h_h2: did not work. This has been
4103 fixed. Similarly, if a system filter added multiple headers in this way,
4104 the routers could not see them.
4106 20. Expanded the error message when iplsearch is called with an invalid key to
4107 suggest using net-iplsearch in a host list.
4109 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4110 ACLs are no longer actually imposed (and a message to that effect is
4113 22. If a "gecos" field in a passwd entry contained escaped characters, in
4114 particular, if it contained a \" sequence, Exim got it wrong when building
4115 a From: or a Sender: header from that name. A second bug also caused
4116 incorrect handling when an unquoted " was present following a character
4117 that needed quoting.
4119 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4120 was not being matched caselessly.
4122 24. Arranged for all hyphens in the exim.8 source to be escaped with
4125 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4126 properly. Recipient callout cache records were still being keyed to include
4127 the sender, even when use_sender was set false. This led to far more
4128 callouts that were necessary. The sender is no longer included in the key
4129 when use_sender is false.
4131 26. Added "control = submission" modifier to ACLs.
4133 27. Added the ${base62d: operator to decode base 62 numbers.
4135 28. dnsdb lookups can now access SRV records.
4137 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4138 the configuration file.
4140 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4141 (-v) mode. This makes the output for a verbose queue run more intelligible.
4143 31. Added a use_postmaster feature to recipient callouts.
4145 32. Added the $body_zerocount variable, containing the number of binary zero
4146 bytes in the message body.
4148 33. The time of last modification of the "new" subdirectory is now used as the
4149 "mailbox time last read" when there is a quota error for a maildir
4152 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4154 35. Added +ignore_unknown as a special item in host lists.
4156 36. Code for decoding IPv6 addresses in host lists is now included, even if
4157 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4158 address was recognized as an IP address, but was then not correctly decoded
4159 into binary, causing unexpected and incorrect effects when compared with
4166 1. Very minor rewording of debugging text in manualroute to say "list of
4167 hosts" instead of "hostlist".
4169 2. If verify=header_syntax was set, and a header line with an unqualified
4170 address (no domain) and a large number of spaces between the end of the
4171 name and the colon was received, the reception process suffered a buffer
4172 overflow, and (when I tested it) crashed. This was caused by some obsolete
4173 code that should have been removed. The fix is to remove it!
4175 3. When running in the test harness, delay a bit after writing a bounce
4176 message to get a bit more predictability in the log output.
4178 4. Added a call to search_tidyup() just before forking a reception process. In
4179 theory, someone could use a lookup in the expansion of smtp_accept_max_
4180 per_host which, without the tidyup, could leave open a database connection.
4182 5. Added the variables $recipient_data and $sender_data which get set from a
4183 lookup success in an ACL "recipients" or "senders" condition, or a router
4184 "senders" option, similar to $domain_data and $local_part_data.
4186 6. Moved the writing of debug_print from before to after the "senders" test
4189 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4190 problems for message scanning, either using a data ACL, or using
4191 local_scan() because the Received: header was not generated till after they
4192 were called (in order to set the time as the time of reception completion).
4193 I have revised the way this works. The header is now generated after the
4194 body is received, but before the ACL or local_scan() are called. After they
4195 are run, the timestamp in the header is updated.
4201 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4202 before starting a queue runner without re-exec. This happened only when
4203 deliver_drop_privilege was set or when the Exim user was set to root. The
4204 effect of the bug was that timeouts during subsequent deliveries caused
4205 crashes instead of being properly handled. The handler is now left at its
4206 default (and expected) setting.
4208 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4209 message, again when deliver_drop_privilege is set or Exim is run as root.
4210 The bug described in (1) was not present in this case, but the tidying up
4211 of the other signals was missing. I have made the two cases consistent.
4213 3. The ignore_target_hosts setting on a manualroute router was being ignored
4214 for hosts that were looked up using the /MX notation.
4216 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4219 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4220 operated on the sender address. After changing the $sender_address to <>
4221 for the sender address verify, Exim was re-instated it as the original
4222 (before rewriting) address, but remembering that it had rewritten it, so it
4223 wasn't rewriting it again. This bug also had the effect of breaking the
4224 sender address verification caching when the sender address was rewritten.
4226 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4227 This has been changed so that if the ip literal address matches
4228 ignore_target_hosts, the router declines.
4230 7. Added expansion conditions match_domain, match_address, and match_local_
4231 part (NOT match_host).
4233 8. The placeholder for the Received: header didn't have a length field set.
4235 9. Added code to Exim itself and to exim_lock to test for a specific race
4236 condition that could lead to file corruption when using MBX delivery. The
4237 issue is with the lockfile that is created in /tmp. If this file is removed
4238 after a process has opened it but before that process has acquired a lock,
4239 there is the potential for a second process to recreate the file and also
4240 acquire a lock. This could lead to two Exim processes writing to the file
4241 at the same time. The added code performs the same test as UW imapd; it
4242 checks after acquiring the lock that its file descriptor still refers to
4243 the same named file.
4245 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4246 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4249 11. Added the smtp_active_hostname option. If used, this will typically be made
4250 to depend on the incoming interface address. Because $interface_address is
4251 not set up until the daemon has forked a reception process, error responses
4252 that can happen earlier (such as "too many connections") no longer contain
4255 12. If an expansion in a condition on a "warn" statement fails because a lookup
4256 defers, the "warn" statement is abandoned, and the next ACL statement is
4257 processed. Previously this caused the whole ACL to be aborted.
4259 13. Added the iplsearch lookup type.
4261 14. Added ident_timeout as a log selector.
4263 15. Added tls_certificate_verified as a log selector.
4265 16. Added a global option tls_require_ciphers (compare the smtp transport
4266 option of the same name). This controls incoming TLS connections.
4268 17. I finally figured out how to make tls_require_ciphers do a similar thing
4269 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4270 before starting the TLS session.
4272 18. Tabs are now shown as \t in -bP output.
4274 19. If the log selector return_path_on_delivery was set, Exim crashed when
4275 bouncing a message because it had too many Received: header lines.
4277 20. If two routers both had headers_remove settings, and the first one included
4278 a superfluous trailing colon, the final name in the first list and the
4279 first name in the second list were incorrectly joined into one item (with a
4280 colon in the middle).
4286 1. Added -C and -D options to the exinext utility, mainly to make it easier
4287 to include in the automated testing, but these could be helpful when
4288 multiple configurations are in use.
4290 2. The exinext utility was not formatting the output nicely when there was
4291 an alternate port involved in the retry record key, nor when there was a
4292 message id as well (for retries that were specific to a specific message
4293 and a specific host). It was also confused by IPv6 addresses, because of
4294 the additional colons they contain. I have fixed the IPv4 problem, and
4295 patched it up to do a reasonable job for IPv6.
4297 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4298 delivery, the log line now contains "pipelined" if PIPELINING was used.
4300 4. An SMTP transport process used to panic and die if the bind() call to set
4301 an explicit outgoing interface failed. This has been changed; it is now
4302 treated in the same way as a connect() failure.
4304 5. A reference to $sender_host_name in the part of a conditional expansion
4305 that was being skipped was still causing a DNS lookup. This no longer
4308 6. The def: expansion condition was not recognizing references to header lines
4309 that used bh_ and bheader_.
4311 7. Added the _cache feature to named lists.
4313 8. The code for checking quota_filecount in the appendfile transport was
4314 allowing one more file than it should have been.
4316 9. For compatibility with Sendmail, the command line option
4324 and sets the incoming protocol and host name (for trusted callers). The
4325 host name and its colon can be omitted when only the protocol is to be set.
4326 Note the Exim already has two private options, -pd and -ps, that refer to
4327 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4328 "s", but I don't think that's a major issue.
4330 10. A number of refactoring changes to the code, none of which should affect
4333 (a) The number of logging options was getting close to filling up the
4334 32-bit word that was used as a bit map. I have split them into two classes:
4335 those that are passed in the argument to log_write(), and those that are
4336 only ever tested independently outside of that function. These are now in
4337 separate 32-bit words, so there is plenty of room for expansion again.
4338 There is no change in the user interface or the logging behaviour.
4340 (b) When building, for example, log lines, the code previously used a
4341 macro that called string_cat() twice, in order to add two strings. This is
4342 not really sufficiently general. Furthermore, there was one instance where
4343 it was actually wrong because one of the argument was used twice, and in
4344 one call a function was used. (As it happened, calling the function twice
4345 did not affect the overall behaviour.) The macro has been replaced by a
4346 function that can join an arbitrary number of extra strings onto a growing
4349 (c) The code for expansion conditions now uses a table and a binary chop
4350 instead of a serial search (which was left over from when there were very
4351 few conditions). Also, it now recognizes conditions like "pam" even when
4352 the relevant support is not compiled in: a suitably worded error message is
4353 given if an attempt is made to use such a condition.
4355 11. Added ${time_interval:xxxxx}.
4357 12. A bug was causing one of the ddress fields not to be passed back correctly
4358 from remote delivery subprocesses. The field in question was not being
4359 subsequently used, so this caused to problems in practice.
4361 13. Added new log selectors queue_time and deliver_time.
4363 14. Might have fixed a bug in maildirsizefile handling that threw up
4364 "unexpected character" debug warnings, and recalculated the data
4365 unnecessarily. In any case, I expanded the warning message to give more
4368 15. Added the message "Restricted characters in address" to the statements in
4369 the default ACL that block characters like @ and % in local parts.
4371 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4372 Three changes have been made:
4374 (a) There was a serious bug; a negative response to MAIL caused the whole
4375 recipient domain to be cached as invalid, thereby blocking all messages
4376 to all local parts at the same domain, from all senders. This bug has
4377 been fixed. The domain is no longer cached after a negative response to
4378 MAIL if the sender used is not empty.
4380 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4383 (c) A new callout option, "use_sender" has been added for people who want
4384 the modified behaviour.
4390 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4393 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4394 indeed breaks things for older releases.
4396 3. Added additional logging to the case where there is a problem reading data
4397 from a filter that is running in a subprocess using a pipe, in order to
4398 try to track down a specific problem.
4400 4. Testing facility fudge: when running in the test harness and attempting
4401 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4402 getting "No route to host". Convert this to a timeout.
4404 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4407 6. Some OS don't have socklen_t but use size_t instead. This affects the
4408 fifth argument of getsockopt() amongst other things. This is now
4409 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4410 can be set for individual OS. I have set it for SunOS5, OSF1, and
4411 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4412 some earlier ones do not.
4414 7. Change 4.30/15 was not doing the test caselessly.
4416 8. The standard form for an IPv6 address literal was being rejected by address
4417 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4418 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4419 this, as well as the form without the "IPv6" on the front (but only when
4420 address literals are enabled, of course).
4422 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4424 10. Exim crashed if a message with an empty sender address specified by -f
4425 encountered a router with an errors_to setting. This could be provoked only
4426 by a command such as
4430 where an empty string was supplied; "<>" did not hit this bug.
4432 11. Installed PCRE release 4.5.
4434 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4435 remained set. It is now erased.
4437 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4438 times from message ids (which are base 36 rather than the normal 62).
4440 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4441 were being counted as actual protocol errors, and logged if the log
4442 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4443 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4444 and DATA following a set of rejected RCPTs do not count as protocol errors.
4445 In other words, Exim assumes they were pipelined, though this may not
4446 actually be the case. Of course, in all cases the client gets an
4447 appropriate error code.
4449 15. If a lookup fails in an ACL condition, a message about the failure may
4450 be available; it is used if testing the ACL cannot continue, because most
4451 such messages specify what the cause of the deferral is. However, some
4452 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4453 that caused an old message to be retained and used if a later statement
4454 caused a defer, replacing the real cause of the deferral.
4456 16. If an IP address had so many PTR records that the DNS lookup buffer
4457 was not large enough to hold them, Exim could crash while trying to process
4458 the truncated data. It now detects and logs this case.
4460 17. Further to 4.21/58, another change has been made: if (and only if) the
4461 first line of a message (the first header line) ends with CRLF, a bare LF
4462 in a subsequent header line has a space inserted after it, so as not to
4463 terminate the header.
4465 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4466 unnecessarily, used atoi() instead of strtol(), and didn't check the
4467 termination when getting file sizes from file names by regex.
4469 19. Completely re-implemented the support for maildirsize files, in the light
4470 of a number of problems with the previous contributed implementation
4471 (4.30/29). In particular:
4473 . If the quota is zero, the maildirsize file is maintained, but no quota is
4476 . If the maildir directory does not exist, it is created before any attempt
4477 to write a maildirsize file.
4479 . The quota value in the file is just a cache; if the quota is changed in
4480 the transport, the new value overrides.
4482 . A regular expression is available for excluding directories from the
4485 20. The autoreply transport checks the characters in options that define the
4486 message's headers; it allows continued headers, but it was checking with
4487 isspace() after an embedded newline instead of explicitly looking for a
4490 21. If all the "regular" hosts to which an address was routed had passed their
4491 expiry times, and had not reached their retry times, the address was
4492 bounced, even if fallback hosts were defined. Now Exim should go on to try
4495 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4496 equivalent code in the SMTP transport. Some hosts send humungous responses
4497 to HELO/EHLO, more than 1024 it seems.
4499 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4500 gives compiler warnings in some environments. I've now done it "properly",
4503 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4504 (because of problems with the built-in one) was declared to return uschar *
4505 instead of char *, causing compiler failure.
4507 25. Fixed a file descriptor leak when processing alias/forward files.
4509 26. Fixed a minor format string issue in dbfn.c.
4511 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4513 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4514 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4517 29. When a sender address is verified, it is cached, to save repeating the test
4518 when there is more than one recipient in a message. However, when the
4519 verification involves a callout, it is possible for different callout
4520 options to be set for different recipients. It is too complicated to keep
4521 track of this in the cache, so now Exim always runs a verification when a
4522 callout is required, relying on the callout cache for the optimization.
4523 The overhead is duplication of the address routing, but this should not be
4526 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4527 to be verified with callout=postmaster, and the main callout worked but the
4528 postmaster check failed, the verification correctly failed. However, if a
4529 subsequent RCPT command asked for sender verification *without* the
4530 postmaster check, incorrect caching caused this verification also to fail,
4533 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4534 it was not caching the DNS options (qualify_single, search_parents) that
4535 were used when the lookup failed. A subsequent lookup with different
4536 options therefore always gave the same answer, though there were cases
4537 where it should not have. (Example: a "domains = !$mx_any" option on a
4538 dnslookup router: the "domains" option is always processed without any
4539 widening, but the router might have qualify_single set.) Now Exim uses the
4540 cached value only when the same options are set.
4542 32. Added John Jetmore's "exipick" utility to the distribution.
4544 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4545 than a timeout (e.g. a certificate is required, and is not provided), an
4546 Exim server now closes the connection immediately. Previously it waited for
4547 the client to close - but if the client is SSL, it seems that they each
4548 wait for each other, leading to a delay before one of them times out.
4550 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4551 maintained 0.8.x compatibility because I don't think many are using it, and
4552 it is clearly obsolete.
4554 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4557 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4558 client certificate was expired. A simple patch fixes this, though I don't
4559 understand the full logic of why the verify callback is called multiple
4562 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4563 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4564 which causes problems with some clients (such as the Certicom SSL Plus
4565 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4566 disables the coutermeasure allowing Eudora to connect."
4568 38. Exim was not checking that a write() to a log file succeeded. This could
4569 lead to Bad Things if a log got too big, in particular if it hit a file
4570 size limit. Exim now panics and dies if it cannot write to a log file, just
4571 as it does if it cannot open a log file.
4573 39. Modified OS/Makefile-Linux so that it now contains
4575 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4577 The two -D definitions ensure that Exim is compiled with large file
4578 support, which makes it possible to handle log files that are bigger than
4581 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4582 instance) a domain was checked against a named list that involved a lookup,
4583 causing $domain_data to be set, then another domain was checked against the
4584 same list, then the first domain was re-checked, the value of $domain_data
4585 after the final check could be wrong. In particular, if the second check
4586 failed, it could be set empty. This bug probably also applied to
4589 41. The strip_trailing_dot option was not being applied to the address given
4590 with the -f command-line option.
4592 42. The code for reading a message's header from the spool was incrementing
4593 $received_count, but never initializing it. This meant that the value was
4594 incorrect (doubled) while delivering a message in the same process in which
4595 it was received. In the most common configuration of Exim, this never
4596 happens - a fresh exec is done - but it can happen when
4597 deliver_drop_privilege is set.
4599 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4600 it now includes up to 150 characters of the unexpected data in the log
4603 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4604 and building data strings. The size of both of these buffers was 10 000
4605 bytes - far larger than anybody would *ever* want, thought I. Needless to
4606 say, somebody hit the limit. I have increased the maximum line length to
4607 20 000 and the maximum data length of concatenated lines to 100 000. I have
4608 also fixed two bugs, because there was no checking on these buffers. Tsk,
4609 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4610 buffer is too small.
4612 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4613 lsearch lookups. Now it does.
4615 46. When parsing a route_list item in a manualroute router, a fixed-length
4616 buffer was used for the list of hosts. I made this 1024 bytes long,
4617 thinking that nobody would ever have a list of hosts that long. Wrong.
4618 Somebody had a whole pile of complicated expansion conditions, and the
4619 string was silently truncated, leading to an expansion error. It turns out
4620 that it is easier to change to an unlimited length (owing to other changes
4621 that have happened since this code was originally written) than to build
4622 structure for giving a limitation error. The length of the item that
4623 expands into the list of hosts is now unlimited.
4625 47. The lsearch lookup could not handle data where the length of text line was
4626 more than 4095 characters. Such lines were truncated, leading to shortened
4627 data being returned. It should now handle lines of any length.
4629 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4630 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4633 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4634 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4635 no longer happens in comment lines. A list of replacements is now placed
4636 at the head of all of the source files, except those whose only change is
4637 to replace PERL_COMMAND in the very first #! line.
4639 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4640 messages on the queue, with a bottom-up merge sort, using code contributed
4641 by Michael Haardt. This should make operations like -bp somewhat faster on
4642 large queues. It won't affect queue runners, except when queue_run_in_order
4645 51. Installed eximstats 1.31 in the distribution.
4647 52. Added support for SRV lookups to the dnslookup router.
4649 53. If an ACL referred to $message_body or $message_body_end, the value was not
4650 reset for any messages that followed in the same SMTP session.
4652 54. The store-handling optimization for building very long strings was not
4653 differentiating between the different store pools. I don't think this
4654 actually made any difference in practice, but I've tidied it.
4656 55. While running the routers to verify a sender address, $sender_address
4657 was still set to the sender address. This is wrong, because when routing to
4658 send a bounce to the sender, it would be empty. Therefore, I have changed
4659 it so that, while verifying a sender address, $sender_address is set to <>.
4660 (There is no change to what happens when verifying a recipient address.)
4662 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4663 target A or AAAA records (if not already returned) without resetting the
4664 qualify_single or search_parents options of the DNS resolver. These are
4665 inappropriate in this case because the targets of MX and SRV records must
4666 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4667 target that, when qualified, matched something in the local domain. These
4668 two options are now turned off when doing these lookups.
4670 57. It seems that at least some releases of Reiserfs (which does not have the
4671 concept of a fixed number of inodes) returns zero and not -1 for the
4672 number of available inodes. This interacted badly with check_spool_inodes,
4673 which assumed that -1 was the "no such thing" setting. What I have done is
4674 to check that the total number of inodes is greater than zero before doing
4675 the test of how many are available.
4677 58. When a "warn" ACL statement has a log_message modifier, the message is
4678 remembered, and not repeated. This is to avoid a lot of repetition when a
4679 message has many recipients that cause the same warning to be written.
4680 Howewer, Exim was preserving the list of already written lines for an
4681 entire SMTP session, which doesn't seem right. The memory is now reset if a
4682 new message is started.
4684 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4685 debugging output unless log_rewrite was also set.
4687 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4688 of (int)(handle) when we know that handle contains (void *)(-1).
4690 61. The Exim daemon panic-logs an error return when it closes the incoming
4691 connection. However "connection reset by peer" seems to be common, and
4692 isn't really an error worthy of noting specially, so that particular error
4695 62. When Exim is trying to find all the local interfaces, it used to panic and
4696 die if the ioctl to get the interface flags failed. However, it seems that
4697 on at least one OS (Solaris 9) it is possible to have an interface that is
4698 included in the list of interfaces, but for which you get a failure error
4699 for this call. This happens when the interface is not "plumbed" into a
4700 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4701 failure of the "get flags" call assumes that the interface is down.
4703 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4704 makes life easier for people who are doing arithmetic on fields extracted
4705 from dates, where you often get leading zeros that should not be
4706 interpreted as octal.
4708 64. Added qualify_domain to the redirect router, to override the global
4711 65. If a pathologically long header line contained very many addresses (the
4712 report of this problem mentioned 10 000) and each of them was rewritten,
4713 Exim could use up a very large amount of memory. (It kept on making new
4714 copies of the header line as it rewrote, and never released the old ones.)
4715 At the expense of a bit more processing, the header rewriting function has
4716 been changed so that it no longer eats memory in this way.
4718 66. The generation of the Received: header has been moved from the time that a
4719 message starts to be received, to the time that it finishes. The timestamp
4720 in the Received: header should now be very close to that of the <= log
4721 line. There are two side-effects of this change:
4723 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4724 logged header lines no longer include the local Received: line, because
4725 it has not yet been created. The same applies to a copy of the message
4726 that is returned to a non-SMTP sender when a message is rejected.
4728 (b) When a filter file is tested using -bf, no additional Received: header
4729 is added to the test message. After some thought, I decided that this
4732 This change does not affect the value of $received_for. It is still set
4733 after address rewriting, but before local_scan() is called.
4735 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4737 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4738 gave an unhelpful panic error message, and a defer error. I have managed to
4739 change this behaviour so that it now rejects any supplied certificate,
4740 which seems right, as the list of acceptable certificates is empty.
4742 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4743 gave an unhelpful defer error. I have not managed to make this reject any
4744 supplied certificates, but the error message it gives is "no certificate
4745 supplied", which is not helpful.
4747 70. exigrep's output now also includes lines that are not associated with any
4748 message, but which match the given pattern. Implemented by a patch from
4749 Martin Sluka, which also tidied up the Perl a bit.
4751 71. Recipient callout verification, like sender verification, was using <> in
4752 the MAIL FROM command. This isn't really the right thing, since the actual
4753 sender may affect whether the remote host accepts the recipient or not. I
4754 have changed it to use the actual sender in the callout; this means that
4755 the cache record is now keyed on a recipient/sender pair, not just the
4756 recipient address. There doesn't seem to be a real danger of callout loops,
4757 since a callout by the remote host to check the sender would use <>.
4758 [SEE ABOVE: changed after hitting problems.]
4760 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4761 temporary errors. However, in the case of such a code being given after
4762 the end of a data transmission (i.e. after ".") Exim was failing to write
4763 a retry record for the message. (Yes, there was some broken host that was
4764 actually sending 8xx at this point.)
4766 73. An unknown lookup type in a host list could cause Exim to panic-die when
4767 the list was checked. (An example that provoked this was putting <; in the
4768 middle of a list instead of at the start.) If this happened during a DATA
4769 ACL check, a -D file could be left lying around. This kind of configuration
4770 error no longer causes Exim to die; instead it causes a defer errror. The
4771 incident is still logged to the main and panic logs.
4773 74. Buglet left over from Exim 3 conversion. The message "too many messages
4774 in one connection" was written to the rejectlog but not the mainlog, except
4775 when address rewriting (yes!) was being logged.
4777 75. Added write_rejectlog option.
4779 76. When a system filter was run not as root (that is, when system_filter_user
4780 was set), the values of the $n variables were not being returned to the
4781 main process; thus, they were not subsequently available in the $sn
4784 77. Added +return_path_on_delivery log selector.
4786 78. A connection timeout was being treated differently from recipients deferred
4787 when testing hosts_max_try with a message that was older than the host's
4788 retry timeout. (The host should not be counted, thus allowing all hosts to
4789 be tried at least once before bouncing.) This may have been the cause of an
4790 occasionally reported bug whereby a message would remain on the queue
4791 longer than the retry timeout, but would be bounced if a delivery was
4792 forced. I say "may" because I never totally pinned down the problem;
4793 setting up timeout/retry tests is difficult. See also the next item.
4795 79. The ultimate address timeout was not being applied to errors that involved
4796 a combination of host plus message (for example, a timeout on a MAIL
4797 command). When an address resolved to a number of possible hosts, and they
4798 were not all tried for each delivery (e.g. because of hosts_max_try), a
4799 message could remain on the queue longer than the retry timeout.
4801 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4804 81. Fixed an obscure SMTP outgoing bug which required at least the following
4805 conditions: (a) there was another message waiting for the same server;
4806 (b) the server returned 5xx to all RCPT commands in the first message so
4807 that the message was not completed; (c) the server dropped the connection
4808 or gave a negative response to the RSET that Exim sends to abort the
4809 transaction. The observed case was a dropped connection after DATA that had
4810 been sent in pipelining mode. That is, the server had advertised PIPELINING
4811 but was not implementing it correctly. The effect of the bug was incorrect
4812 behaviour, such as trying another host, and this could lead to a crash.
4818 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4819 and daemon.c were passed as pointers to ints; they should have been
4820 pointers to socklen_t variables (which are typically unsigned ints).
4822 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4825 3. Fixed a really odd bug that affected only the testing scheme; patching a
4826 certain fixed string in the binary changed the value of another string that
4827 happened to be identical to the end of the original first string.
4829 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4830 name", it returns that address as the IP address. On some operating
4831 systems (e.g. Solaris), it also passes back the IP address string as the
4832 "host name". However, on others (e.g. Linux), it passes back an empty
4833 string. Exim wasn't checking for this, and was changing the host name to an
4834 empty string, assuming it had been canonicized.
4836 5. Although rare, it is permitted to have more than one PTR record for a given
4837 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4838 all the names associated with an address, because they do in Solaris.
4839 However, it seems that they do not in Linux for data that comes from the
4840 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4841 I found this out when I moved to a new Linux workstation and tried to run
4842 the Exim test suite.
4844 To get round this problem I have changed the code so that it now does its
4845 own call to the DNS to look up PTR records when searching for a host name.
4846 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4847 addresses that are only in /etc/hosts are still found.
4849 This behaviour is, however, controlled by an option called host_lookup_
4850 order, which defaults to "bydns:byaddr". If people want to use the other
4851 order, or indeed, just use one or the other means of lookup, they can
4852 specify it in this variable.
4854 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4855 some operating systems, this comes back from gethostbyaddr() as an empty
4856 string, and this is what Exim used to test for. However, it seems that in
4857 other systems, "." is yielded. Exim now tests for this case too.
4859 7. The values of check_spool_space and check_log_space are now held internally
4860 as a number of kilobytes instead of an absolute number of bytes. If a
4861 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4862 kilobyte. This means that much larger values can be stored.
4864 8. Exim monitor: an attempt to get the action menu when not actually pointing
4865 at a message produces an empty menu entitled "No message selected". This
4866 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4867 no entries in it ("Shell widget menu has zero width and/or height"). So I
4868 have added a single, blank menu entry in this case.
4870 9. Added ${quote_local_part.
4872 10. MIME decoding is now applied to the contents of Subject: header lines when
4875 11. Now that a reference to $sender_host_address automatically causes a reverse
4876 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4877 host lookup before query-style lookups in lists that might use this
4878 variable. This has therefore been abolished, and the "net-" prefix is no
4879 longer necessary for query-style lookups.
4881 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4882 have been a typo for LFLAGS, so it has been changed.
4884 13. The install script calls Exim with "-C /dev/null" in order to find the
4885 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4886 to be output. Howeve, since Exim outputs its version number before the
4887 error, it didn't break the script. It just looked ugly. I fixed this by
4888 always allowing "-C /dev/null" if the caller is root.
4890 14. Ignore overlarge ACL variable number when reading spool file - insurance
4891 against a later release with more variables having written the file.
4893 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4894 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4895 this, as well as the form without the "IPv6" on the front.
4897 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4898 OS/Makefile-Darwin file.
4900 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4901 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4903 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4904 variable. However, some OS use ints for inodes. I've added cast to long int
4905 to get rid of the compiler warning.
4907 19. I had forgotten to lock out "/../" in configuration file names when
4908 ALT_CONFIG_PREFIX was set.
4910 20. Routers used for verification do not need to specify transports. However,
4911 if such a router generated a host list, and callout was configured, Exim
4912 crashed, because it could not find a port number from the (non-existent)
4913 transport. It now assumes port 25 in this circumstance.
4915 21. Added the -t option to exigrep.
4917 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4918 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4919 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4920 them is set, LOOKUP_LSEARCH is forced.
4922 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4924 24. Added sender and host information to the "rejected by local_scan()" log
4925 line; previously there was no indication of these.
4927 25. Added .include_if_exists.
4929 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4930 Linux. It turns out that not all file systems support this. Apparently some
4931 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4932 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4933 not supported on the file descriptor, is now ignored when Exim is trying to
4934 sync a directory. This applies only to Linux.
4936 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4938 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4939 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4942 29. Added support for maildirsize files from supplied patch (modified a bit).
4944 30. The use of :fail: followed by an empty string could lead Exim to respond to
4945 sender verification failures with (e.g.):
4947 550 Verification failed for <xxx>
4948 550 Sender verify failed
4950 where the first response line was missing the '-' that indicates it is not
4951 the final line of the response.
4953 31. The loop for finding the name of the user that called Exim had a hardwired
4954 limit of 10; it now uses the value of finduser_retries, which is used for
4955 all other user lookups.
4957 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4960 33. Exim was neglecting to zero errno before one call of strtol() when
4961 expanding a string and expecting an integer value. On some systems this
4962 resulted in spurious "integer overflow" errors. Also, it was casting the
4963 result into an int without checking.
4965 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4966 did not work. The code looks as if it has *never* worked, though it appears
4967 to have been documented since at least releast 1.62. I have made it work.
4969 35. The "timeout_DNS" error in retry rules, also documented since at least
4970 1.62, also never worked. As it isn't clear exactly what this means, and
4971 clearly it isn't a major issue, I have abolished the feature by treating it
4972 as "timeout", and writing a warning to the main and panic logs.
4974 36. The display of retry rules for -brt wasn't always showing the error code
4977 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4978 timeout_connect_A, timeout_connect_MX.
4980 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4981 to the empty sender.
4983 39. The daemon was not analysing the content of -oX till after it had closed
4984 stderr and disconnected from the controlling terminal. This meant that any
4985 syntax errors were only noted on the panic log, and the return code from
4986 the command was 0. By re-arranging the code a little, I've made the
4987 decoding happen first, so such errors now appear on stderr, and the return
4988 code is 1. However, the actual setting up of the sockets still happens in
4989 the disconnected process, so errors there are still only recorded on the
4992 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4993 connections (as happens on some IP stacks) was logged at start up time as
4994 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4995 it from "IPv6 and IPv4", which means that two separate sockets are being
4998 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4999 says whether AF_INET or AF_INET6 was passed as an argument.
5001 42. Exiwhat output was messed up when time zones were included in log
5004 43. Exiwhat now gives more information about the daemon's listening ports,
5005 and whether -tls-on-connect was used.
5007 44. The "port" option of the smtp transport is now expanded.
5009 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5010 silently ignored. Now an error message is written to the main and panic
5013 46. There's a new ACL modifier called "logwrite" which writes to a log file
5014 as soon as it is encountered.
5016 47. Added $local_user_uid and $local_user_gid at routing time.
5018 48. Exim crashed when trying to verify a sender address that was being
5021 49. Exim was recognizing only a space character after ".include". It now also
5022 recognizes a tab character.
5024 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5025 extracting the relevant information from the specification. The man page no
5026 longer contains scrambled data for the -d option, and I've added a section
5027 at the front about calling Exim under different names.
5029 51. Added "extra_headers" argument to the "mail" command in filter files.
5031 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5034 53. Installed eximstats 1.29.
5036 54. Added transport_filter_timeout as a generic transport option.
5038 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5039 Cc: header lines. This was required by RFC 822, but it not required by RFC
5042 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5043 incoming messages that did not have them. Now it does so only if the
5044 message originates locally, that is, if there is no associated remote host
5045 address. When Resent- header lines are present, this applies to the Resent-
5046 lines rather than the non-Resent- lines.
5048 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5049 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5051 58. Messages for configuration errors now include the name of the main
5052 configuration file - useful now that there may be more than one file in a
5053 list (.included file names were always shown).
5055 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5056 for those rare installations that do not start the daemon as root or run it
5057 setuid root. I've cut out the call to initgroups() if the daemon is not
5060 60. The Exim user and group can now be bound into the binary as text strings
5061 that are looked up at the start of Exim's processing.
5063 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5065 62. Added $mailstore_basename variable.
5067 63. Installed patch to sieve.c from Michael Haardt.
5069 64. When Exim failed to open the panic log after failing to open the main log,
5070 the original message it was trying to log was written to stderr and debug
5071 output, but if they were not available (the usual case in production), it
5072 was lost. Now it is written to syslog before the two lines that record the
5073 failures to open the logs.
5075 65. Users' Exim filters run in subprocesses under the user's uid. It is
5076 possible for a "deliver" command or an alias in a "personal" command to
5077 provoke an address rewrite. If logging of address rewriting is configured,
5078 this fails because the process is not running as root or exim. There may be
5079 a better way of dealing with this, but for the moment (because 4.30 needs
5080 to be released), I have disabled address rewrite logging when running a
5081 filter in a non-root, non-exim process.
5087 1. The buildconfig auxiliary program wasn't quoting the value set for
5088 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5089 not defined. This bug was masked in 4.22 by the effect that was fixed in
5092 2. Some messages that were rejected after a message id was allocated were
5093 shown as "incomplete" by exigrep. It no longer does this for messages that
5094 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5096 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5097 have allow_domain_literals set, the ID did not get logged in the <= line.
5098 Domain literals are now always recognized in Message-ID: header lines.
5100 4. The first argument for a ${extract expansion item is the key name or field
5101 number. Leading and trailing spaces in this item were not being ignored,
5102 causing some misleading effects.
5104 5. When deliver_drop_privilege was set, single queue runner processes started
5105 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5106 same command in the process it spins off) were not dropping privilege.
5108 6. When the daemon running as "exim" started a queue runner, it always
5109 re-executed Exim in the spun-off process. This is a waste of effort when
5110 deliver_drop_privilege is set. The new process now just calls the
5111 queue-runner function directly.
5117 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5120 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5121 ignored. Though the use of -oP was forcing the writing of a pid file, it
5122 was always written to the default place.
5124 3. If the message "no IP address found for host xxxx" is generated during
5125 incoming verification, it is now followed by identification of the incoming
5126 connection (so you can more easily find what provoked it).
5128 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5130 5. Added some features to "harden" Exim a bit more against certain attacks:
5132 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5133 be put in Local/Makefile. This is like the never_users runtime option,
5134 but it cannot be overridden. The default setting is "root".
5136 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5137 prefix string with which any file named in a -C command line option
5140 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5141 is retained for -C and -D only if the caller of Exim is root. Without
5142 it, the exim user may also use -C and -D and retain privilege.
5144 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5145 command line option is disabled.
5147 6. Macro names set by the -D option must start with an upper case letter, just
5148 like macro names defined in the configuration file.
5150 7. Added "dereference=" facility to LDAP.
5152 8. Two instances of the typo "uknown" in the source files are fixed.
5154 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5155 the Configure-Makefile script screwed up while processing it.
5157 10. Incorporated PCRE 4.4.
5159 11. The SMTP synchronization check was not operating right at the start of an
5160 SMTP session. For example, it could not catch a HELO sent before the client
5161 waited for the greeting. There is now a check for outstanding input at the
5162 point when the greeting is written. Because of the duplex, asynchronous
5163 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5164 way, but not yet received, when the check is performed.
5166 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5167 on TCP/IP sockets, because this apparently causes some broken clients to
5170 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5171 unchanged) from the Cygwin maintainer.
5173 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5174 was supported instead of testing for mbx. Effectively a typo.
5176 15. The spa authenticator server code was not checking that the input it
5177 received was valid base64.
5179 16. The debug output line for the "set" modifier in ACLs was not showing the
5180 name of the variable that was being set.
5182 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5184 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5185 lookup on its own. Something else was needed to trigger the lookup. For
5186 example, a match in host_lookup or the need for a host name in a host list.
5187 Now, if $sender_host_name is referenced and the host name has not yet been
5188 looked up, a lookup is performed. If the lookup fails, the variable remains
5189 empty, and $host_lookup_failed is set to "1".
5191 19. Added "eqi" as a case-independent comparison operator.
5193 20. The saslauthd authentication condition could segfault if neither service
5194 nor realm was specified.
5196 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5197 error message that was logged was misleading, and incoming SMTP
5198 connections were dropped. The message is now more accurate, and temporary
5199 errors are given to SMTP connections.
5201 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5202 (except RSET) with a 421 error, until QUIT is received. However, it was
5203 failing to send a response to QUIT.
5205 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5206 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5207 ACL is now run after the helo_try_verify_hosts code.
5209 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5210 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5211 case-independent, so other case variants are also recognized). Apparently
5212 some systems use these upper case variants.
5214 25. If more than two messages were waiting for the same host, and a transport
5215 filter was specified for the transport, Exim sent two messages over the
5216 same TCP/IP connection, and then failed with "socket operation on non-
5217 socket" when it tried to send the third.
5219 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5221 27. The extern definition of crypt16() in expand.c was not being excluded when
5222 the OS had its own crypt16() function.
5224 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5225 as a preferred synonym for return_size_limit, both as an option and as an
5228 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5230 30. Changed the default configuration ACL to relax the local part checking rule
5231 for addresses that are not in any local domains. For these addresses,
5232 slashes and pipe symbols are allowed within local parts, but the sequence
5233 /../ is explicitly forbidden.
5235 31. SPA server authentication was not clearing the challenge buffer before
5238 32. log_message in a "warn" ACL statement was writing to the reject log as
5239 well as to the main log, which contradicts the documentation and doesn't
5240 seem right (because no rejection is happening). So I have stopped it.
5242 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5243 However, I am unable to do any testing of this.
5245 34. Fixed an infelicity in the appendfile transport. When checking directories
5246 for a mailbox, to see if any needed to be created, it was accidentally
5247 using path names with one or more superfluous leading slashes; tracing
5248 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5250 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5251 contents are added to the log line that is written for every discarded
5252 recipient. (Previously a log_message setting was ignored.)
5254 36. The ${quote: operator now quotes the string if it is empty.
5256 37. The install script runs exim in order to find its version number. If for
5257 some reason other than non-existence or emptiness, which it checks, it
5258 could not run './exim', it was installing it with an empty version number,
5259 i.e. as "exim-". This error state is now caught, and the installation is
5262 38. An argument was missing from the function that creates an error message
5263 when Exim fails to connect to the socket for saslauthd authentication.
5264 This could cause Exim to crash, or give a corrupted message.
5266 39. Added isip, isip4, and isip6 to ${if conditions.
5268 40. The ACL variables $acl_xx are now saved with the message, and can be
5269 accessed later in routers, transports, and filters.
5271 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5272 strings in the file are not string-expanded.
5274 42. If a MAIL command specified a SIZE value that was too large to fit into an
5275 int variable, the check against message_size_limit failed. Such values are
5276 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5277 day this will have to be increased, but I don't think I want to be around
5278 when emails are that large.
5285 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5286 iconv() is not standard in FreeBSD.
5288 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5289 IPv6 enabled. The observed symptom was a segmentation fault on return from
5290 the function os_common_find_running_interfaces() in src/os.c.
5292 3. In the check_special_case() function in daemon.c I had used "errno" as an
5293 argument name, which causes warnings on some systems. This was basically a
5294 typo, since it was named "eno" in the comments!
5296 4. The code that waits for the clock to tick (at a resolution of some fraction
5297 of a second) so as to ensure message-id uniqueness was always waiting for
5298 at least one whole tick, when it could have waited for less. [This is
5299 almost certainly not relevant at current processor speeds, where it is
5300 unlikely to ever wait at all. But we try to future-proof.]
5302 5. The function that sleeps for a time interval that includes fractions of a
5303 second contained a race. It did not block SIGALRM between setting the
5304 timer, and suspending (a couple of lines later). If the interval was short
5305 and the sigsuspend() was delayed until after it had expired, the suspension
5306 never ended. On busy systems this could lead to processes getting stuck for
5309 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5310 process, before it forks any delivery processes. The open lookup caching
5311 mechanism meant that the open file or database connection was passed into
5312 the delivery process. The problem was that delivery processes always tidy
5313 up cached lookup data. This could cause a problem for the next delivery
5314 process started by the queue runner, because the external queue runner
5315 process does not know about the closure. So the next delivery process
5316 still has data in the lookup cache. In the case of a file lookup, there was
5317 no problem because closing a file descriptor in a subprocess doesn't affect
5318 the parent. However, if the lookup was caching a connection to a database,
5319 the connection was closed, and the second delivery process was likely to
5320 see errors such as "PGSQL: query failed: server closed the connection
5321 unexpectedly". The problem has been fixed by closing all cached lookups
5322 in a queue runner before running a delivery process.
5324 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5325 seem to have the "const" qualifier which it has on other OS. I've
5328 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5329 *of the same type* (client or server) with the same public name that an
5330 error should be diagnosed.
5332 9. When Exim looked up a host name for an IP address, but failed to find the
5333 original IP address when looking up the host name (a safety check), it
5334 output the message "<ip address> does not match any IP for NULL", which was
5335 confusing, to say the least. The bug was that the host name should have
5336 appeared instead of "NULL".
5338 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5339 user that is built into the binary, and the -C or -D options is used, root
5340 privilege is dropped before the configuration file is read. In addition,
5341 logging is switched to stderr instead of the normal log files. If the
5342 configuration then re-defines the Exim user, the unprivileged environment
5343 is probably not what is expected, so Exim logs a panic warning message (but
5346 However, if deliver_drop_privilege is set, the unprivileged state may well
5347 be exactly what is intended, so the warning has been cut out in that case,
5348 and Exim is allowed to try to write to its normal log files.
5354 1. smtp_return_error_details was not giving details for temporary sender
5355 or receiver verification errors.
5357 2. Diagnose a configuration error if two authenticators have the same public
5360 3. Exim used not to create the message log file for a message until the first
5361 delivery attempt. This could be confusing when incoming messages were held
5362 for policy or load reasons. The message log file is now created at the time
5363 the message is received, and an initial "Received" line is written to it.
5365 4. The automatically generated man page for command line options had a minor
5366 bug that caused no ill effects; however, a more serious problem was that
5367 the procedure for building the man page automatically didn't always
5368 operate. Consequently, release 4.20 contains an out-of-date version. This
5369 shouldn't happen again.
5371 5. When building Exim with embedded Perl support, the script that builds the
5372 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5373 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5375 6. The freeze_tell option was not being used for messages that were frozen on
5376 arrival, either by an ACL or by local_scan().
5378 7. Added the smtp_incomplete_transaction log selector.
5380 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5381 was accepting AUTH without a new EHLO.
5383 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5384 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5387 10. Logging of TCP/IP connections (when configured) now happens in the main
5388 daemon process instead of the child process, so that the TCP/IP connection
5389 count is more accurate (but it can never be perfect).
5391 11. The use of "drop" in a nested ACL was not being handled correctly in the
5392 outer ACL. Now, if condition failure induced by the nested "drop" causes
5393 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5394 or "require"), the connection is dropped.
5396 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5397 that yield "discard" can now be used with an "accept" or a "discard" verb,
5398 but an error is generated for any others (because I can't see a useful way
5399 to define what should happen).
5401 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5402 are now processed in the same way as lines in the Exim configuration file.
5403 In particular, continuation lines are supported.
5405 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5407 15. Added -ti meaning -t -i.
5409 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5410 domains, and warn by logging if others are found.
5412 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5413 returned from GIFCONF when Exim is trying to find the list of interfaces on
5414 a host. The code in os.c has been modified to copy each ifreq to an aligned
5415 structure in all cases.
5417 Also, in some cases, the returned ifreq's were being copied to a 'struct
5418 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5419 means the last couple of bytes of an IPv6 address could be chopped if the
5420 ifreq contained only a normal sockaddr (14 bytes storage).
5422 18. Named domain lists were not supported in the hosts_treat_as_local option.
5423 An entry such as +xxxx was not recognized, and was treated as a literal
5426 19. Ensure that header lines added by a DATA ACL are included in the reject log
5427 if the ACL subsequently rejects the message.
5429 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5430 MD5 (which is deprecated).
5432 21. When testing a filter file using -bf, Exim was writing a message when it
5433 took the sender from a "From " line in the message, but it was not doing so
5434 when it took $return_path from a Return-Path: header line. It now does.
5436 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5437 with a valid header line field name (a series of printing characters
5438 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5440 23. Changed "disc" in the source to "disk" to conform to the documentation and
5441 the book and for uniformity.
5443 24. Ignore Sendmail's -Ooption=value command line item.
5445 25. When execve() failed while trying to run a command in a pipe transport,
5446 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5447 could be confused with a return value of 69 from the command itself. This
5448 has been changed to 127, the value the shell returns if it is asked to run
5449 a non-existent command. The wording for the related log line suggests a
5450 non-existent command as the problem.
5452 26. If received_header_text expands to an empty string, do not add a Received:
5453 header line to the message. (Well, it adds a token one on the spool, but
5454 marks it "old" so that it doesn't get used or transmitted.)
5456 27. Installed eximstats 1.28 (addition of -nt option).
5458 28. There was no check for failure on the call to getsockname() in the daemon
5459 code. This can fail if there is a shortage of resources on the system, with
5460 ENOMEM, for example. A temporary error is now given on failure.
5462 29. Contrary to the C standard, it seems that in some environments, the
5463 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5464 program. Exim now does this explicitly; it affects the formatting of
5465 timestamps using strftime().
5467 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5468 complaints. I've now initialized all the variables, to avoid this.
5470 32. Header lines added by a system filter were not being "seen" during
5471 transport-time rewrites.
5473 33. The info_callback() function passed to OpenSSL is set up with type void
5474 (*)(SSL *, int, int), as described somewhere. However, when calling the
5475 function (actually a macro) that sets it up, the type void(*)() is
5476 expected. I've put in a cast to prevent warnings from picky compilers.
5478 34. If a DNS black list lookup found a CNAME record, but there were no A
5479 records associated with the domain it pointed at, Exim crashed.
5481 35. If a DNS black list lookup returned more than one A record, Exim ignored
5482 all but the first. It now scans all returned addresses if a particular IP
5483 value is being sought. In this situation, the contents of the
5484 $dnslist_value variable are a list of all the addresses, separated by a
5487 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5488 to accept a host name and all its aliases if the forward lookup for any of
5489 them yielded the IP address of the incoming connection. Now it accepts only
5490 those names whose forward lookup yields the correct IP address. Any other
5491 names are discarded. This closes a loophole whereby a rogue DNS
5492 administrator could create reverse DNS records to break through a
5493 wildcarded host restriction in an ACL.
5495 37. If a user filter or a system filter that ran in a subprocess used any of
5496 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5497 the wrong values were passed to the pipe command ($thisaddress had the
5498 value of $0, $0 had the value of $1, etc). This bug was introduced by
5499 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5501 38. Improved the line breaking for long SMTP error messages from ACLs.
5502 Previously, if there was no break point between 40 and 75 characters, Exim
5503 left the rest of the message alone. Two changes have been made: (a) I've
5504 reduced the minimum length to 35 characters; (b) if it can't find a break
5505 point between 35 and 75 characters, it looks ahead and uses the first one
5506 that it finds. This may give the occasional overlong line, but at least the
5507 remaining text gets split now.
5509 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5510 file descriptors might be low, and that setting 1000 would always raise it.
5511 It turns out that in some environments, the limit is already over 1000 and
5512 that lowering it causes trouble. So now Exim takes care not to decrease it.
5514 40. When delivering a message, the value of $return_path is set to $sender_
5515 address at the start of routing (routers may change the value). By an
5516 oversight, this default was not being set up when an address was tested by
5517 -bt or -bv, which affected the outcome if any router or filter referred to
5520 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5521 log only when "message" or "log_message" are set. However, if one of the
5522 conditions was an address verification, or a call to a nested ACL, the
5523 messages generated by the underlying test were being passed through. This
5524 no longer happens. The underlying message is available in $acl_verify_
5525 message for both "message" and "log_message" expansions, so it can be
5526 passed through if needed.
5528 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5529 new expansion $bh_ to give the encoded byte string without charset
5530 translation. Translation happens only if iconv() is available; HAVE_ICONV
5531 indicates this at build time. HEADERS_CHARSET gives the charset to
5532 translate to; headers_charset can change it in the configuration, and
5533 "headers charset" can change it in an individual filter file.
5535 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5536 that creates RFC 2047 encoded "words" labels them as that charset instead
5537 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5538 expansion operator; (ii) when Exim creates a From: line for a local
5539 message; (iii) when a header line is rewritten to include a "phrase" part.
5541 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5542 buggy, causing it to skip the first lines of messages whose message ID
5543 ended in 'D'. This would not have bitten before Exim release 4.14, because
5544 message IDs were unlikely to end in 'D' before then. The effect was to have
5545 incorrect size information for certain domains.
5547 45. #include "config.h" was missing at the start of the crypt16.c module. This
5548 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5551 46. If there was a timeout during a "random" callout check, Exim treated it as
5552 a failure of the random address, and carried on sending RSET and the real
5553 address. If the delay was just some slowness somewhere, the response to the
5554 original RCPT would be taken as a response to RSET and so on, causing
5555 mayhem of various kinds.
5557 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5558 when I implemented it. It didn't allow for the fact that some option values
5559 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5560 the right test for positive values.
5562 48. Domain names in DNS records are case-independent. Exim always looks them up
5563 in lower case. Some resolvers return domain names in exactly the case they
5564 appear in the zone file, that is, they may contain uppercase letters. Not
5565 all resolvers do this - some return always lower case. Exim was treating a
5566 change of case by a resolver as a change of domain, similar to a widening
5567 of a domain abbreviation. This triggered its re-routing code and so it was
5568 trying to route what was effectively the same domain again. This normally
5569 caused routing to fail (because the router wouldn't handle the domain
5570 twice). Now Exim checks for this case specially, and just changes the
5571 casing of the domain that it ultimately uses when it transmits the message
5574 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5577 50. If a filter generated a file delivery with a non-absolute name (possible if
5578 no home directory exists for the router), the forbid_file option was not
5581 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5582 the existing equality matching.
5584 52. Exim was using ints instead of ino_t variables in some places where it was
5585 dealing with inode numbers.
5587 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5588 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5589 called TMPDIR, and if it finds it is different, it changes its value.
5591 54. The smtp_printf() function is now made available to local_scan() so
5592 additional output lines can be written before returning. There is also an
5593 smtp_fflush() function to enable the detection of a dropped connection.
5594 The variables smtp_input and smtp_batched_input are exported to
5597 55. Changed the default runtime configuration: the message "Unknown user"
5598 has been removed from the ACL, and instead placed on the localuser router,
5599 using the cannot_route_message feature. This means that any verification
5600 failures that generate their own messages won't get overridden. Similarly,
5601 the "Unrouteable address" message that was in the ACL for unverifiable
5602 relay addresses has also been removed.
5604 56. Added hosts_avoid_esmtp to the smtp transport.
5606 57. The exicyclog script was not checking for the esoteric option
5607 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5608 will work only if exicyclog is run under the appropriate euid.
5610 58. Following a discussion on the list, the rules by which Exim recognises line
5611 endings on incoming messages have been changed. The -dropcr and drop_cr
5612 options are now no-ops, retained only for backwards compatibility. The
5613 following line terminators are recognized: LF CRLF CR. However, special
5614 processing applies to CR:
5616 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5617 nor a local message in the state where . is a terminator.
5619 (ii) If a bare CR is encountered in a header line, an extra space is added
5620 after the line terminator so as not to end the header. The reasoning
5621 behind this is that bare CRs in header lines are most likely either
5622 to be mistakes, or people trying to play silly games.
5624 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5625 was being incorrectly given as 18 bytes larger than it should have been.
5626 This is a VOB (very old bug).
5628 60. This may never have affected anything current, but just in case it has:
5629 When the local host is found other than at the start of a list of hosts,
5630 the local host, those with the same MX, and any that follow, are discarded.
5631 When the list in question was part of a longer list of hosts, the following
5632 hosts (not currently being processed) were also being discarded. This no
5633 longer happens. I'm not sure if this situation could ever has previously
5636 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5639 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5640 header. This is something that is recommended in a new Internet Draft, and
5641 is something that is documented as being done by Sendmail. There are two
5642 possible values. For messages generated by the autoreply transport, Exim
5645 Auto-Submitted: auto-replied
5647 whereas for all other generated messages (e.g. bounces) it adds
5649 Auto-Submitted: auto-generated
5651 63. The "personal" condition in filters now includes a test for the
5652 Auto-Submitted: header. If it contains the string "auto-" the message it
5653 not considered personal.
5655 64. Added rcpt_include_affixes as a generic transport option.
5657 65. Added queue_only_override (default true).
5659 66. Added the syslog_duplication option.
5661 67. If what should have been the first header line of a message consisted of
5662 a space followed by a colon, Exim was mis-interpreting it as a header line.
5663 It isn't of course - it is syntactically invalid and should therefore be
5664 treated as the start of the message body. The misbehaviour could have
5665 caused a number of strange effects, including loss of data in subsequent
5666 header lines, and spool format errors.
5668 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5669 client host had authenticated. This control can now be exercised by an ACL
5670 for more flexibility.
5672 69. By default, callouts do not happen when testing with -bh. There is now a
5673 variant, -bhc, which does actually run the callout code, including
5674 consulting and updating the callout cache.
5676 70. Added support for saslauthd authentication, courtesy of Alexander
5679 71. If statvfs() failed on the spool or log directories while checking their
5680 size for availability, Exim confusingly gave the error "space shortage".
5681 Furthermore, in debugging mode it crashed with a floating point exception.
5682 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5683 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5684 serious problem, Exim now writes to the main and panic logs when this
5685 happens, with details of the failure. It then refuses to accept the
5686 incoming message, giving the message "spool directory problem" or "log
5687 directory problem" with a 421 code for SMTP messages.
5689 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5690 0, 1, and 2 exist, because some OS complain for execs without them (see
5691 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5692 descriptors. However, the code omitted to check that the open succeeded,
5693 causing mysterious errors if for some reason the permissions on /dev/null
5694 got screwed. Now Exim writes a message to the main and panic logs, and
5695 bombs out if it can't open /dev/null.
5697 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5698 interact so that it is all more flexible. It is supposed to remain
5699 backwards compatible. Also added extra_local_interfaces.
5701 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5702 to bomb out with an assertion failure - to the client this appears as a
5703 connection drop. This problem occurs in the part of the code that was taken
5704 from the Samba project. Fortunately, the assertion is in a very simple
5705 function, so I have fixed this by reproducing the function inline in the
5706 one place where it is called, and arranging for authentication to fail
5707 instead of killing the process with assert().
5709 75. The SPA client code was not working when the server requested OEM rather
5710 than Unicode encoding.
5712 76. Added code to make require_files with a specific uid setting more usable in
5713 the case where statting the file as root fails - usually a non-root-mounted
5714 NFS file system. When this happens and the failure is EACCES, Exim now
5715 forks a subprocess and does the per-uid checking as the relevant uid.
5717 77. Added process_log_path.
5719 78. If log_file_path was not explicitly set, a setting of check_log_space or
5720 check_log_inodes was ignored.
5722 79. If a space check for the spool or log partitions fails, the incident is now
5723 logged. Of course, in the latter case the data may get lost...
5725 80. Added the %p formatting code to string_format() so that it can be used to
5726 print addresses in debug_print(). Adjusted all the address printing in the
5727 debugging in store.c to use %p rather than %d.
5729 81. There was a concern that a line of code in smtp_in.c could overflow a
5730 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5731 initially expressed, the concern was not well-founded, because trailing
5732 spaces are removed early. However, if the trailing spaces were followed by
5733 a NULL, they did not get removed, so the overflow was possible. Two fixes
5736 (a) I re-wrote the offending code in a cleaner fashion.
5737 (b) If an incoming SMTP command contains a NULL character, it is rejected
5740 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5741 runs initgroups(), so that if the Exim user is in any additional groups,
5742 they will be used during message reception.
5748 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob