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.
74 PP/01 Handle short writes when writing local log-files.
75 In practice, only affects FreeBSD (8 onwards).
76 Bugzilla 1053, with thanks to Dmitry Isaikin.
78 NM/01 Bugzilla 949 - Documentation tweak
80 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
83 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
85 PP/02 Implemented gsasl authenticator.
87 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
89 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
90 `pkg-config foo` for cflags/libs.
92 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
93 with rest of GSASL and with heimdal_gssapi.
95 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
96 `pkg-config foo` for cflags/libs for the TLS implementation.
98 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
99 properties get this fed in as external SSF. A number of robustness
100 and debugging improvements to the cyrus_sasl authenticator.
102 PP/08 cyrus_sasl server now expands the server_realm option.
104 PP/09 Bugzilla 1214 - Log authentication information in reject log.
105 Patch by Jeremy Harris.
107 PP/10 Added dbmjz lookup type.
109 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
111 PP/12 MAIL args handles TAB as well as SP, for better interop with
112 non-compliant senders.
113 Analysis and variant patch by Todd Lyons.
115 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
116 Bug report from Lars Müller <lars@samba.org> (via SUSE),
117 Patch from Dirk Mueller <dmueller@suse.com>
119 PP/13 tls_peerdn now print-escaped for spool files.
120 Observed some $tls_peerdn in wild which contained \n, which resulted
121 in spool file corruption.
123 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
124 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
125 or write after TLS renegotiation, which otherwise led to messages
128 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
129 as a tracking header (ie: a signed header comes before the signature).
130 Patch from Wolfgang Breyha.
132 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
133 comma-sep list; embedded commas doubled.
135 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
137 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
139 Report and patch from Dmitry Banschikov.
141 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
142 Removed SSL_clear() after SSL_new() which led to protocol negotiation
143 failures. We appear to now support TLS1.1+ with Exim.
145 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
146 lets Exim select keys and certificates based upon TLS SNI from client.
147 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
148 before an outbound SMTP session. New log_selector, +tls_sni.
150 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
151 NULL dereference. Report and patch from Alun Jones.
153 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
154 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
155 Not seeing resolver debug output on NetBSD, but suspect this is a
156 resolver implementation change.
158 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
159 Left warnings. Added "eximon gdb" invocation mode.
161 PP/21 Defaulting "accept_8bitmime" to true, not false.
163 PP/22 Added -bw for inetd wait mode support.
165 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
166 locate the relevant includes and libraries. Made this the default.
168 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
169 Bugzilla 1246, report and most of solution from Tomasz Kusy.
171 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
172 This may cause build issues on older platforms.
174 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
175 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
176 gnutls_require_mac & gnutls_require_protocols (no longer supported).
177 Added SNI support via GnuTLS too.
178 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
180 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
182 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
183 Applied second patch from Janne, implementing suggestion to default
184 multiple-strings-in-record handling to match SPF spec.
186 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
188 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
189 read-only, out of scope).
190 Patch from Wolfgang Breyha, report from Stuart Northfield.
192 PP/29 Fix three issues highlighted by clang analyser static analysis.
193 Only crash-plausible issue would require the Cambridge-specific
194 iplookup router and a misconfiguration.
195 Report from Marcin Mirosław.
197 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
199 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
200 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
201 As part of this, removing so much warning spew let me fix some minor
202 real issues in debug logging.
204 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
205 assignment on my part. Fixed.
207 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
208 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
209 Janne Snabb (who went above and beyond: thank you).
211 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
212 string otherwise requires a connection and a bunch more work and it's
213 relatively easy to get wrong. Should also expose TLS library linkage
216 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
217 64-bit ${eval} (JH/03).
219 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
220 GNU libc to support some of the 64-bit stuff, should not lead to
221 conflicts. Defined before os.h is pulled in, so if a given platform
222 needs to override this, it can.
224 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
225 protection layer was required, which is not implemented.
226 Bugzilla 1254, patch from Wolfgang Breyha.
228 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
229 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
230 tls_dhparam take prime identifiers. Also unbreak combination of
231 OpenSSL+DH_params+TLSSNI.
233 PP/39 Disable SSLv2 by default in OpenSSL support.
239 PP/01 Solaris build fix for Oracle's LDAP libraries.
240 Bugzilla 1109, patch from Stephen Usher.
242 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
244 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
247 TF/02 Fix a couple more cases where we did not log the error message
248 when unlink() failed. See also change 4.74-TF/03.
250 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
251 lock up or crash if it happened to be inside a call to libc when it
252 got a SIGUSR1 from exiwhat.
254 The SIGUSR1 handler appends the current process status to the process
255 log which is later printed by exiwhat. It used to use the general
256 purpose logging code to do this, but several functions it calls are
257 not safe for signals.
259 The new output code in the SIGUSR1 handler is specific to the process
260 log, and simple enough that it's easy to inspect for signal safety.
261 Removing some special cases also simplifies the general logging code.
262 Removing the spurious timestamps from the process log simplifies
265 TF/04 Improved ratelimit ACL condition.
267 The /noupdate option has been deprecated in favour of /readonly which
268 has clearer semantics. The /leaky, /strict, and /readonly update modes
269 are mutually exclusive. The update mode is no longer included in the
270 database key; it just determines when the database is updated. (This
271 means that when you upgrde Exim will forget old rate measurements.)
273 Exim now checks that the per_* options are used with an update mode that
274 makes sense for the current ACL. For example, when Exim is processing a
275 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
276 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
277 must specify per_mail/readonly. If you omit the update mode it defaults to
278 /leaky where that makes sense (as before) or /readonly where required.
280 The /noupdate option is now undocumented but still supported for
281 backwards compatibility. It is equivalent to /readonly except that in
282 ACLs where /readonly is required you may specify /leaky/noupdate or
283 /strict/noupdate which are treated the same as /readonly.
285 A useful new feature is the /count= option. This is a generalization
286 of the per_byte option, so that you can measure the throughput of other
287 aggregate values. For example, the per_byte option is now equivalent
288 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
290 The per_rcpt option has been generalized using the /count= mechanism
291 (though it's more complicated than the per_byte equivalence). When it is
292 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
293 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
294 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
295 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
296 Note that using per_rcpt with a non-readonly update mode in more than
297 one ACL will cause the recipients to be double-counted. (The per_mail
298 and per_byte options don't have this problem.)
300 The handling of very low rates has changed slightly. If the computed rate
301 is less than the event's count (usually one) then this event is the first
302 after a long gap. In this case the rate is set to the same as this event's
303 count, so that the first message of a spam run is counted properly.
305 The major new feature is a mechanism for counting the rate of unique
306 events. The new per_addr option counts the number of different
307 recipients that someone has sent messages to in the last time period. It
308 behaves like per_rcpt if all the recipient addresses are different, but
309 duplicate recipient addresses do not increase the measured rate. Like
310 the /count= option this is a general mechanism, so the per_addr option
311 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
312 example, measure the rate that a client uses different sender addresses
313 with the options per_mail/unique=$sender_address. There are further
314 details in the main documentation.
316 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
318 TF/06 Removed a few PCRE remnants.
320 TF/07 Automatically extract Exim's version number from tags in the git
321 repository when doing development or release builds.
323 PP/02 Raise smtp_cmd_buffer_size to 16kB.
324 Bugzilla 879. Patch from Paul Fisher.
326 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
327 Heavily based on revision 40f9a89a from Simon Arlott's tree.
330 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
332 PP/05 Variable $av_failed, true if the AV scanner deferred.
333 Bugzilla 1078. Patch from John Horne.
335 PP/06 Stop make process more reliably on build failure.
336 Bugzilla 1087. Patch from Heiko Schlittermann.
338 PP/07 Make maildir_use_size_file an _expandable_ boolean.
339 Bugzilla 1089. Patch from Heiko Schlittermann.
341 PP/08 Handle ${run} returning more data than OS pipe buffer size.
342 Bugzilla 1131. Patch from Holger Weiß.
344 PP/09 Handle IPv6 addresses with SPF.
345 Bugzilla 860. Patch from Wolfgang Breyha.
347 PP/10 GnuTLS: support TLS 1.2 & 1.1.
349 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
352 PP/11 match_* no longer expand right-hand-side by default.
353 New compile-time build option, EXPAND_LISTMATCH_RHS.
354 New expansion conditions, "inlist", "inlisti".
356 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
358 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
360 PP/14 fix log_write() format string regression from TF/03.
361 Bugzilla 1152. Patch from Dmitry Isaikin.
367 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
369 PP/02 Harmonised TLS library version reporting; only show if debugging.
370 Layout now matches that introduced for other libraries in 4.74 PP/03.
372 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
374 PP/04 New "dns_use_edns0" global option.
376 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
379 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
380 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
382 TK/01 Updated PolarSSL code to 0.14.2.
383 Bugzilla 1097. Patch from Andreas Metzler.
385 PP/07 Catch divide-by-zero in ${eval:...}.
388 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
391 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
392 format-string attack -- SECURITY: remote arbitrary code execution.
394 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
395 time unintentionally subject to list matching rules, letting the header
396 cause arbitrary Exim lookups (of items which can occur in lists, *not*
397 arbitrary string expansion). This allowed for information disclosure.
399 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
400 INT_MIN/-1 -- value coerced to INT_MAX.
406 NM/01 Workround for PCRE version dependency in version reporting
409 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
410 This fixes portability to compilers other than gcc, notably
411 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
413 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
414 makefiles for portability to HP-UX and POSIX correctness.
416 PP/01 Permit LOOKUP_foo enabling on the make command-line.
417 Also via indented variable definition in the Makefile.
418 (Debugging by Oliver Heesakkers).
420 PP/02 Restore caching of spamd results with expanded spamd_address.
421 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
423 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
424 Improves build reliability. Fix from: Frank Elsner
426 NM/02 Fix wide character breakage in the rfc2047 coding
427 Fixes bug 1064. Patch from Andrey N. Oktyabrski
429 NM/03 Allow underscore in dnslist lookups
430 Fixes bug 1026. Patch from Graeme Fowler
432 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
433 Code patches from Adam Ciarcinski of NetBSD.
435 NM/04 Fixed exiqgrep to cope with mailq missing size issue
438 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
439 is logged, to avoid truncation. Patch from John Horne.
441 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
442 Patch from Jakob Hirsch.
444 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
445 SQL string expansion failure details.
446 Patch from Andrey Oktyabrski.
448 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
449 Patch from Simon Arlott.
451 PP/09 New lookups functionality failed to compile on old gcc which rejects
452 extern declarations in function scope.
453 Patch from Oliver Fleischmann
455 PP/10 Use sig_atomic_t for flags set from signal handlers.
456 Check getgroups() return and improve debugging.
457 Fixed developed for diagnosis in bug 927 (which turned out to be
460 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
461 Patch from Mark Zealey.
463 PP/12 Bugzilla 1056: Improved spamd server selection.
464 Patch from Mark Zealey.
466 PP/13 Bugzilla 1086: Deal with maildir quota file races.
467 Based on patch from Heiko Schlittermann.
469 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
470 Patch from Uwe Doering, sign-off by Michael Haardt.
472 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
473 variable declaration deep within a block. Bug and patch from
476 PP/15 lookups-Makefile IRIX compatibilty coercion.
478 PP/16 Make DISABLE_DKIM build knob functional.
480 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
481 Patch by Simon Arlott
483 TF/03 Fix valgrind.h portability to C89 compilers that do not support
484 variable argument macros. Our copy now differs from upstream.
490 TF/01 Failure to get a lock on a hints database can have serious
491 consequences so log it to the panic log.
493 TF/02 Log LMTP confirmation messages in the same way as SMTP,
494 controlled using the smtp_confirmation log selector.
496 TF/03 Include the error message when we fail to unlink a spool file.
498 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
499 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
500 for maintaining out-of-tree patches for some time.
502 PP/01 Bugzilla 139: Documentation and portability issues.
503 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
504 Handle per-OS dynamic-module compilation flags.
506 PP/02 Let /dev/null have normal permissions.
507 The 4.73 fixes were a little too stringent and complained about the
508 permissions on /dev/null. Exempt it from some checks.
509 Reported by Andreas M. Kirchwitz.
511 PP/03 Report version information for many libraries, including
512 Exim version information for dynamically loaded libraries. Created
513 version.h, now support a version extension string for distributors
514 who patch heavily. Dynamic module ABI change.
516 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
517 privilege escalation vulnerability whereby the Exim run-time user
518 can cause root to append content of the attacker's choosing to
521 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
524 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
525 If dropping privileges for untrusted macros, we disabled normal logging
526 on the basis that it would fail; for the Exim run-time user, this is not
527 the case, and it resulted in successful deliveries going unlogged.
528 Fixed. Reported by Andreas Metzler.
534 PP/01 Date: & Message-Id: revert to normally being appended to a message,
535 only prepend for the Resent-* case. Fixes regression introduced in
536 Exim 4.70 by NM/22 for Bugzilla 607.
538 PP/02 Include check_rfc2047_length in configure.default because we're seeing
539 increasing numbers of administrators be bitten by this.
541 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
543 PP/03 Bugzilla 994: added openssl_options main configuration option.
545 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
547 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
549 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
551 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
552 without a peer certificate, leading to a segfault because of an
553 assumption that peers always have certificates. Be a little more
554 paranoid. Problem reported by Martin Tscholak.
556 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
557 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
558 NB: ClamAV planning to remove STREAM in "middle of 2010".
559 CL also introduces -bmalware, various -d+acl logging additions and
560 more caution in buffer sizes.
562 PP/09 Implemented reverse_ip expansion operator.
564 PP/10 Bugzilla 937: provide a "debug" ACL control.
566 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
568 PP/12 Bugzilla 973: Implement --version.
570 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
572 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
574 PP/15 Bugzilla 816: support multiple condition rules on Routers.
576 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
577 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
578 ignore trailing whitespace.
580 JJ/02 prevent non-panic DKIM error from being sent to paniclog
582 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
585 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
586 Notification from Dr Andrew Aitchison.
588 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
589 ExtendedDetectionInfo response format.
590 Notification from John Horne.
592 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
595 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
596 XSL and documented dependency on system catalogs, with examples of how
599 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
602 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
603 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
604 configuration file which is writeable by the Exim user or group.
606 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
607 of configuration files to cover files specified with the -C option if
608 they are going to be used with root privileges, not just the default
611 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
612 option (effectively making it always true).
614 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
615 files to be used while preserving root privileges.
617 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
618 that rogue child processes cannot use them.
620 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
621 run-time user, instead of root.
623 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
624 Exim run-time user without dropping privileges.
626 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
627 result string, instead of calling string_vformat() twice with the same
630 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
631 for other users. Others should always drop root privileges if they use
632 -C on the command line, even for a whitelisted configure file.
634 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
636 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
642 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
643 $data_path, and $header_path variables; fixed documentation bugs and
646 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
647 exipick to access non-standard spools, including the "frozen" queue
650 NM/01 Bugzilla 965: Support mysql stored procedures.
651 Patch from Alain Williams
653 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
655 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
656 Patch from Andreas Metzler
658 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
659 Patch from Kirill Miazine
661 NM/05 Bugzilla 671: Added umask to procmail example.
663 JJ/03 installed exipick 20100323.0, fixing doc bug
665 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
666 directory. Notification and patch from Dan Rosenberg.
668 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
670 TK/02 Improve log output when DKIM signing operation fails.
672 MH/01 Treat the transport option dkim_domain as a colon separated
673 list, not as a single string, and sign the message with each element,
674 omitting multiple occurences of the same signer.
676 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
677 Bugzilla 985, 986. Patch by Simon Arlott
679 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
680 Patch by Simon Arlott
682 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
683 MBX locking. Notification from Dan Rosenberg.
689 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
691 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
693 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
695 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
697 NM/04 Bugzilla 915: Flag broken perl installation during build.
703 TK/01 Added patch by Johannes Berg that expands the main option
704 "spamd_address" if it starts with a dollar sign.
706 TK/02 Write list of recipients to X-Envelope-Sender header when building
707 the mbox-format spool file for content scanning (suggested by Jakob
710 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
711 (http://www.dcc-servers.net/) support via dccifd. Activated by
712 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
714 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
715 by Mark Daniel Reidel <mr@df.eu>.
717 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
718 When building exim an external PCRE library is now needed -
719 PCRE is a system library on the majority of modern systems.
720 See entry on PCRE_LIBS in EDITME file.
722 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
723 conversation. Added nologin parameter to request.
724 Patch contributed by Kirill Miazine.
726 TF/01 Do not log submission mode rewrites if they do not change the address.
728 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
730 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
731 log files in place. Contributed by Roberto Lima.
733 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
735 TF/03 Bugzilla 615: When checking the local_parts router precondition
736 after a local_part_suffix or local_part_prefix option, Exim now
737 does not use the address's named list lookup cache, since this
738 contains cached lookups for the whole local part.
740 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
741 Robert Millan. Documentation is in experimental-spec.txt.
743 TF/04 Bugzilla 668: Fix parallel build (make -j).
745 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
747 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
748 Patch provided by Jan Srzednicki.
750 TF/05 Leading white space used to be stripped from $spam_report which
751 wrecked the formatting. Now it is preserved.
753 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
754 that they are available at delivery time.
756 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
758 TF/08 TLS error reporting now respects the incoming_interface and
759 incoming_port log selectors.
761 TF/09 Produce a more useful error message if an SMTP transport's hosts
762 setting expands to an empty string.
764 NM/06 Bugzilla 744: EXPN did not work under TLS.
765 Patch provided by Phil Pennock.
767 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
768 Patch provided by Richard Godbee.
770 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
771 acl_smtp_notquit, added index entry.
773 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
774 Patch provided by Eugene Bujak.
776 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
777 accept(). Patch provided by Maxim Dounin.
779 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
780 Patch provided by Phil Pennock.
782 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
784 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
785 Patch provided by Brad "anomie" Jorsch.
787 NM/14 Bugzilla 622: Added timeout setting to transport filter.
788 Patch provided by Dean Brooks.
790 TK/05 Add native DKIM support (does not depend on external libraries).
792 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
793 Patch provided by Graeme Fowler.
795 NM/16 Bugzilla 851: Documentation example syntax fix.
797 NM/17 Changed NOTICE file to remove references to embedded PCRE.
799 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
802 NM/19 Bugzilla 745: TLS version reporting.
803 Patch provided by Phil Pennock.
805 NM/20 Bugzilla 167: bool: condition support.
806 Patch provided by Phil Pennock.
808 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
809 clients. Patch provided by Phil Pennock.
811 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
812 Patch provided by Brad "anomie" Jorsch.
814 NM/23 Bugzilla 687: Fix misparses in eximstats.
815 Patch provided by Heiko Schlittermann.
817 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
818 Patch provided by Heiko Schlittermann.
820 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
821 plus update to original patch.
823 NM/26 Bugzilla 799: Documentation correction for ratelimit.
825 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
826 Patch provided by David Brownlee.
828 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
830 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
832 NM/30 Bugzilla 888: TLS documentation bugfixes.
834 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
836 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
837 Unlike the original bugzilla I have changed all shell scripts in src tree.
839 NM/33 Bugzilla 898: Transport filter timeout fix.
840 Patch by Todd Rinaldo.
842 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
843 Patch by Serge Demonchaux.
845 NM/35 Bugzilla 39: Base64 decode bug fixes.
846 Patch by Jakob Hirsch.
848 NM/36 Bugzilla 909: Correct connect() call in dcc code.
850 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
852 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
854 NM/39 Bugzilla 911: Fixed MakeLinks build script.
860 TK/01 Add preliminary DKIM support. Currently requires a forked version of
861 ALT-N's libdkim that I have put here:
862 http://duncanthrax.net/exim-experimental/
864 Note to Michael Haardt: I had to rename some vars in sieve.c. They
865 were called 'true' and it seems that C99 defines that as a reserved
866 keyword to be used with 'bool' variable types. That means you could
867 not include C99-style headers which use bools without triggering
868 build errors in sieve.c.
870 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
871 as mailq or other aliases. Changed the --help handling significantly
872 to do whats expected. exim_usage() emits usage/help information.
874 SC/01 Added the -bylocaldomain option to eximstats.
876 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
878 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
880 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
886 PH/01 Another patch from the Sieve maintainer.
888 PH/02 When an IPv6 address is converted to a string for single-key lookup
889 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
890 dots are used instead of colons so that keys in lsearch files need not
891 contain colons. This was done some time before quoting was made available
892 in lsearch files. However, iplsearch files do require colons in IPv6 keys
893 (notated using the quote facility) so as to distinguish them from IPv4
894 keys. This meant that lookups for IP addresses in host lists did not work
895 for iplsearch lookups.
897 This has been fixed by arranging for IPv6 addresses to be expressed with
898 colons if the lookup type is iplsearch. This is not incompatible, because
899 previously such lookups could never work.
901 The situation is now rather anomolous, since one *can* have colons in
902 ordinary lsearch keys. However, making the change in all cases is
903 incompatible and would probably break a number of configurations.
905 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
908 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
909 conversion specification without a maximum field width, thereby enabling
910 a rogue spamd server to cause a buffer overflow. While nobody in their
911 right mind would setup Exim to query an untrusted spamd server, an
912 attacker that gains access to a server running spamd could potentially
913 exploit this vulnerability to run arbitrary code as the Exim user.
915 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
916 $primary_hostname instead of what libspf2 thinks the hosts name is.
918 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
919 a directory entry by the name of the lookup key. Previously, if a
920 symlink pointed to a non-existing file or a file in a directory that
921 Exim lacked permissions to read, a lookup for a key matching that
922 symlink would fail. Now it is enough that a matching directory entry
923 exists, symlink or not. (Bugzilla 503.)
925 PH/03 The body_linecount and body_zerocount variables are now exported in the
928 PH/04 Added the $dnslist_matched variable.
930 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
931 This means they are set thereafter only if the connection becomes
934 PH/06 Added the client_condition to authenticators so that some can be skipped
935 by clients under certain conditions.
937 PH/07 The error message for a badly-placed control=no_multiline_responses left
938 "_responses" off the end of the name.
940 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
942 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
943 (without spaces) instead of just copying the configuration text.
945 PH/10 Added the /noupdate option to the ratelimit ACL condition.
947 PH/11 Added $max_received_linelength.
949 PH/12 Added +ignore_defer and +include_defer to host lists.
951 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
952 way in which PCRE > 7.0 is built.
954 PH/14 Implemented queue_only_load_latch.
956 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
957 MAIL command. The effect was to mangle the value on 64-bit systems.
959 PH/16 Another patch from the Sieve maintainer.
961 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
963 PH/18 If a system quota error occurred while trying to create the file for
964 a maildir delivery, the message "Mailbox is full" was not appended to the
965 bounce if the delivery eventually timed out. Change 4.67/27 below applied
966 only to a quota excession during the actual writing of the file.
968 PH/19 It seems that peer DN values may contain newlines (and other non-printing
969 characters?) which causes problems in log lines. The DN values are now
970 passed through string_printing() before being added to log lines.
972 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
973 and InterBase are left for another time.)
975 PH/21 Added message_body_newlines option.
977 PH/22 Guard against possible overflow in moan_check_errorcopy().
979 PH/23 POSIX allows open() to be a macro; guard against that.
981 PH/24 If the recipient of an error message contained an @ in the local part
982 (suitably quoted, of course), incorrect values were put in $domain and
983 $local_part during the evaluation of errors_copy.
989 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
990 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
993 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
994 issue a MAIL command.
996 PH/02 In an ACL statement such as
998 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1000 if a client was not listed at all, or was listed with a value other than
1001 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1002 the condition was not true (as it should be), so access was not denied.
1003 The bug was that the ! inversion was incorrectly passed on to the second
1004 item. This has been fixed.
1006 PH/03 Added additional dnslists conditions == and =& which are different from
1007 = and & when the dns lookup returns more than one IP address.
1009 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1010 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1012 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1013 FSYNC, which compiles an option called disable_fsync that allows for
1014 bypassing fsync(). The documentation is heavily laced with warnings.
1016 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1018 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1019 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1020 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1021 including adding "make clean"; (3) Added -fPIC when compiling the test
1022 dynamically loaded module, to get rid of a warning.
1024 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1025 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1026 = 0s. The bug is otherwise harmless.
1028 PH/07 There was a bug in the dovecot authenticator such that the value of
1029 $auth1 could be overwritten, and so not correctly preserved, after a
1030 successful authentication. This usually meant that the value preserved by
1031 the server_setid option was incorrect.
1033 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1035 PH/09 Installed PCRE release 7.0.
1037 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1038 run for batched SMTP input. It is now run at the start of every message
1039 in the batch. While fixing this I discovered that the process information
1040 (output by running exiwhat) was not always getting set for -bs and -bS
1041 input. This is fixed, and it now also says "batched" for BSMTP.
1043 PH/11 Added control=no_pipelining.
1045 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1046 patch, slightly modified), and move the expansion of helo_data till after
1047 the connection is made in the smtp transport (so it can use these
1050 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1052 PH/14 Added log_selector = +pid.
1054 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1056 PH/16 Add ${if forany and ${if forall.
1058 PH/17 Added dsn_from option to vary the From: line in DSNs.
1060 PH/18 Flush SMTP output before performing a callout, unless control =
1061 no_callout_flush is set.
1063 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1064 was true (the default) a successful delivery failed to delete the retry
1065 item, thus causing premature timeout of the address. The bug is now
1068 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1070 PH/21 Long custom messages for fakedefer and fakereject are now split up
1071 into multiline reponses in the same way that messages for "deny" and
1072 other ACL rejections are.
1074 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1075 with slight modification.
1077 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1078 draft, changing the syntax and factoring some duplicate code".
1080 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1081 for deliveries of the second and subsequent messages over the same SMTP
1084 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1085 ${reduce, with only minor "tidies".
1087 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1089 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1090 expansion side effects.
1092 PH/27 When a message times out after an over-quota error from an Exim-imposed
1093 quota, the bounce message says "mailbox is full". This message was not
1094 being given when it was a system quota that was exceeded. It now should
1097 MH/03 Made $recipients available in local_scan(). local_scan() already has
1098 better access to the recipient list through recipients_list[], but
1099 $recipients can be useful in postmaster-provided expansion strings.
1101 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1102 in the case of a MAIL command with additional options following the
1103 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1104 were accidentally chopped off.
1106 PH/29 SMTP synchronization checks are implemented when a command is read -
1107 there is a check that no more input is waiting when there shouldn't be
1108 any. However, for some commands, a delay in an ACL can mean that it is
1109 some time before the response is written. In this time, more input might
1110 arrive, invalidly. So now there are extra checks after an ACL has run for
1111 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1112 pipelining has not been advertised.
1114 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1116 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1117 returned in the SMTP response when smtp_return_error_details was false.
1118 This has been fixed.
1120 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1121 instead of the C I/O that was originally supplied, because problems were
1122 reported on Solaris.
1124 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1125 Exim which did not show up earlier: it was assuming that a call to
1126 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1127 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1128 was a macro that became an assignment, so it seemed to work. This has
1129 changed to a proper function call with a void return, hence the compile
1130 error. Exim's code has been fixed.
1132 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1135 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1137 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1139 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1140 because it turns out to be too big or there is a timeout) there is no
1141 "Completed" line in the log. When some messages of this type were
1142 selected by exigrep, they were listed as "not completed". Others were
1143 picked up by some special patterns. I have improved the selection
1144 criteria to be more general.
1146 PH/38 The host_find_failed option in the manualroute router can now be set
1147 to "ignore", to completely ignore a host whose IP address cannot be
1148 found. If all hosts are ignored, the behaviour is controlled by the new
1149 host_all_ignored option.
1151 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1152 homing or because of multiple MX records with /mx) generated more than
1153 one IP address, and the following item turned out to be the local host,
1154 all the secondary addresses of the first item were incorrectly removed
1155 from the list, along with the local host and any following hosts (which
1156 is what is supposed to happen).
1158 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1159 whoever called Exim into the -H file. In the case of the daemon it was
1160 behaving confusingly. When first started, it used values for whoever
1161 started the daemon, but after a SIGHUP it used the Exim user (because it
1162 calls itself on a restart). I have changed the code so that it now always
1165 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1166 message are rejected with the same error (e.g. no authentication or bad
1167 sender address), and a DATA command is nevertheless sent (as can happen
1168 with PIPELINING or a stupid MUA), the error message that was given to the
1169 RCPT commands is included in the rejection of the DATA command. This is
1170 intended to be helpful for MUAs that show only the final error to their
1173 PH/42 Another patch from the Sieve maintainer.
1175 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1176 Eximstats - Fixed some broken HTML links and added missing column headers
1178 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1179 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1181 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1183 PH/43 Yet another patch from the Sieve maintainer.
1185 PH/44 I found a way to check for a TCP/IP connection going away before sending
1186 the response to the final '.' that terminates a message, but only in the
1187 case where the client has not sent further data following the '.'
1188 (unfortunately, this is allowed). However, in many cases there won't be
1189 any further data because there won't be any more messages to send. A call
1190 to select() can be used: if it shows that the input is "ready", there is
1191 either input waiting, or the socket has been closed. An attempt to read
1192 the next input character can distinguish the two cases. Previously, Exim
1193 would have sent an OK response which the client would never have see.
1194 This could lead to message repetition. This fix should cure that, at
1195 least in a lot of common cases.
1197 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1198 advertised in response to EHLO.
1204 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1205 fixed by 4.65/MH/01 (is this a record?) are fixed:
1207 (i) An empty string was always treated as zero by the numeric comparison
1208 operators. This behaviour has been restored.
1210 (ii) It is documented that the numeric comparison operators always treat
1211 their arguments as decimal numbers. This was broken in that numbers
1212 starting with 0 were being interpreted as octal.
1214 While fixing these problems I realized that there was another issue that
1215 hadn't been noticed. Values of message_size_limit (both the global option
1216 and the transport option) were treated as octal if they started with 0.
1217 The documentation was vague. These values are now always treated as
1218 decimal, and I will make that clear in the documentation.
1224 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1225 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1228 MH/01 Don't check that the operands of numeric comparison operators are
1229 integers when their expansion is in "skipping" mode (fixes bug
1230 introduced by 4.64-PH/07).
1232 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1233 child addresses, Exim now panics and dies. Previously, because the count
1234 is held in a short int, deliveries were likely to be lost. As such a
1235 large number of recipients for a single message is ridiculous
1236 (performance will be very, very poor), I have chosen to impose a limit
1237 rather than extend the field.
1243 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1244 leftover -K file (the existence of which was triggered by #402).
1245 While we were at it, introduced process PID as part of the -K
1246 filename. This should rule out race conditions when creating
1249 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1250 processing considerably. Previous code took too long for large mails,
1251 triggering a timeout which in turn triggers #401.
1253 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1254 in the DK code in transports.c. sendfile() is not really portable,
1255 hence the _LINUX specificness.
1257 TF/01 In the add_headers option to the mail command in an Exim filter,
1258 there was a bug that Exim would claim a syntax error in any
1259 header after the first one which had an odd number of characters
1262 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1263 callout verification, Exim cached a "reject" for the entire domain. This
1264 is correct for most verifications, but it is not correct for a recipient
1265 verification with use_sender or use_postmaster set, because in that case
1266 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1267 case of MAIL FROM:<> rejection from other early rejections (e.g.
1268 rejection of HELO). When verifying a recipient using a non-null MAIL
1269 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1270 Whatever the result of the callout, the value of the domain cache is
1271 left unchanged (for any other kind of callout, getting as far as trying
1272 RCPT means that the domain itself is ok).
1274 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1277 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1278 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1279 session, the daemon ignores ECONNECT errors and logs others; it now
1280 ignores EPIPE as well.
1282 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1283 (quoted-printable decoding).
1285 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1286 later the small subsequent patch to fix an introduced bug.
1288 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1290 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1292 PH/08 An error is now given if message_size_limit is specified negative.
1294 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1295 to be given (somewhat) arbitrary names.
1297 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1300 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1301 miscellaneous code fixes
1303 PH/10 Added the log_reject_target ACL modifier to specify where to log
1306 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1307 hostname. This is wrong, because it relates to the incoming message (and
1308 probably the interface on which it is arriving) and not to the outgoing
1309 callout (which could be using a different interface). This has been
1310 changed to use the value of the helo_data option from the smtp transport
1311 instead - this is what is used when a message is actually being sent. If
1312 there is no remote transport (possible with a router that sets up host
1313 addresses), $smtp_active_hostname is used.
1315 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1316 tweaks were necessary in order to get it to work (see also 21 below):
1317 (a) The code assumed that strncpy() returns a negative number on buffer
1318 overflow, which isn't the case. Replaced with Exim's string_format()
1320 (b) There were several signed/unsigned issues. I just did the minimum
1321 hacking in of casts. There is scope for a larger refactoring.
1322 (c) The code used strcasecmp() which is not a standard C function.
1323 Replaced with Exim's strcmpic() function.
1324 (d) The code set only $1; it now sets $auth1 as well.
1325 (e) A simple test gave the error "authentication client didn't specify
1326 service in request". It would seem that Dovecot has changed its
1327 interface. Fortunately there's a specification; I followed it and
1328 changed what the client sends and it appears to be working now.
1330 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1333 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1334 address A is aliased to B and C, where B exists and C does not. Without
1335 -v the output is "A verified" because verification stops after a
1336 successful redirection if more than one address is generated. However,
1337 with -v the child addresses are also verified. Exim was outputting "A
1338 failed to verify" and then showing the successful verification for C,
1339 with its parentage. It now outputs "B failed to verify", showing B's
1340 parentage before showing the successful verification of C.
1342 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1343 look up a TXT record in a specific list after matching in a combined
1346 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1347 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1348 they consult the DNS. I had assumed they would set it the way they
1349 wanted; and indeed my experiments on Linux seem to show that in some
1350 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1351 To be on the safe side, however, I have now made the interface to
1352 host_find_byname() similar to host_find_bydns(), with an argument
1353 containing the DNS resolver options. The host_find_byname() function now
1354 sets these options at its start, just as host_find_bydns() does. The smtp
1355 transport options dns_qualify_single and dns_search_parents are passed to
1356 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1357 of host_find_byname() use the default settings of RES_DEFNAMES
1358 (qualify_single) but not RES_DNSRCH (search_parents).
1360 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1361 spool_read_header() do less string testing, by means of a preliminary
1362 switch on the second character of optional "-foo" lines. (This is
1363 overdue, caused by the large number of possibilities that now exist.
1364 Originally there were few.) While I was there, I also converted the
1365 str(n)cmp tests so they don't re-test the leading "-" and the first
1366 character, in the hope this might squeeze out yet more improvement.
1368 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1369 flag allowing group syntax was set by the header_syntax check but not
1370 turned off, possible causing trouble later; (2) The flag was not being
1371 set at all for the header_verify test, causing "group"-style headers to
1372 be rejected. I have now set it in this case, and also caused header_
1373 verify to ignore an empty address taken from a group. While doing this, I
1374 came across some other cases where the code for allowing group syntax
1375 while scanning a header line wasn't quite right (mostly, not resetting
1376 the flag correctly in the right place). These bugs could have caused
1377 trouble for malformed header lines. I hope it is now all correct.
1379 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1380 with the "reply" argument non-NULL. The code, however (which originally
1381 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1382 but it didn't always do it. This confused somebody who was copying the
1383 code for some other use. I have removed all the tests.
1385 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1386 feature that was used to support insecure browsers during the U.S. crypto
1387 embargo. It requires special client support, and Exim is probably the
1388 only MTA that supported it -- and would never use it because real RSA is
1389 always available. This code has been removed, because it had the bad
1390 effect of slowing Exim down by computing (never used) parameters for the
1391 RSA_EXPORT functionality.
1393 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1394 authenticator to fail if there's a tab character in the incoming data
1395 (there should never be unless someone is messing about, as it's supposed
1396 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1397 if the connection is using TLS or if the remote IP is the same as the
1398 local IP, and the "valid-client-cert option" if a client certificate has
1401 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1402 authenticators. This can be used for authorization after authentication
1403 succeeds. (In the case of plaintext, it servers for both authentication
1406 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1407 if any retry times were supplied.
1409 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1410 connection, where there is no client IP address to check. In this
1411 situation, the verify now always succeeds.
1413 PH/25 Applied John Jetmore's -Mset patch.
1415 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1417 PH/27 In a string expansion for a processed (not raw) header when multiple
1418 headers of the same name were present, leading whitespace was being
1419 removed from all of them, but trailing whitespace was being removed only
1420 from the last one. Now trailing whitespace is removed from each header
1421 before concatenation. Completely empty headers in a concatenation (as
1422 before) are ignored.
1424 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1425 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1427 PH/29 [Removed. This was a change that I later backed out, and forgot to
1428 correct the ChangeLog entry (that I had efficiently created) before
1429 committing the later change.]
1431 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1432 address errors (4xx response to RCPT) over the same connection as other
1433 messages routed to the same hosts. Such deliveries are always "forced",
1434 so retry times are not inspected. This resulted in far too many retries
1435 for the affected addresses. The effect occurred only when there were more
1436 hosts than the hosts_max_try setting in the smtp transport when it had
1437 the 4xx errors. Those hosts that it had tried were not added to the list
1438 of hosts for which the message was waiting, so if all were tried, there
1439 was no problem. Two fixes have been applied:
1441 (i) If there are any address or message errors in an SMTP delivery, none
1442 of the hosts (tried or untried) are now added to the list of hosts
1443 for which the message is waiting, so the message should not be a
1444 candidate for sending over the same connection that was used for a
1445 successful delivery of some other message. This seems entirely
1446 reasonable: after all the message is NOT "waiting for some host".
1447 This is so "obvious" that I'm not sure why it wasn't done
1448 previously. Hope I haven't missed anything, but it can't do any
1449 harm, as the worst effect is to miss an optimization.
1451 (ii) If, despite (i), such a delivery is accidentally attempted, the
1452 routing retry time is respected, so at least it doesn't keep
1453 hammering the server.
1455 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1456 in ${readsocket because some servers need this prod.
1458 PH/32 Added some extra debug output when updating a wait-xxx database.
1460 PH/33 The hint "could be header name not terminated by colon", which has been
1461 given for certain expansion errors for a long time, was not being given
1462 for the ${if def:h_colon_omitted{... case.
1464 PH/34 The spec says: "With one important exception, whenever a domain list is
1465 being scanned, $domain contains the subject domain." There was at least
1466 one case where this was not true.
1468 PH/35 The error "getsockname() failed: connection reset by peer" was being
1469 written to the panic log as well as the main log, but it isn't really
1470 panic-worthy as it just means the connection died rather early on. I have
1471 removed the panic log writing for the ECONNRESET error when getsockname()
1474 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1475 runs only) independently of the message's sender address. This meant
1476 that, if the 4xx error was in fact related to the sender, a different
1477 message to the same recipient with a different sender could confuse
1478 things. In particualar, this can happen when sending to a greylisting
1479 server, but other circumstances could also provoke similar problems.
1480 I have changed the default so that the retry time for these errors is now
1481 based a combination of the sender and recipient addresses. This change
1482 can be overridden by setting address_retry_include_sender=false in the
1485 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1486 remote server are returned as part of bounce messages. This was not
1487 happening for LMTP over a pipe (the lmtp transport), but now it is the
1488 same for both kinds of LMTP.
1490 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1491 in header lines that were in fact CNAMEs. This is no longer the case.
1493 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1494 and queue runs started by the daemon processed all messages. This has
1495 been fixed so that -R and -S can now usefully be given with -q<time>.
1497 PH/40 Import PCRE release 6.7 (fixes some bugs).
1499 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1501 PH/42 Give an error if -q is specified more than once.
1503 PH/43 Renamed the variables $interface_address and $interface_port as
1504 $received_ip_address and $received_port, to make it clear that these
1505 values apply to message reception, and not to the outgoing interface when
1506 a message is delivered. (The old names remain recognized, of course.)
1508 PH/44 There was no timeout on the connect() call when using a Unix domain
1509 socket in the ${readsocket expansion. There now is.
1511 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1512 be meaningful with "accept".
1514 SC/01 Eximstats V1.43
1515 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1517 SC/02 Eximstats V1.44
1518 Use a glob alias rather than an array ref in the generated
1519 parser. This improves both readability and performance.
1521 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1522 Collect SpamAssassin and rejection statistics.
1523 Don't display local sender or destination tables unless
1524 there is data to show.
1525 Added average volumes into the top table text output.
1527 SC/04 Eximstats V1.46
1528 Collect data on the number of addresses (recipients)
1529 as well as the number of messages.
1531 SC/05 Eximstats V1.47
1532 Added 'Message too big' to the list of mail rejection
1533 reasons (thanks to Marco Gaiarin).
1535 SC/06 Eximstats V1.48
1536 Mainlog lines which have GMT offsets and are too short to
1537 have a flag are now skipped.
1539 SC/07 Eximstats V1.49 (Alain Williams)
1540 Added the -emptyok flag.
1542 SC/08 Eximstats V1.50
1543 Fixes for obtaining the IP address from reject messages.
1545 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1546 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1547 whitesspace changes from 4.64-PH/27
1549 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1552 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1553 are found, allow negative numbers in numeric criteria)
1555 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1557 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1558 to match changes made in 4.64-PH/43
1560 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1562 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1563 "domain" is actually an IP address, and force a failure. This locks out
1564 those revolvers/nameservers that support "A-for-A" lookups, in
1565 contravention of the specifications.
1567 PH/48 When a host name was looked up from an IP address, and the subsequent
1568 forward lookup of the name timed out, the host name was left in
1569 $sender_host_name, contrary to the specification.
1571 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1572 restricted to single-key lookups, Exim was not diagnosing an error if
1573 * or *@ was used with a query-style lookup.
1575 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1577 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1578 long ago, but noone interested enough thought of it. Let's just say that
1579 the "1.1" means that there are some new functions that weren't there at
1580 some point in the past.
1582 PH/51 Error processing for expansion failure of helo_data from an smtp
1583 transport during callout processing was broken.
1585 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1586 tested/used via the -bh/-bhc/-bs options.
1588 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1589 bug, fixed in subsequent PCRE releases).
1591 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1592 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1594 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1600 SC/01 Use a glob alias rather than an array ref in eximstats generated
1601 parser. This improves both readability and performance.
1603 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1604 Don't display local sender or destination tables in eximstats unless
1605 there is data to show.
1606 Added average volumes into the eximstats top table text output.
1608 SC/03 Collect data on the number of addresses (recipients) as well
1609 as the number of messages in eximstats.
1611 TF/01 Correct an error in the documentation for the redirect router. Exim
1612 does (usually) call initgroups() when daemonizing.
1614 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1615 with consistent privilege compared to when running as a daemon.
1617 TF/03 Note in the spec that $authenticated_id is not set for local
1618 submissions from trusted users.
1620 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1621 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1623 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1624 by adding some example configuration directives to the default
1625 configuration file. A little bit of work is required to uncomment the
1626 directives and define how usernames and passwords are checked, but
1627 there is now a framework to start from.
1629 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1630 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1631 without this. I don't know how relevant this is to other LDAP libraries.
1633 PH/02 Add the verb name to the "unknown ACL verb" error.
1635 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1637 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1639 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1640 directory not expanded when it should be if an expanded home directory
1641 was set for the address (which is overridden by the transport).
1643 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1646 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1647 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1648 because it is too late at that time, and has no effect.
1650 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1651 security issue with \' (bugzilla #107). I could not use the
1652 PQescapeStringConn() function, because it needs a PGconn value as one of
1655 PH/08 When testing addresses using -bt, indicate those final addresses that
1656 are duplicates that would not cause an additional delivery. At least one
1657 person was confused, thinking that -bt output corresponded to deliveries.
1658 (Suppressing duplicates isn't a good idea as you lose the information
1659 about possibly different redirections that led to the duplicates.)
1661 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1662 systems where poll() doesn't work, in particular OS X.
1664 PH/10 Added more information to debugging output for retry time not reached.
1666 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1667 operations in malware.c.
1669 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1672 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1673 syslog_duplication set false, log lines that would normally be written
1674 both the the main log and to the reject log were not written to syslog at
1677 PH/14 In the default configuration, change the use of "message" in ACL warn
1678 statements to "add_header".
1680 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1681 not followed by a command (e.g. "seen endif").
1683 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1684 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1687 PH/17 Added extra conditions to the default value of delay_warning_condition
1691 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1692 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1693 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1696 The Auto-Submitted: and various List- headers are standardised, whereas I
1697 don't think Precedence: ever was.
1699 PH/18 Refactored debugging code in route_finduser() to show more information,
1700 in particular, the error code if getpwnam() issues one.
1702 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1703 This is apparently needed in addition to the PH/07 change above to avoid
1704 any possible encoding problems.
1706 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1707 but not after initializing Perl.
1709 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1710 output them only if debugging. By default they are written stderr,
1711 apparently, which is not desirable.
1713 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1716 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1719 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1721 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1722 authenticated or an ident call has been made. Suppress the default
1723 values for $authenticated_id and $authenticated_sender (but permit -oMai
1724 and -oMas) when testing with -bh.
1726 PH/24 Re-jigged the order of the tests in the default configuration so that the
1727 tests for valid domains and recipients precede the DNS black list and CSA
1728 tests, on the grounds that those ones are more expensive.
1730 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1731 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1732 This bug exists in every version of Exim that I still have, right back to
1735 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1736 However, an attempt to turn on case-sensitivity in a regex key by
1737 including (?-i) didn't work because the subject string was already
1738 lowercased, and the effects were non-intuitive. It turns out that a
1739 one-line patch can be used to allow (?-i) to work as expected.
1745 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1746 other effects) broke the use of negated acl sub-conditions.
1748 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1751 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1752 "Deny" causes Exim to reject the incoming connection with a 554 error.
1753 Unfortunately, if there is a major crisis, such as a disk failure,
1754 tcp-wrappers gives "deny", whereas what one would like would be some
1755 kind of temporary error. A kludge has been added to help with this.
1756 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1757 554 error is used if errno is still zero or contains ENOENT (which occurs
1758 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1761 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1763 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1764 errors. Otherwise a message that provokes a temporary error (when other
1765 messages do not) can cause a whole host to time out.
1767 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1768 addresses were routed directly to files or pipes from a redirect router.
1769 File deliveries just didn't batch; pipe deliveries might have suffered
1772 PH/06 A failure to get a lock for a hints database would erroneously always say
1773 "Failed to get write lock", even when it was really a read lock.
1775 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1776 of 0600. This has been changed to use the value of the lockfile_mode
1777 option (which defaults to 0600).
1779 PH/08 Applied small patch from the Sieve maintainer.
1781 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1782 folder from quota calculations, a direct delivery into this folder messed
1783 up the contents of the maildirsize file. This was because the regex was
1784 used only to exclude .Trash (or whatever) when the size of the mailbox
1785 was calculated. There was no check that a delivery was happening into an
1786 excluded directory. This bug has been fixed by ignoring all quota
1787 processing for deliveries into excluded directories.
1789 PH/10 Added the maildirfolder_create_regex option to appendfile.
1795 PH/01 The code for finding all the local interface addresses on a FreeBSD
1796 system running IPv6 was broken. This may well have applied to all BSD
1797 systems, as well as to others that have similar system calls. The broken
1798 code found IPv4 interfaces correctly, but gave incorrect values for the
1799 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1800 that it would not match correctly against @[] and not recognize the IPv6
1803 PH/02 The ipliteral router was not recognizing addresses of the form user@
1804 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1806 PH/03 Added disable_ipv6.
1808 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1809 decoded form, because it is most often used to construct To: headers
1810 lines in autoreplies, and the decoded form may well be syntactically
1811 invalid. However, $reply_address has leading white space removed, and all
1812 newlines turned into spaces so that the autoreply transport does not
1815 PH/05 If group was specified without a user on a router, and no group or user
1816 was specified on a transport, the group from the router was ignored.
1818 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1819 for visible compile-time settings that can be used to change these
1820 numbers, for those that want even more. Backwards compatibility with old
1821 spool files has been maintained. However, going back to a previous Exim
1822 release will lost any variables that are in spool files.
1824 PH/07 Two small changes when running in the test harness: increase delay when
1825 passing a TCP/IP connection to a new process, in case the original
1826 process has to generate a bounce, and remove special handling of
1827 127.0.0.2 (sic), which is no longer necessary.
1829 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1830 be the same on different OS.
1832 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1835 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1836 whether --show-vars was specified or not
1838 JJ/02 exipick: Added support for new ACL variable spool format introduced
1841 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1842 syntactically invalid From: or Reply-to: line, and a filter used this to
1843 generate an autoreply, and therefore failed to obtain an address for the
1844 autoreply, Exim could try to deliver to a non-existent relative file
1845 name, causing unrelated and misleading errors. What now happens is that
1846 it logs this as a hard delivery error, but does not attempt to create a
1849 PH/11 The exinext utility has a -C option for testing purposes, but although
1850 the given file was scanned by exinext itself; it wasn't being passed on
1851 when Exim was called.
1853 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1854 an end-of-file indication when reading a command response.
1856 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1857 compiled. In many other places in Exim, IPv6 addresses are always
1858 recognized, so I have changed this. It also means that IPv4 domain
1859 literals of the form [IPV4:n.n.n.n] are now always recognized.
1861 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1862 used if the router is not running as root, for example, when verifying at
1863 ACL time, or when using -bh. The debugging output from this situation was
1864 non-existent - all you got was a failure to exec. I have made two
1867 (a) Failures to set uid/gid, the current directory, or a process leader
1868 in a subprocess such as that created by queryprogram now generate
1869 suitable debugging ouput when -d is set.
1871 (b) The queryprogram router detects when it is not running as root,
1872 outputs suitable debugging information if -d is set, and then runs
1873 the subprocess without attempting to change uid/gid.
1875 PH/15 Minor change to Makefile for building test_host (undocumented testing
1878 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1879 additional section of a DNS packet that returns MX or SRV records.
1880 Instead, it always explicitly searches for A/AAAA records. This avoids
1881 major problems that occur when a DNS server includes only records of one
1882 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1883 fixed another bug: if SRV records were looked up and the corresponding
1884 address records were *not* found in the additional section, the port
1885 values from the SRV records were lost.
1887 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1888 using the correct key (the original address) when searching the retry
1889 rules in order to find which one to use for generating the retry hint.
1891 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1892 adding the default one. Similarly, if it contains a Reply-To: header, the
1893 errors_reply_to option, if set, is not used.
1895 PH/19 When calculating a retry time, Exim used to measure the "time since
1896 failure" by looking at the "first failed" field in the retry record. Now
1897 it does not use this if it is later than than the arrival time of the
1898 message. Instead it uses the arrival time. This makes for better
1899 behaviour in cases where some deliveries succeed, thus re-setting the
1900 "first failed" field. An example is a quota failure for a huge message
1901 when small messages continue to be delivered. Without this change, the
1902 "time since failure" will always be short, possible causing more frequent
1903 delivery attempts for the huge message than are intended.
1904 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1906 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1907 $1, $2, $3) because the numerical variables can be reset during some
1908 expansion items (e.g. "match"), thereby losing the authentication data.
1910 PH/21 Make -bV show the size of off_t variables so that the test suite can
1911 decide whether to run tests for quotas > 2G.
1913 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1914 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1915 filecount value is greater than 2G or if a quota value is greater than 2G
1916 on a system where the size of off_t is not greater than 4, a panic error
1919 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1920 never match. The debug and -bh output now contains an explicit error
1921 message indicating a malformed IPv4 address or mask.
1923 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1924 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1925 PH/23 above applies.
1927 PH/25 Do not write to syslog when running in the test harness. The only
1928 occasion when this arises is a failure to open the main or panic logs
1929 (for which there is an explicit test).
1931 PH/26 Added the /no_tell option to "control=freeze".
1933 PH/27 If a host name lookup failed very early in a connection, for example, if
1934 the IP address matched host_lookup and the reverse lookup yielded a name
1935 that did not have a forward lookup, an error message of the form "no IP
1936 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1937 could be logged. Now it outputs the IP address instead of "NULL".
1939 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1940 allows the sender and the authenticated sender to be set when
1941 submitting a message from within Exim. Since child_open_exim() is
1942 documented for local_scan(), the new function should be too.
1944 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1945 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1946 results in an empty string is now treated as unset.
1948 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1950 PH/31 Added sender_verify_fail logging option.
1952 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1953 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1954 tidied the source and removed it altogether.
1956 PH/33 When a queue run was abandoned because the load average was too high, a
1957 log line was always written; now it is written only if the queue_run log
1958 selector is set. In addition, the log line for abandonment now contains
1959 information about the queue run such as the pid. This is always present
1960 in "start" and "stop" lines but was omitted from the "abandon" line.
1962 PH/34 Omit spaces between a header name and the colon in the error message that
1963 is given when verify = headers_syntax fails (if there are lots of them,
1964 the message gets confusing).
1966 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1967 names, as there are now some PTR records that contain slashes. This check
1968 is only to protect against broken name servers that fall over on strange
1969 characters, so the fact that it applies to all lookups doesn't matter.
1971 PH/36 Now that the new test suite is complete, we can remove some of the
1972 special code in Exim that was needed for the old test suite. For example,
1973 sorting DNS records because real resolvers return them in an arbitrary
1974 order. The new test suite's fake resolver always returns records in the
1977 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1978 bounces) except when queue_only is set, to avoid logging races between
1979 the different processes.
1981 PH/38 Panic-die if .include specifies a non-absolute path.
1983 PH/39 A tweak to the "H" retry rule from its user.
1985 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1986 a label. They prevented compilation on older perls.
1988 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1989 a warning to be raised on newish perls.
1991 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1992 on queue. Changes to match documented behaviour of showing count of
1993 messages matching specified criteria.
1995 PH/40 Changed the default ident timeout from 30s to 5s.
1997 PH/41 Added support for the use of login_cap features, on those BSD systems
1998 that have them, for controlling the resources used by pipe deliveries.
2000 PH/42 The content-scanning code uses fopen() to create files in which to put
2001 message data. Previously it was not paying any attention to the mode of
2002 the files. Exim runs with umask(0) because the rest of the code creates
2003 files with open(), and sets the required mode explicitly. Thus, these
2004 files were ending up world-writeable. This was not a big issue, because,
2005 being within the spool directory, they were not world-accessible. I have
2006 created a function called modefopen, which takes an additional mode
2007 argument. It sets umask(777), creates the file, chmods it to the required
2008 mode, then resets the umask. All the relevant calls to fopen() in the
2009 content scanning code have been changed to use this function.
2011 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2012 to 24 hours. This avoids potential overflow problems when processing G
2013 and H retry rules. I suspect nobody ever tinkers with this value.
2015 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2017 PH/45 When the plaintext authenticator is running as a client, the server's
2018 challenges are checked to ensure they are valid base64 strings. By
2019 default, the authentication attempt is cancelled if an invalid string is
2020 received. Setting client_ignore_invalid_base64 true ignores these errors.
2021 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2022 they are received. Thus, the responses can be made to depend on the
2023 challenges. If an invalid string is ignored, an empty string is placed in
2026 PH/46 Messages that are created by the autoreply transport now contains a
2027 References: header, in accordance with RFCs 2822 and 3834.
2029 PH/47 Added authenticated_sender_force to the smtp transport.
2031 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2033 PH/49 Installed latest patch from the Sieve maintainer.
2035 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2036 was also set, the appendfile transport was unnecessarily scanning a
2037 directory of message files (e.g. for maildir delivery) to find the count
2038 of files (along with the size), even though it did not need this
2039 information. It now does the scan only if it needs to find either the
2040 size of the count of files.
2042 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2044 PH/52 Two bugs concerned with error handling when the smtp transport is
2047 (i) Exim was not creating retry information for temporary errors given
2048 for individual recipients after the DATA command when the smtp transport
2049 was used in LMTP mode. This meant that they could be retried too
2050 frequently, and not timed out correctly.
2052 (ii) Exim was setting the flag that allows error details to be returned
2053 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2054 recipients that were returned after the DATA command.
2056 PH/53 This is related to PH/52, but is more general: for any failing address,
2057 when detailed error information was permitted to be returned to the
2058 sender, but the error was temporary, then after the final timeout, only
2059 "retry timeout exceeded" was returned. Now it returns the full error as
2060 well as "retry timeout exceeded".
2062 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2063 do this, and (what is worse) MTAs that accept it.
2065 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2066 will now be deprecated.
2068 PH/56 New os.c-cygwin from the Cygwin maintainer.
2070 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2071 formats (previously only available in exim formats via -bpr, -bpru,
2072 and -bpra. Now also available in native and exiqgrep formats)
2074 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2075 with very large, slow to parse queues
2077 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2079 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2081 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2082 responses to authentication challenges, though it was showing the
2083 challenges; (ii) I've removed the CR characters from the debug output for
2086 PH/58 Allow for the insertion of a newline as well as a space when a string
2087 is turned into more than one encoded-word during RFC 2047 encoding. The
2088 Sieve code now uses this.
2090 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2091 data_4xx, lost_connection, tls_required.
2093 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2094 message was being sent as an SMTP response.
2096 PH/61 Add -l and -k options to exicyclog.
2098 PH/62 When verifying, if an address was redirected to one new address, so that
2099 verification continued, and the new address failed or deferred after
2100 having set something in $address_data, the value of $address_data was not
2101 passed back to the ACL. This was different to the case when no
2102 redirection occurred. The value is now passed back in both cases.
2104 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2105 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2106 use login_cap.h, so on its own it isn't the distinguishing feature. The
2107 new name refers directly to the setclassresources() function.
2109 PH/65 Added configuration files for NetBSD3.
2111 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2113 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2114 is preferred over IPv4.
2116 PH/68 The bounce_return_message and bounce_return_body options were not being
2117 honoured for bounces generated during the reception of non-SMTP messages.
2118 In particular, this applied to messages rejected by the ACL. This bug has
2119 been fixed. However, if bounce_return_message is true and bounce_return_
2120 body is false, the headers that are returned for a non-SMTP message
2121 include only those that have been read before the error was detected.
2122 (In the case of an ACL rejection, they have all been read.)
2124 PH/69 The HTML version of the specification is now built in a directory called
2125 spec_html instead of spec.html, because the latter looks like a path with
2126 a MIME-type, and this confuses some software.
2128 PH/70 Catch two compiler warnings in sieve.c.
2130 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2131 function verify_get_ident() calls ip_connect() to connect a socket, but
2132 if the "connect()" function timed out, ip_connect() used to close the
2133 socket. However, verify_get_ident() also closes the socket later, and in
2134 between Exim writes to the log, which may get opened at this point. When
2135 the socket was closed in ip_connect(), the log could get the same file
2136 descriptor number as the socket. This naturally causes chaos. The fix is
2137 not to close the socket in ip_connect(); the socket should be closed by
2138 the function that creates it. There was only one place in the code where
2139 this was missing, in the iplookup router, which I don't think anybody now
2140 uses, but I've fixed it anyway.
2142 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2143 well as to direct DNS lookups. Otherwise the handling of names in host
2144 lists is inconsistent and therefore confusing.
2150 PH/01 Two changes to the default runtime configuration:
2152 (1) Move the checks for relay_from_hosts and authenticated clients from
2153 after to before the (commented out) DNS black list checks.
2155 (2) Add control=submission to the relay_from_hosts and authenticated
2156 clients checks, on the grounds that messages accepted by these
2157 statements are most likely to be submissions.
2159 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2161 (1) Generate an error if the third argument for the ${prvs expansion is
2164 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2167 (3) Reset the variables that are obtained from the first argument of
2168 ${prvscheck and used in the second argument before leaving the code,
2169 because their memory is reclaimed, so using them afterwards may do
2172 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2173 one (it's much easier than Tom thought :-).
2175 (5) Because of (4), we can now allow for the use of $prvscheck_result
2176 inside the third argument.
2178 PH/03 For some reason, the default setting of PATH when running a command from
2179 a pipe transport was just "/usr/bin". I have changed it to
2182 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2183 anything to be listed in the output from -bV.
2185 PH/05 When a filter generated an autoreply, the entire To: header line was
2186 quoted in the delivery log line, like this:
2188 => >A.N.Other <ano@some.domain> <original@ddress> ...
2190 This has been changed so that it extracts the operative address. There
2191 may be more than one such address. If so, they are comma-separated, like
2194 => >ano@some.domain,ona@other.domain <original@ddress> ...
2196 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2197 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2198 not being looked up in the rDNS to get a host name, Exim was showing the
2199 IP address twice in Received: lines, even though the IP addresses were
2200 identical. For example:
2202 Received: from [1.2.3.4] (helo=[1.2.3.4])
2204 However, if the real host name was known, it was omitting the HELO data
2205 if it matched the actual IP address. This has been tidied up so that it
2206 doesn't show the same IP address twice.
2208 PH/07 When both +timestamp and +memory debugging was on, the value given by
2209 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2210 called by the debug printing, thereby overwriting the timestamp buffer.
2211 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2213 PH/08 When the original message was included in an autoreply transport, it
2214 always said "this is a copy of the message, including all the headers",
2215 even if body_only or headers_only was set. It now gives an appropriate
2218 PH/09 Applied a patch from the Sieve maintainer which:
2220 o fixes some comments
2221 o adds the (disabled) notify extension core
2222 o adds some debug output for the result of if/elsif tests
2223 o points to the current vacation draft in the documentation
2224 and documents the missing references header update
2228 o fixes a bug in processing the envelope test (when testing
2229 multiple envelope elements, the last element determinted the
2232 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2233 Electronic Mail") by including:
2235 Auto-submitted: auto-generated
2237 in the messages that it generates (bounce messages and others, such as
2238 warnings). In the case of bounce messages for non-SMTP mesages, there was
2239 also a typo: it was using "Auto_submitted" (underscore instead of
2240 hyphen). Since every message generated by Exim is necessarily in response
2241 to another message, thes have all been changed to:
2243 Auto-Submitted: auto-replied
2245 in accordance with these statements in the RFC:
2247 The auto-replied keyword:
2249 - SHOULD be used on messages sent in direct response to another
2250 message by an automatic process,
2252 - MUST NOT be used on manually-generated messages,
2254 - MAY be used on Delivery Status Notifications (DSNs) and Message
2255 Disposition Notifications (MDNs),
2257 - MUST NOT be used on messages generated by automatic or periodic
2258 processes, except for messages which are automatic responses to
2261 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2262 to the default Received: header definition.
2264 PH/12 Added log selector acl_warn_skipped (default on).
2266 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2267 variables because (a) they are in the wrong storage pool and (b) even if
2268 they were copied, it wouldn't work properly because of the caching.
2270 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2271 checking when decoding. Apparently there are clients that generate
2272 overlong encoded strings. Why am I not surprised?
2274 PH/15 If the first argument of "${if match_address" was not empty, but did not
2275 contain an "@" character, Exim crashed. Now it writes a panic log message
2276 and treats the condition as false.
2278 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2280 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2281 extension "envelope-auth". The code is finished and in agreement with
2282 other implementations, but there is no documentation so far and in fact,
2283 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2284 not changing the active code.
2286 Print executed "if" and "elsif" statements when debugging is used. This
2287 helps a great deal to understand what a filter does.
2289 Document more things not specified clearly in RFC3028. I had all this
2290 sorted out, when out of a sudden new issues came to my mind. Oops."
2292 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2295 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2296 canonical form (as documented). However, after a host name lookup from
2297 the IP address, check_host() was doing a simple string comparison with
2298 addresses acquired from the DNS when checking that the found name did
2299 have the original IP as one of its addresses. Since any found IPv6
2300 addresses are likely to be in abbreviated form, the comparison could
2301 fail. Luckily, there already exists a function for doing the comparison
2302 by converting both addresses to binary, so now that is used instead of
2303 the text comparison.
2305 PH/20 There was another similar case to PH/19, when a complete host name was
2306 given in a host list; looking up its IP address could give an abbreviated
2307 form, whereas the current host's name might or might not be abbreviated.
2308 The same fix has been applied.
2314 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2315 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2318 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2319 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2321 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2323 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2324 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2325 Also, for "Auto-Submitted", treat anything other than "no" as
2326 non-personal, in accordance with RFC 3834. (Previously it treated
2327 anything starting "auto-" as non-personal.)
2329 TF/01 The control=submission/name=... option had a problem with syntax
2330 errors if the name included a slash character. The /name= option
2331 now slurps the rest of the string, so it can include any characters
2332 but it must come last in the list of options (after /sender_retain
2335 PH/05 Some modifications to the interface to the fake nameserver for the new
2343 TK/01 Added the "success_on_redirect" address verification option. See
2344 NewStuff for rationale and an example.
2346 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2348 PH/02 Patch to exigrep to allow it to work on syslog lines.
2350 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2351 fread() to skip over the body file's header line, because in Cygwin the
2352 header line is locked and is inaccessible.
2354 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2355 co-exist for some time) to make it clear that it is the Exim ID that is
2356 referenced, not the Message-ID: header line.
2358 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2359 string_format() function, because snprintf() does not exist on all
2362 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2363 ${stat: expansion item.
2365 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2366 protocol synchronization error", to keep the pedants happy.
2368 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2369 well as for IRIX systems, when gcc is being used. See the host.c source
2372 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2374 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2377 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2378 transport and to the smtp transport in LMTP mode.
2380 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2382 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2383 run a filter in a subprocess. This could lead to confusion in subsequent
2384 lookups in the parent process. There should also be a search_tidyup() at
2385 the end of the subprocess.
2387 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2388 only if the host matched helo_try_verify_hosts, which caused the
2389 verification to occur when the EHLO/HELO command was issued. The ACL just
2390 tested the remembered result. Now, if a previous verification attempt has
2391 not happened, "verify = helo" does it there and then.
2393 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2395 TK/03 Fix log output including CR from clamd.
2397 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2398 exist provoked a memory error which could cause a segfault.
2400 PH/15 Installed PCRE 6.2
2402 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2404 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2405 of the problem. Specifically, suggested +O2 rather than +O1 for the
2408 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2410 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2411 record, the port was not correctly logged when the outgoing_port log
2412 selector was set (it logged the transort's default port).
2414 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2415 fallback_hosts, and "hosts" in the smtp transport.
2417 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2418 host errors such as "Connection refused".
2420 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2421 authentication with radiusclient 0.4.9:
2423 - Error returned from rc_read_config was caught wrongly
2424 - Username/password not passed on to radius server due to wrong length.
2426 The presumption is that some radiusclient API changes for 4.51/PH/17
2427 were not taken care of correctly. The code is still untested by me (my
2428 Linux distribution still has 0.3.2 of radiusclient), but it was
2429 contributed by a Radius user.
2431 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2432 expanding the "port" option of the smtp transport.
2434 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2435 while reading a MIME header. Thanks to Tom Hughes for a patch.
2437 PH/24 Include config.h inside local_scan.h so that configuration settings are
2440 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2441 that in an ACL for RCPT (for example), you can examine exactly what was
2444 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2445 commands, but it was not correctly comparing the address with the actual
2446 client host address. Thus, it would show the EHLO address in Received:
2447 header lines when this was not necessary.
2449 PH/27 Added the % operator to ${eval:}.
2451 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2452 it should be ignoring failures (because with -C, for example, it has lost
2453 privilege). It wasn't ignoring creation failures other than "already
2456 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2459 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2460 because an input line was too long, either on its own, or by virtue of
2461 too many continuations, the temporary file was not being removed, and the
2462 return code was incorrect.
2464 PH/31 Missing "BOOL" in function definition in filtertest.c.
2466 PH/32 Applied Sieve patches from the maintainer.
2468 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2470 PH/33 Added "verify = not_blind".
2472 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2473 Local/Makefile (with some defaults set). These are used in built scripts
2474 such as exicyclog, but they have never been used in the exim_install
2475 script (though there are many overriding facilities there). I have
2476 arranged that the exim_install script now takes note of these two
2479 PH/35 Installed configuration files for Dragonfly.
2481 PH/36 When a locally submitted message by a trusted user did not contain a
2482 From: header, and the sender address was obtained from -f or from an SMTP
2483 MAIL command, and the trusted user did not use -F to supply a sender
2484 name, $originator_name was incorrectly used when constructing a From:
2485 header. Furthermore, $originator_name was used for submission mode
2486 messages from external hosts without From: headers in a similar way,
2487 which is clearly wrong.
2489 PH/37 Added control=suppress_local_fixups.
2491 PH/38 When log_selector = +received_sender was set, and the addition of the
2492 sender made the log line's construction buffer exactly full, or one byte
2493 less than full, an overflow happened when the terminating "\n" was
2496 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2497 when the result of a list match is failure because a DNS lookup failed.
2499 PH/40 RM_COMMAND is now used in the building process.
2501 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2502 the "build-* directories that it finds.
2504 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2505 address in a domain literal was a prefix of an interface address.
2507 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2508 when verifying a sender address, unless rewrite_headers is false.
2510 PH/44 Wrote a long comment about why errors_to addresses are verified as
2511 recipients, not senders.
2513 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2514 the ratelimit ACL was added.
2516 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2518 PH/46 Added extra information about PostgreSQL errors to the error string.
2520 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2521 avoiding the need to install special zones in a real server. This is
2522 backwards compatible; if it can't find the fake resolver, it drops back.
2523 Thus, both old and new test suites can be run.
2525 TF/02 Added util/ratelimit.pl
2527 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2528 clock is set back in time.
2530 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2531 Candler <B.Candler@pobox.com>.
2533 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2534 OK for addresses that are the result of redirections.
2536 PH/48 A number of further additions for the benefit of the new test suite,
2537 including a fake gethostbyname() that interfaces to the fake DNS resolver
2540 TF/06 The fix for widen_domains has also been applied to qualify_single and
2541 search_parents which are the other dnslookup options that can cause
2544 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2547 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2549 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2550 service" if it encountered an MX record with an empty target hostname.
2551 The message is now "an MX or SRV record indicated no SMTP service".
2553 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2554 if the DNS of the sending site is misconfigured. This is quite a
2555 common situation. This change restores the behaviour of treating a
2556 helo verification defer as a failure.
2558 PH/51 If self=fail was set on a router, the bounce message did not include the
2559 actual error message.
2565 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2567 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2568 command itself ended in error, the underlying message about the transport
2569 filter timeout was being overwritten with the pipe command error. Now the
2570 underlying error message should be appended to the second error message.
2572 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2574 PH/02 Reduce the amount of output that "make" produces by default. Full output
2575 can still be requested.
2577 PH/03 The warning log line about a condition test deferring for a "warn" verb
2578 was being output only once per connection, rather than after each
2579 occurrence (because it was using the same function as for successful
2580 "warn" verbs). This seems wrong, so I have changed it.
2582 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2583 it should not have, which might have caused a crash in the right
2584 circumstances, but probably never did.
2586 PH/04 Installed a modified version of Tony Finch's patch to make submission
2587 mode fix the return path as well as the Sender: header line, and to
2588 add a /name= option so that you can make the user's friendly name appear
2591 TF/03 Added the control = fakedefer ACL modifier.
2593 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2594 Mark Lowes for thorough testing.
2596 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2598 TK/03 Merged latest SRS patch from Miles Wilton.
2600 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2601 with the definition in sysexits.h (which is #included earlier).
2602 Fortunately, Exim does not actually use EX_OK. The code used to try to
2603 preserve the sysexits.h value, by assumimg that macro definitions were
2604 scanned for macro replacements. I have been disabused of this notion,
2605 so now the code just undefines EX_OK before #including unistd.h.
2607 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2608 in the smtp transport. When a block could not be written in a single
2609 write() function, the timeout was being re-applied to each part-write.
2610 This seems wrong - if the receiver was accepting one byte at a time it
2611 would take for ever. The timeout is now adjusted when this happens. It
2612 doesn't have to be particularly precise.
2614 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2615 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2617 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2618 without a domain if the check to <postmaster@domain> fails.
2620 SC/01 Eximstats: added -xls and the ability to specify output files
2621 (patch written by Frank Heydlauf).
2623 SC/02 Eximstats: use FileHandles for outputing results.
2625 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2627 SC/04 Eximstats: fixed display of large numbers with -nvr option
2629 SC/05 Eximstats: fixed merging of reports with empty tables.
2631 SC/06 Eximstats: added the -include_original_destination flag
2633 SC/07 Eximstats: removed tabs and trailing whitespace.
2635 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2637 TK/06 MBOX spool code: Add real "From " MBOX separator line
2638 so the .eml file is really in mbox format (even though
2639 most programs do not really care). Patch from Alex Miller.
2641 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2642 The latter is generated from $received_to and is only set if the
2643 message has one envelope recipient. SA can use these headers,
2644 obviously out-of-the-box. Patch from Alex Miller.
2646 PH/08 The ${def test on a variable was returning false if the variable's
2647 value was "0", contrary to what the specification has always said!
2648 The result should be true unless the variable is empty.
2650 PH/09 The syntax error of a character other than { following "${if
2651 def:variable_name" (after optional whitespace) was not being diagnosed.
2652 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2653 accidental colon was present, for example, could give incorrect results.
2655 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2656 result is used (not including appendfile, where other changes are about
2659 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2660 This involved changing a lot of size variables from int to off_t. It
2661 should work with maildirs and everything.
2663 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2664 spamd dying while we are connected to it.
2666 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2669 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2670 function for BDB 4.3.
2672 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2674 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2675 the sources for PCRE have been re-arranged and more files are now
2678 PH/15 The code I had for printing potentially long long variables in PH/11
2679 above was not the best (it lost precision). The length of off_t variables
2680 is now inspected at build time, and an appropriate printing format (%ld
2681 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2682 to be "long long int" or "long int". This is needed for the internal
2683 formatting function string_vformat().
2685 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2686 the configuration file to be ":syslog", then the script "guesses" where
2687 the logs files are, rather than using the compiled in default. In our
2688 case the guess is not the same as the compiled default, so the script
2689 suddenly stopped working when I started to use syslog. The patch checks
2690 to see if log_file_path is "". If so, it attempts to read it from exim
2691 with no configuration file to get the compiled in version, before it
2692 falls back to the previous guessing code."
2694 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2695 implementing BATV in an Exim configuration. See NewStuff for the gory
2698 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2699 Makefile that are specific to HP-UX.
2701 PH/18 If the "use_postmaster" option was set for a recipient callout together
2702 with the "random" option, the postmaster address was used as the MAIL
2703 FROM address for the random test, but not for the subsequent recipient
2704 test. It is now used for both.
2706 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2707 patch removes a few documentation additions to RFC 3028, because the
2708 latest draft now contains them. It adds the new en;ascii-case comparator
2709 and a new error check for 8bit text in MIME parts. Comparator and
2710 require names are now matched exactly. I enabled the subaddress
2711 extension, but it is not well tested yet (read: it works for me)."
2713 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2714 rework some of the code of TK/09 above to avoid the hardwired use of
2715 "%lld" and "long long". Replaced the call to snprintf() with a call to
2718 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2719 records point to non-existent hosts", "retry timeout exceeded", and
2720 "retry time not reached for any host after a long failure period".
2722 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2723 experimental DomainKeys support:
2725 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2726 (2) On an error such as an illegally used "control", the wrong name for
2727 the control was given.
2729 These problems did NOT occur unless DomainKeys support was compiled.
2731 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2733 PH/24 Added ${if match_ip condition.
2735 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2736 databases so that it will be absolutely obvious if a crash occurs in the
2737 DB library. This is a regular occurrence (often caused by mis-matched
2740 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2741 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2742 picked up on a user's system that detects such things. There doesn't seem
2743 to be a gcc warning option for this - only an attribute that has to be
2744 put on the function's prototype. It seems that in Fedora Core 4 they have
2745 set this on a number of new functions. No doubt there will be more in due
2748 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2749 specify a transport. However, if an address that was verified by such a
2750 router was the subject of a callout, Exim crashed because it tried to
2751 read the rcpt_include_affixes from the non-existent transport. Now it
2752 just assumes that the setting of that option is false. This bug was
2753 introduced by 4.51/PH/31.
2755 PH/28 Changed -d+all to exclude +memory, because that information is very
2756 rarely of interest, but it makes the output a lot bigger. People tend to
2757 do -d+all out of habit.
2759 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2760 code in os-type was giving problems when libc.so lives in lib64, like on
2763 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2764 aren't the modern standard, and it seems that some systems' include files
2765 don't always have them. Exim was already checking for some of the newer
2766 ones like T_AAAA, and defining it itself. I've added checks for all the
2767 record types that Exim uses.
2769 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2770 not automatically generating a new one, as it is supposed to. This
2771 prevented TLS from working. If the file did exist, but contained invalid
2772 data, a new version was generated, as expected. It was only the case of a
2773 non-existent file that was broken.
2775 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2776 with a change in libdomainkeys > 0.64.
2778 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2779 from DNS. If the selector record carries the flag, it now has
2780 precedence over the domain-wide flag.
2782 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2784 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2785 the use of a port name that isn't defined in /etc/services) occurred, the
2786 message was deferred as in a normal delivery, and thus remained on the
2787 spool, instead of being failed because of the mua_wrapper setting. This
2788 is now fixed, and I tidied up some of the mua_wrapper messages at the
2791 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2792 the messages in a hash of arrays rather than using individual hashes.
2793 This is a bit cleaner and results in dramatic memory savings, albeit
2794 at a slight CPU cost.
2796 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2797 as requested by Marc Sherman.
2799 SC/10 Eximstats: added histograms for user specified patterns as requested
2802 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2804 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2805 fopen() in the content-scanning modules that did not already have it.
2811 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2812 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2814 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2816 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2818 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2819 to test Sieve filters that use "vacation".
2821 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2822 that changes the way the GnuTLS parameters are stored in the cache file.
2823 The new format can be generated externally. For backward compatibility,
2824 if the data in the cache doesn't make sense, Exim assumes it has read an
2825 old-format file, and it generates new data and writes a new file. This
2826 means that you can't go back to an older release without removing the
2829 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2830 work if there are any delivery delays because "one_time" forces the
2831 parent to be marked "delivered", so its unseen clone is never tried
2832 again. For this reason, Exim now forbids the simultaneous setting of
2835 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2836 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2837 entry if you want to know the details. The fix, however, neglected to
2838 consider the case when local delivery batching is involved. The test for
2839 "previously delivered" was not happening when checking to see if an
2840 address could be batched with a previous (undelivered) one; under
2841 certain circumstances this could lead to multiple deliveries to the same
2844 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2845 in its include files, and this causes problems building Exim.
2847 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2848 header_syntax) but Exim was just ignoring anything given after a slash.
2849 In particular, this caused confusion with an attempt to use "verify =
2850 reverse_host_lookup/defer_ok". An error is now given when options are
2851 supplied for verify items that do not have them. (Maybe reverse_host_
2852 lookup should have a defer_ok option, but that's a different point.)
2854 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2855 defined by RFC 821) to 2048, because there were problems with some AUTH
2856 commands, and RFC 1869 says the size should be increased for extended
2857 SMTP commands that take arguments.
2859 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2862 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2863 "unknown" error; now it says that the functionality isn't in the binary.
2865 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2866 an address' error message when a string expansion fails (syntax or
2867 whatever). Otherwise the password may appear in the log. Following change
2868 PH/42 below, there is no longer a chance of it appearing in a bounce
2871 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2873 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2874 of its addresses was ever tried. (Bugzilla bug #2.)
2876 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2877 the result incorrectly in the debug output. (It correctly added a newline
2878 to what was transported.)
2880 TF/01 Added $received_time.
2882 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2883 commented out examples of how to interface to a virus scanner and to
2884 SpamAssassin. Also added commented examples of av_scanner and
2885 spamd_address settings.
2887 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2888 and controls are allowed in which ACLs. There were a couple of minor
2889 errors. Some of the entries in the conditions table (which is a table of
2890 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2891 negation of where the condition IS allowed.
2893 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2895 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2896 header file does not have a version number, so I've had to invent a new
2897 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2898 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2899 radiusclient), but it was contributed by a Radius user.
2901 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2902 files or directories, for OpenSSL.
2904 PH/19 When an Exim process that is running as root has to create an Exim log
2905 file, it does so in a subprocess that runs as exim:exim so as to get the
2906 ownership right at creation (otherwise, other Exim processes might see
2907 the file with the wrong ownership). There was no test for failure of this
2908 fork() call, which would lead to the process getting stuck as it waited
2909 for a non-existent subprocess. Forks do occasionally fail when resources
2910 run out. I reviewed all the other calls to fork(); they all seem to check
2913 PH/20 When checking for unexpected SMTP input at connect time (before writing
2914 the banner), Exim was not dealing correctly with a non-positive return
2915 from the read() function. If the client had disconnected by this time,
2916 the result was a log entry for a synchronization error with an empty
2917 string after "input=" when read() returned zero. If read() returned -1
2918 (an event I could not check), uninitialized data bytes were printed.
2919 There were reports of junk text (parts of files, etc) appearing after
2922 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2924 PH/22 Added support for macro redefinition, and (re)definition in between
2925 driver and ACL definitions.
2927 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2928 forgetting to use the resulting value; it was using the unexpanded value.
2930 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2931 hadn't been configured. The fix is from Juergen Kreileder, who
2932 understands it better than I do:
2934 "Here's what I see happening with three configured cyrus_sasl
2935 authenticators configured (plain, login, cram-md5):
2937 On startup auth_cyrus_sasl_init() gets called for each of these.
2938 This means three calls to sasl_listmech() without a specified mech_list.
2939 => SASL tests which mechs of all available mechs actually work
2940 => three warnings about OTP not working
2941 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2943 With the patch, sasl_listmech() also gets called three times. But now
2944 SASL's mech_list option is set to the server_mech specified in the the
2945 authenticator. Or in other words, the answer from sasl_listmech()
2946 gets limited to just the mech you're testing for (which is different
2948 => the return list contains just 'plain' or 'login', 'cram-md5' or
2949 nothing depending on the value of ob->server_mech.
2951 I've just tested the patch: Authentication still works fine,
2952 unavailable mechs specified in the exim configuration are still
2953 caught, and the auth.log warnings about OTP are gone."
2955 PH/25 When debugging is enabled, the contents of the command line are added
2956 to the debugging output, even when log_selector=+arguments is not
2959 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2960 answer is "GNU", and only if the return is "GNU/something" is the answer
2963 PH/27 $acl_verify_message is now set immediately after the failure of a
2964 verification in an ACL, and so is available in subsequent modifiers. In
2965 particular, the message can be preserved by coding like this:
2967 warn !verify = sender
2968 set acl_m0 = $acl_verify_message
2970 Previously, $acl_verify_message was set only while expanding "message"
2971 and "log_message" when a very denied access.
2973 PH/28 Modified OS/os.c-Linux with
2975 -#ifndef OS_LOAD_AVERAGE
2976 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2978 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2979 nomenclature these days.)
2981 PH/29 Installed patch from the Sieve maintainer that adds the options
2982 sieve_useraddress and sieve_subaddress to the redirect router.
2984 PH/30 In these circumstances:
2985 . Two addresses routed to the same list of hosts;
2986 . First host does not offer TLS;
2987 . First host accepts first address;
2988 . First host gives temporary error to second address;
2989 . Second host offers TLS and a TLS session is established;
2990 . Second host accepts second address.
2991 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2992 and peerdn, if requested) that were in fact used only for the second
2995 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2996 attention to any local part prefix or suffix that was matched by the
2997 router that accepted the address. It now behaves in the same way as it
2998 does for delivery: the affixes are removed from the local part unless
2999 rcpt_include_affixes is set on the transport.
3001 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3002 timeout during the DATA phase of an incoming message.
3004 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3005 applied a patch sanctioned by the Sieve maintainer.
3007 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3008 the uid or gid is negative. A case of a negative gid caused this to be
3009 noticed. The fix allows for either to be negative.
3011 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3012 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3015 PH/36 The change PH/12 above was broken. Fixed it.
3017 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3018 the grounds that routing the same address twice would always produce the
3019 same answer. This might have been true once, but it is certainly no
3020 longer true now. Routing a child address may depend on the previous
3021 routing that produced that child. Some complicated redirection strategies
3022 went wrong when messages had multiple recipients, and made Exim's
3023 behaviour dependent on the order in which the addresses were given.
3025 I have moved the duplicate checking until after the routing is complete.
3026 Exim scans the addresses that are assigned to local and remote
3027 transports, and removes any duplicates. This means that more work will be
3028 done, as duplicates will always all be routed, but duplicates are
3029 presumably rare, so I don't expect this is of any significance.
3031 For deliveries to pipes, files, and autoreplies, the duplicate checking
3032 still happens during the routing process, since they are not going to be
3035 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3036 It corrects a timeout issue with spamd. This is Ian's comment: "The
3037 background is that sometimes spamd either never reads data from a
3038 connection it has accepted, or it never writes response data. The exiscan
3039 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3040 blindly assumes that writes won't block so it may never time out."
3042 PH/39 Allow G after quota size as well as K and M.
3044 PH/40 The value set for $authenticated_id in an authenticator may not contain
3045 binary zeroes or newlines because the value is written to log lines and
3046 to spool files. There was no check on this. Now the value is run through
3047 the string_printing() function so that such characters are converted to
3048 printable escape sequences.
3050 PH/41 $message_linecount is a new variable that contains the total number of
3051 lines in the message. Compare $body_linecount, which is the count for the
3054 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3055 bounce and delay warning messages, except in certain special cases, which
3058 (a) An SMTP error message from a remote host;
3059 (b) A message specified in a :fail: redirection;
3060 (c) A message specified in a "fail" command in a system filter;
3061 (d) A message specified in a FAIL return from the queryprogram router;
3062 (e) A message specified by the cannot_route_message router option.
3064 In these cases only, Exim does include the error details in bounce and
3065 warning messages. There are also a few cases where bland messages such
3066 as "unrouteable address" or "local delivery error" are given.
3068 PH/43 $value is now also set for the "else" part of a ${run expansion.
3070 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3071 being worked on, but at least Exim now implements the latest version to
3074 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3075 process to complete was treated as a delivery failure, a timeout while
3076 writing the message to the pipe was logged, but erroneously treated as a
3077 successful delivery. Such timeouts include transport filter timeouts. For
3078 consistency with the overall process timeout, these timeouts are now
3079 treated as errors, giving rise to delivery failures by default. However,
3080 there is now a new Boolean option for the pipe transport called
3081 timeout_defer, which, if set TRUE, converts the failures into defers for
3082 both kinds of timeout. A transport filter timeout is now identified in
3085 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3086 systems where "make" and "gmake" are different, calling "gmake" at top
3087 level broke things. I've arranged for the value of $(MAKE) to be passed
3088 from the Makefile to this script so that it can call the same version of
3092 A note about Exim versions 4.44 and 4.50
3093 ----------------------------------------
3095 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3096 changes of various kinds. As a consequence, a big documentation update was
3097 needed. This delayed the release for rather longer than seemed good, especially
3098 in the light of a couple of (minor) security issues. Therefore, the changes
3099 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3100 release. So 4.44 and 4.50 are in effect two different branches that both start
3103 I have left the 4.50 change log unchanged; it contains all the changes since
3104 4.43. The change log for 4.44 is below; many of its items are identical to
3105 those for 4.50. This seems to be the most sensible way to preserve the
3106 historical information.
3112 1. Minor wording change to the doc/README.SIEVE file.
3114 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3115 computation of the current number of files was incorrect.
3117 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3118 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3121 4. Give more explanation in the error message when the command for a transport
3122 filter fails to execute.
3124 5. There are several places where Exim runs a non-Exim command in a
3125 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3126 was being done only for the command run by the queryprogram router. It is
3127 now done for all such subprocesses. The other cases are: ${run, transport
3128 filters, and the commands run by the lmtp and pipe transports.
3130 6. Added CONFIGURE_GROUP build-time option.
3132 7. Some older OS have a limit of 256 on the maximum number of file
3133 descriptors. Exim was using setrlimit() to set 1000 as a large value
3134 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3135 systems. I've change it so that if it can't get 1000, it tries for 256.
3137 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3138 was an oversight, and furthermore, ever since the addition of extra
3139 controls (e.g. 4.43/32), the checks on when to allow different forms of
3140 "control" were broken. There should now be diagnostics for all cases when a
3141 control that does not make sense is encountered.
3143 9. Added the /retain_sender option to "control=submission".
3145 10. $recipients is now available in the predata ACL (oversight).
3147 11. Tidy the search cache before the fork to do a delivery from a message
3148 received from the command line. Otherwise the child will trigger a lookup
3149 failure and thereby defer the delivery if it tries to use (for example) a
3150 cached ldap connection that the parent has called unbind on.
3152 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3153 of $address_data from the recipient verification was clobbered by the
3154 sender verification.
3156 13. The value of address_data from a sender verification is now available in
3157 $sender_address_data in subsequent conditions in the ACL statement.
3159 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3161 15. Added a new option "connect=<time>" to callout options, to set a different
3164 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3165 was its contents. (It was OK if the option was not defined at all.)
3167 17. A "Completed" log line is now written for messages that are removed from
3168 the spool by the -Mrm option.
3170 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3171 information about exactly what failed.
3173 19. Added -dd to debug only the daemon process.
3175 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3176 handles timeouts, both on the server side and network timeouts. Renamed the
3177 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3179 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3180 It is now set to "smtps".
3182 22. $host_address is now set to the target address during the checking of
3183 ignore_target_hosts.
3185 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3186 being passed; this would have caused $sender_host_name to have been used if
3187 matching the list had actually called for a host name (not very likely,
3188 since this list is usually IP addresses). A host name is now passed as
3191 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3192 code that specifies a non-restarting handler (typically sigaction() in
3193 modern systems) in an attempt to fix a rare and obscure crash bug.
3195 25. Narrowed the window for a race in the daemon that could cause it to ignore
3196 SIGCHLD signals. This is not a major problem, because they are used only to
3197 wake it up if nothing else does.
3199 26. A malformed maildirsize file could cause Exim to calculate negative values
3200 for the mailbox size or file count. Odd effects could occur as a result.
3201 The maildirsize information is now recalculated if the size or filecount
3204 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3205 support for a long time. Removed HAVE_SYS_VFS_H.
3207 28. Installed the latest version of exipick from John Jetmore.
3209 29. In an address list, if the pattern was not a regular expression, an empty
3210 subject address (from a bounce message) matched only if the pattern was an
3211 empty string. Non-empty patterns were not even tested. This was the wrong
3212 because it is perfectly reasonable to use an empty address as part of a
3213 database query. An empty address is now tested by patterns that are
3214 lookups. However, all the other forms of pattern expect the subject to
3215 contain a local part and a domain, and therefore, for them, an empty
3216 address still always fails if the pattern is not itself empty.
3218 30. Exim went into a mad DNS loop when attempting to do a callout where the
3219 host was specified on an smtp transport, and looking it up yielded more
3220 than one IP address.
3222 31. Re-factored the code for checking spool and log partition space into a
3223 function that finds that data and another that does the check. The former
3224 is then used to implement four new variables: $spool_space, $log_space,
3225 $spool_inodes, and $log_inodes.
3227 32. The RFC2047 encoding function was originally intended for short strings
3228 such as real names; it was not keeping to the 75-character limit for
3229 encoded words that the RFC imposes. It now respects the limit, and
3230 generates multiple encoded words if necessary. To be on the safe side, I
3231 have increased the buffer size for the ${rfc2047: expansion operator from
3234 33. It is now permitted to omit both strings after an "if" condition; if the
3235 condition is true, the result is "true". As before, when the second string
3236 is omitted, a false condition yields an empty string. This makes it less
3237 cumbersome to write custom ACL and router conditions.
3239 34. Failure to deliver a bounce message always caused it to be frozen, even if
3240 there was an errors_to setting on the router. The errors_to setting is now
3243 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3244 canonical form (fully expanded) before being placed in
3245 $sender_host_address.
3247 36. The table in the code that translates DNS record types into text (T_A to
3248 "A" for instance) was missing entries for NS and CNAME. It is just possible
3249 that this could have caused confusion if both these types were looked up
3250 for the same domain, because the text type is used as part of Exim's
3251 per-process caching. But the chance of anyone hitting this buglet seems
3254 37. The dnsdb lookup has been extended in a number of ways.
3256 (1) There is a new type, "zns", which walks up the domain tree until it
3257 finds some nameserver records. It should be used with care.
3259 (2) There is a new type, "mxh", which is like "mx" except that it returns
3260 just the host names, not the priorities.
3262 (3) It is now possible to give a list of domains (or IP addresses) to be
3263 looked up. The behaviour when one of the lookups defers can be
3264 controlled by a keyword.
3266 (4) It is now possible to specify the separator character for use when
3267 multiple records are returned.
3269 38. The dnslists ACL condition has been extended: it is now possible to supply
3270 a list of IP addresses and/or domains to be looked up in a particular DNS
3273 39. Added log_selector=+queue_time_overall.
3275 40. When running the queue in the test harness, wait just a tad after forking a
3276 delivery process, to get repeatability of debugging output.
3278 41. Include certificate and key file names in error message when GnuTLS fails
3279 to set them up, because the GnuTLS error message doesn't include the name
3280 of the failing file when there is a problem reading it.
3282 42. Allow both -bf and -bF in the same test run.
3284 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3286 44. The "Exiscan patch" is now merged into the mainline Exim source.
3288 45. Sometimes the final signoff response after QUIT could fail to get
3289 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3290 < 0 before doing a fflush(). This bug looks as though it goes back to the
3291 introduction of TLS in release 3.20, but "sometimes" must have been rare
3292 because the tests only now provoked it.
3294 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3295 (this can affect the format of dates).
3297 47. exim_tidydb, when checking for the continued existence of a message for
3298 which it has found a message-specific retry record, was not finding
3299 messages that were in split spool directories. Consequently, it was
3300 deleting retry records that should have stayed in existence.
3302 48. Steve fixed some bugs in eximstats.
3304 49. The SPA authentication driver was not abandoning authentication and moving
3305 on to the next authenticator when an expansion was forced to fail,
3306 contradicting the general specification for all authenticators. Instead it
3307 was generating a temporary error. It now behaves as specified.
3309 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3310 (the order specifies the preference for clients). The order is now AES256,
3311 AES128, 3DES, ARCFOUR128.
3313 51. Small patch to Sieve code - explicitly set From: when generating an
3316 52. Exim crashed if a remote delivery caused a very long error message to be
3317 recorded - for instance if somebody sent an entire SpamAssassin report back
3318 as a large number of 550 error lines. This bug was coincidentally fixed by
3319 increasing the size of one of Exim's internal buffers (big_buffer) that
3320 happened as part of the Exiscan merge. However, to be on the safe side, I
3321 have made the code more robust (and fixed the comments that describe what
3324 53. Now that there can be additional text after "Completed" in log lines (if
3325 the queue_time_overall log selector is set), a one-byte patch to exigrep
3326 was needed to allow it to recognize "Completed" as not the last thing in
3329 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3330 patch that reportedly fixes this has been added. I am not expert enough to
3331 create a test for it. This is what the patch creator wrote:
3333 "I found a little strange behaviour of ldap code when working with
3334 Windows 2003 AD Domain, where users was placed in more than one
3335 Organization Units. When I tried to give exim partial DN, the exit code
3336 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3337 But simultaneously result of request was absolutely normal ldap result,
3338 so I produce this patch..."
3340 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3341 REFERENCE, so I have modified the code to exclude the patch when that macro
3344 55. Some experimental protocols are using DNS PTR records for new purposes. The
3345 keys for these records are domain names, not reversed IP addresses. The
3346 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3347 leaves it alone. Component reversal etc. now happens only for IP addresses.
3350 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3352 57. Double the size of the debug message buffer (to 2048) so that more of very
3353 long debug lines gets shown.
3355 58. The exicyclog utility now does better if the number of log files to keep
3356 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3358 59. Two changes related to the smtp_active_hostname option:
3360 (1) $smtp_active_hostname is now available as a variable.
3361 (2) The default for smtp_banner uses $smtp_active_hostname instead
3362 of $primary_hostname.
3364 60. The host_aton() function is supposed to be passed a string that is known
3365 to be a valid IP address. However, in the case of IPv6 addresses, it was
3366 not checking this. This is a hostage to fortune. Exim now panics and dies
3367 if the condition is not met. A case was found where this could be provoked
3368 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3369 components; fortuitously, this particular loophole had already been fixed
3370 by change 4.50/55 above.
3372 If there are any other similar loopholes, the new check in host_aton()
3373 itself should stop them being exploited. The report I received stated that
3374 data on the command line could provoke the exploit when Exim was running as
3375 exim, but did not say which command line option was involved. All I could
3376 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3377 running as the user.
3380 61. There was a buffer overflow vulnerability in the SPA authentication code
3381 (which came originally from the Samba project). I have added a test to the
3382 spa_base64_to_bits() function which I hope fixes it.
3385 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3386 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3388 63. The daemon start-up calls getloadavg() while still root for those OS that
3389 need the first call to be done as root, but it missed one case: when
3390 deliver_queue_load_max is set with deliver_drop_privilege. This is
3391 necessary for the benefit of the queue runner, because there is no re-exec
3392 when deliver_drop_privilege is set.
3394 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3395 This has been fixed.
3397 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3398 was in use, was not putting the data itself into the right store pool;
3399 consequently, it could be overwritten for a subsequent message in the same
3400 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3403 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3405 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3406 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3407 the function were treating the return as a boolean value, which happened to
3408 work because 0=false and not-0=true, but is not correct code.
3410 68. The host_aton() function was not handling scoped IPv6 addresses (those
3411 with, for example, "%eth0" on the end) correctly.
3413 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3414 negated items (that is, ~something) in unsigned ints. Some compilers
3415 apparently mutter when there is no cast.
3417 70. If an address verification called from an ACL failed, and did not produce a
3418 user-specific message (i.e. there was only a "system" message), nothing was
3419 put in $acl_verify_message. In this situation, it now puts the system
3422 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3423 session; change 4.31/43 added the unwanted input to the log line - except
3424 that it did not do this in the start of session case. It now does.
3426 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3427 This is wrong and can cause the other end to generate a synchronization
3428 error if it is another Exim or anything else that does the synchronization
3429 check. A QUIT command is no longer sent after a timeout.
3431 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3432 during host lookups.
3434 74. The defer_ok option of callout verification was not working if it was used
3435 when verifying addresses in header lines, that is, for this case:
3437 verify = header_sender/callout=defer_ok
3439 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3440 those file descriptors could be used for SMTP connections. If anything
3441 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3442 could be sent to the SMTP client, causing chaos. The daemon now opens
3443 stdin, stdout, and stderr to /dev/null when it puts itself into the
3446 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3447 log by default. The user can override this with suitable Perl magic.
3449 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3450 the log message when discard triggers, was not working for the DATA ACL or
3451 for the non-SMTP ACL.
3453 78. Error message wording change in sieve.c.
3455 79. If smtp_accept_max_per_host was set, the number of connections could be
3456 restricted to fewer than expected, because the daemon was trying to set up
3457 a new connection before checking whether the processes handling previous
3458 connections had finished. The check for completed processes is now done
3459 earlier. On busy systems, this bug wouldn't be noticed because something
3460 else would have woken the daemon, and it would have reaped the completed
3463 80. If a message was submitted locally by a user whose login name contained one
3464 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3465 It caused a spool format error. I have fixed the spool reading code. A
3466 related problem was that the "from" clause in the Received: line became
3467 illegal because of the space(s). It is now covered by ${quote_local_part.
3469 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3472 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3474 83. Patch from Sieve maintainer to fix unterminated string problem in
3475 "vacation" handling.
3477 84. Some minor changes to the Linux configuration files to help with other
3478 OS variants using glibc.
3480 85. One more patch for Sieve to update vacation handling to latest spec.
3483 ----------------------------------------------------
3484 See the note above about the 4.44 and 4.50 releases.
3485 ----------------------------------------------------
3491 1. Change 4.43/35 introduced a bug that caused file counts to be
3492 incorrectly computed when quota_filecount was set in an appendfile
3495 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3496 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3499 3. Give more explanation in the error message when the command for a transport
3500 filter fails to execute.
3502 4. There are several places where Exim runs a non-Exim command in a
3503 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3504 was being done only for the command run by the queryprogram router. It is
3505 now done for all such subprocesses. The other cases are: ${run, transport
3506 filters, and the commands run by the lmtp and pipe transports.
3508 5. Some older OS have a limit of 256 on the maximum number of file
3509 descriptors. Exim was using setrlimit() to set 1000 as a large value
3510 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3511 systems. I've change it so that if it can't get 1000, it tries for 256.
3513 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3514 was an oversight, and furthermore, ever since the addition of extra
3515 controls (e.g. 4.43/32), the checks on when to allow different forms of
3516 "control" were broken. There should now be diagnostics for all cases when a
3517 control that does not make sense is encountered.
3519 7. $recipients is now available in the predata ACL (oversight).
3521 8. Tidy the search cache before the fork to do a delivery from a message
3522 received from the command line. Otherwise the child will trigger a lookup
3523 failure and thereby defer the delivery if it tries to use (for example) a
3524 cached ldap connection that the parent has called unbind on.
3526 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3527 of $address_data from the recipient verification was clobbered by the
3528 sender verification.
3530 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3531 was its contents. (It was OK if the option was not defined at all.)
3533 11. A "Completed" log line is now written for messages that are removed from
3534 the spool by the -Mrm option.
3536 12. $host_address is now set to the target address during the checking of
3537 ignore_target_hosts.
3539 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3540 being passed; this would have caused $sender_host_name to have been used if
3541 matching the list had actually called for a host name (not very likely,
3542 since this list is usually IP addresses). A host name is now passed as
3545 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3546 code that specifies a non-restarting handler (typically sigaction() in
3547 modern systems) in an attempt to fix a rare and obscure crash bug.
3549 15. Narrowed the window for a race in the daemon that could cause it to ignore
3550 SIGCHLD signals. This is not a major problem, because they are used only to
3551 wake it up if nothing else does.
3553 16. A malformed maildirsize file could cause Exim to calculate negative values
3554 for the mailbox size or file count. Odd effects could occur as a result.
3555 The maildirsize information is now recalculated if the size or filecount
3558 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3559 support for a long time. Removed HAVE_SYS_VFS_H.
3561 18. Updated exipick to current release from John Jetmore.
3563 19. Allow an empty sender to be matched against a lookup in an address list.
3564 Previously the only cases considered were a regular expression, or an
3567 20. Exim went into a mad DNS lookup loop when doing a callout where the
3568 host was specified on the transport, if the DNS lookup yielded more than
3571 21. The RFC2047 encoding function was originally intended for short strings
3572 such as real names; it was not keeping to the 75-character limit for
3573 encoded words that the RFC imposes. It now respects the limit, and
3574 generates multiple encoded words if necessary. To be on the safe side, I
3575 have increased the buffer size for the ${rfc2047: expansion operator from
3578 22. Failure to deliver a bounce message always caused it to be frozen, even if
3579 there was an errors_to setting on the router. The errors_to setting is now
3582 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3583 canonical form (fully expanded) before being placed in
3584 $sender_host_address.
3586 24. Updated eximstats to version 1.33
3588 25. Include certificate and key file names in error message when GnuTLS fails
3589 to set them up, because the GnuTLS error message doesn't include the name
3590 of the failing file when there is a problem reading it.
3592 26. Expand error message when OpenSSL has problems setting up cert/key files.
3595 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3596 (this can affect the format of dates).
3598 28. exim_tidydb, when checking for the continued existence of a message for
3599 which it has found a message-specific retry record, was not finding
3600 messages that were in split spool directories. Consequently, it was
3601 deleting retry records that should have stayed in existence.
3603 29. eximstats updated to version 1.35
3604 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3605 1.35 - bugfix such that pie charts by volume are generated correctly
3607 30. The SPA authentication driver was not abandoning authentication and moving
3608 on to the next authenticator when an expansion was forced to fail,
3609 contradicting the general specification for all authenticators. Instead it
3610 was generating a temporary error. It now behaves as specified.
3612 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3613 (the order specifies the preference for clients). The order is now AES256,
3614 AES128, 3DES, ARCFOUR128.
3616 31. Small patch to Sieve code - explicitly set From: when generating an
3619 32. Exim crashed if a remote delivery caused a very long error message to be
3620 recorded - for instance if somebody sent an entire SpamAssassin report back
3621 as a large number of 550 error lines. This bug was coincidentally fixed by
3622 increasing the size of one of Exim's internal buffers (big_buffer) that
3623 happened as part of the Exiscan merge. However, to be on the safe side, I
3624 have made the code more robust (and fixed the comments that describe what
3627 33. Some experimental protocols are using DNS PTR records for new purposes. The
3628 keys for these records are domain names, not reversed IP addresses. The
3629 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3630 leaves it alone. Component reversal etc. now happens only for IP addresses.
3633 34. The host_aton() function is supposed to be passed a string that is known
3634 to be a valid IP address. However, in the case of IPv6 addresses, it was
3635 not checking this. This is a hostage to fortune. Exim now panics and dies
3636 if the condition is not met. A case was found where this could be provoked
3637 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3638 components; fortuitously, this particular loophole had already been fixed
3639 by change 4.50/55 or 4.44/33 above.
3641 If there are any other similar loopholes, the new check in host_aton()
3642 itself should stop them being exploited. The report I received stated that
3643 data on the command line could provoke the exploit when Exim was running as
3644 exim, but did not say which command line option was involved. All I could
3645 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3646 running as the user.
3649 35. There was a buffer overflow vulnerability in the SPA authentication code
3650 (which came originally from the Samba project). I have added a test to the
3651 spa_base64_to_bits() function which I hope fixes it.
3654 36. The daemon start-up calls getloadavg() while still root for those OS that
3655 need the first call to be done as root, but it missed one case: when
3656 deliver_queue_load_max is set with deliver_drop_privilege. This is
3657 necessary for the benefit of the queue runner, because there is no re-exec
3658 when deliver_drop_privilege is set.
3660 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3661 was in use, was not putting the data itself into the right store pool;
3662 consequently, it could be overwritten for a subsequent message in the same
3663 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3666 38. Sometimes the final signoff response after QUIT could fail to get
3667 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3668 < 0 before doing a fflush(). This bug looks as though it goes back to the
3669 introduction of TLS in release 3.20, but "sometimes" must have been rare
3670 because the tests only now provoked it.
3676 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3677 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3678 Now it returns an integer. A number of calls were still expecting a T/F
3679 return. Fortuitously, in all cases, the tests worked in OK situations,
3680 which is the norm. However, things would have gone wrong on any write
3681 failures on the smtp file descriptor. This function is used when sending
3682 messages over SMTP and also when doing verify callouts.
3684 2. When Exim is called to do synchronous delivery of a locally submitted
3685 message (the -odf or -odi options), it no longer closes stderr before doing
3688 3. Implemented the mua_wrapper option.
3690 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3692 5. Implemented the functions header_remove(), header_testname(),
3693 header_add_at_position(), and receive_remove_recipient(), and exported them
3696 6. If an ACL "warn" statement specified the addition of headers, Exim already
3697 inserted X-ACL-Warn: at the start if there was no header name. However, it
3698 was not making this test for the second and subsequent header lines if
3699 there were newlines in the string. This meant that an invalid header could
3700 be inserted if Exim was badly configured.
3702 7. Allow an ACL "warn" statement to add header lines at the start or after all
3703 the Received: headers, as well as at the end.
3705 8. Added the rcpt_4xx retry error code.
3707 9. Added postmaster_mailfrom=xxx to callout verification option.
3709 10. Added mailfrom=xxxx to the callout verification option, for verify=
3712 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3713 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3715 12. Inserted some casts to stop certain compilers warning when using pointer
3716 differences as field lengths or precisions in printf-type calls (mostly
3717 affecting debugging statements).
3719 13. Added optional readline() support for -be (dynamically loaded).
3721 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3722 same clock tick as a message's arrival, so that its received time was the
3723 same as the "first fail" time on the retry record, and that message
3724 remained on the queue past the ultimate address timeout, every queue runner
3725 would try a delivery (because it was past the ultimate address timeout) but
3726 after another failure, the ultimate address timeout, which should have then
3727 bounced the address, did not kick in. This was a "< instead of <=" error;
3728 in most cases the first failure would have been in the next clock tick
3729 after the received time, and all would be well.
3731 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3732 being recognized when the domain list was tested by the match_domain
3733 condition in an expansion string.
3735 16. Added the ${str2b64: operator.
3737 17. Exim was always calling setrlimit() to set a large limit for the number of
3738 processes, without checking whether the existing limit was already
3739 adequate. (It did check for the limit on file descriptors.) Furthermore,
3740 errors from getrlimit() and setrlimit() were being ignored. Now they are
3741 logged to the main and panic logs, but Exim does carry on, to try to do its
3742 job under whatever limits there are.
3744 18. Imported PCRE 5.0.
3746 19. Trivial typo in log message " temporarily refused connection" (the leading
3749 20. If the log selector return_path_on_delivery was set and an address was
3750 redirected to /dev/null, the delivery process crashed because it assumed
3751 that a return path would always be set for a "successful" delivery. In this
3752 case, the whole delivery is bypassed as an optimization, and therefore no
3755 21. Internal re-arrangement: the function for sending a challenge and reading
3756 a response while authentication was assuming a zero-terminated challenge
3757 string. It's now changed to take a pointer and a length, to allow for
3758 binary data in such strings.
3760 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3762 23. Exim was not respecting finduser_retries when seeking the login of the
3763 uid under which it was called; it was always trying 10 times. (The default
3764 setting of finduser_retries is zero.) Also, it was sleeping after the final
3765 failure, which is pointless.
3767 24. Implemented tls_on_connect_ports.
3769 25. Implemented acl_smtp_predata.
3771 26. If the domain in control=submission is set empty, Exim assumes that the
3772 authenticated id is a complete email address when it generates From: or
3773 Sender: header lines.
3775 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3776 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3777 chown and chgrp in /bin and hostname in /usr/bin.
3779 28. Exim was keeping the "process log" file open after each use, just as it
3780 does for the main log. This opens the possibility of it remaining open for
3781 long periods when the USR1 signal hits a daemon. Occasional processlog
3782 errors were reported, that could have been caused by this. Anyway, it seems
3783 much more sensible not to leave this file open at all, so that is what now
3786 29. The long-running daemon process does not normally write to the log once it
3787 has entered its main loop, and it closes the log before doing so. This is
3788 so that log files can straightforwardly be renamed and moved. However,
3789 there are a couple of unusual error situations where the daemon does write
3790 log entries, and I had neglected to close the log afterwards.
3792 30. The text of an SMTP error response that was received during a remote
3793 delivery was being truncated at 512 bytes. This is too short for some of
3794 the long messages that one sometimes sees. I've increased the limit to
3797 31. It is now possible to make retry rules that apply only when a message has a
3798 specific sender, in particular, an empty sender.
3800 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3801 it possible to be selective about when SMTP synchronization is enforced.
3803 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3805 32. Implemented hosts_connection_nolog.
3807 33. Added an ACL for QUIT.
3809 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3812 35. Added mailbox_size and mailbox_filecount to appendfile.
3814 36. Added control = no_multiline_responses to ACLs.
3816 37. There was a bug in the logic of the code that waits for the clock to tick
3817 in the case where the clock went backwards by a substantial amount such
3818 that the microsecond fraction of "now" was more than the microsecond
3819 fraction of "then" (but the whole seconds number was less).
3821 38. Added support for the libradius Radius client library this is found on
3822 FreeBSD (previously only the radiusclient library was supported).
3828 1. When certain lookups returned multiple values in the form name=value, the
3829 quoting of the values was not always being done properly. Specifically:
3830 (a) If the value started with a double quote, but contained no whitespace,
3832 (b) If the value contained whitespace other than a space character (i.e.
3833 tabs or newlines or carriage returns) it was not quoted.
3834 This fix has been applied to the mysql and pgsql lookups by writing a
3835 separate quoting function and calling it from the lookup code. The fix
3836 should probably also be applied to nisplus, ibase and oracle lookups, but
3837 since I cannot test any of those, I have not disturbed their existing code.
3839 2. A hit in the callout cache for a specific address caused a log line with no
3840 reason for rejecting RCPT. Now it says "Previous (cached) callout
3841 verification failure".
3843 3. There was an off-by-one bug in the queryprogram router. An over-long
3844 return line was truncated at 256 instead of 255 characters, thereby
3845 overflowing its buffer with the terminating zero. As well as fixing this, I
3846 have increased the buffer size to 1024 (and made a note to document this).
3848 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3849 when Exim is waiting for an SMTP response from a remote server, Exim
3850 restarts its select() call on the socket, thereby resetting its timeout.
3851 This is not a problem when such interrupts are rare. Somebody set up a cron
3852 job to run exiwhat every 2 minutes, which is less than the normal select()
3853 timeout (5 or 10 minutes). This meant that the select() timeout never
3854 kicked in because it was always reset. I have fixed this by comparing the
3855 time when an interrupt arrives with the time at the start of the first call
3856 to select(). If more time than the timeout has elapsed, the interrupt is
3857 treated as a timeout.
3859 5. Some internal re-factoring in preparation for the addition of Sieve
3860 extensions (by MH). In particular, the "personal" test is moved to a
3861 separate function, and given an option for scanning Cc: and Bcc: (which is
3862 not set for Exim filters).
3864 6. When Exim created an email address using the login of the caller as the
3865 local part (e.g. when creating a From: or Sender: header line), it was not
3866 quoting the local part when it contained special characters such as @.
3868 7. Installed new OpenBSD configuration files.
3870 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3871 try to make them clearer.
3873 9. Callout options, other than the timeout value, were being ignored when
3874 verifying sender addresses in header lines. For example, when using
3876 verify = header_sender/callout=no_cache
3878 the cache was (incorrectly) being used.
3880 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3881 only the Cygwin environment.
3883 11. When return_path_on_delivery was set as a log selector, if different remote
3884 addresses in the same message used different return paths and parallel
3885 remote delivery occurred, the wrong values would sometimes be logged.
3886 (Whenever a remote delivery process finished, the return path value from
3887 the most recently started remote delivery process was logged.)
3889 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3890 lines when AUTH and/or TLS are in use. This is the "received protocol"
3891 field. Exim used to use "asmtp" for authenticated SMTP, without any
3892 indication (in the protocol name) for TLS use. Now it follows the RFC and
3893 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3894 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3895 names appear in log lines as well as in Received: header lines.
3897 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3898 extensions, and comparison tests, and to fix some bugs.
3900 14. Changes to the "personal" filter test:
3902 (1) The test was buggy in that it was just doing the equivalent of
3903 "contains" tests on header lines. For example, if a user's address was
3904 anne@some.where, the "personal" test would incorrectly be true for
3906 To: susanne@some.where
3908 This test is now done by extracting each address from the header in turn,
3909 and checking the entire address. Other tests that are part of "personal"
3910 are now done using regular expressions (for example, to check local parts
3911 of addresses in From: header lines).
3913 (2) The list of non-personal local parts in From: addresses has been
3914 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3915 taken from the Sieve specification recommendations.
3917 (3) If the message contains any header line starting with "List-" it is
3918 treated as non-personal.
3920 (4) The test for "circular" in the Subject: header line has been removed
3921 because it now seems ill-conceived.
3923 15. Minor typos in src/EDITME comments corrected.
3925 16. Installed latest exipick from John Jetmore.
3927 17. If headers_add on a router specified a text string that was too long for
3928 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3929 of string_sprintf() is now avoided.
3931 18. $message_body_size was not set (it was always zero) when running the DATA
3932 ACL and the local_scan() function.
3934 19. For the "mail" command in an Exim filter, no default was being set for
3935 the once_repeat time, causing a random time value to be used if "once" was
3936 specified. (If the value happened to be <= 0, no repeat happened.) The
3937 default is now 0s, meaning "never repeat". The "vacation" command was OK
3938 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3939 (I found it when inspecting the code).
3941 20. There is now an overall timeout for performing a callout verification. It
3942 defaults to 4 times the callout timeout, which applies to individual SMTP
3943 commands during the callout. The overall timeout applies when there is more
3944 than one host that can be tried. The timeout is checked before trying the
3945 next host. This prevents very long delays if there are a large number of
3946 hosts and all are timing out (e.g. when the network connections are timing
3947 out). The value of the overall timeout can be changed by specifying an
3948 additional sub-option for "callout", called "maxwait". For example:
3950 verify = sender/callout=5s,maxwait=20s
3952 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3953 to the end before writing, but this should make it even safer).
3955 22. Exim was forgetting that it had advertised PIPELINING for the second and
3956 subsequent messages on an SMTP connection. It was also not resetting its
3957 memory on STARTTLS and an internal HELO.
3959 23. When Exim logs an SMTP synchronization error within a session, it now
3960 records whether PIPELINING has been advertised or not.
3962 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3963 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3964 rather than long int.
3966 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3968 26. Added the never_mail option to autoreply.
3974 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3975 crash if the getsockname() call failed; this can happen if a connection is
3976 closed very soon after it is established. The problem was simply in the
3977 order in which certain operations were done, causing Exim to try to write
3978 to the SMTP stream before it had set up the file descriptor. The bug has
3979 been fixed by making things happen in the correct order.
3985 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3986 before the connection was closed, thus losing the rejection response.
3988 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3989 some early Solaris releases, but causes trouble in current releases where
3990 socklen_t is defined.
3992 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3995 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3998 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3999 that releasing the top of it at the end releases what was used for sub-
4000 expansions (unless the block got too big). However, discard this block if
4001 the first thing is a variable or header, so that we can use its block when
4002 it is dynamic (useful for very large $message_headers, for example).
4004 6. Lookups now cache *every* query, not just the most recent. A new, separate
4005 store pool is used for this. It can be recovered when all lookup caches are
4006 flushed. Lookups now release memory at the end of their result strings.
4007 This has involved some general refactoring of the lookup sources.
4009 7. Some code has been added to the store_xxx() functions to reduce the amount
4010 of flapping under certain conditions.
4012 8. log_incoming_interface used to affect only the <= reception log lines. Now
4013 it causes the local interface and port to be added to several more SMTP log
4014 lines, for example "SMTP connection from", and rejection lines.
4016 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4018 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4020 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4021 could be overwritten at the end of the current message (or the start of a
4022 new message if it was set in a HELO ACL). The value is now preserved for
4023 the duration of the SMTP connection.
4025 12. If a transport had a headers_rewrite setting, and a matching header line
4026 contained an unqualified address, that address was qualified, even if it
4027 did not match any rewriting rules. The underlying bug was that the values
4028 of the flags that permit the existence of unqualified sender and recipient
4029 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4030 non-local messages, and by -bnq for local messages) were not being
4031 preserved with the message after it was received.
4033 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4034 "next input=" as part of the text comprising the host identity instead of
4035 the correct text. The code was using the same buffer for two different
4036 strings. However, depending on which order the printing function evaluated
4037 its arguments, the bug did not always show up. Under Linux, for example, my
4038 test suite worked just fine.
4040 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4041 doesn't work with some older versions of Perl. It has been changed to "my",
4042 which in any case is probably the better facility to use.
4044 15. A really picky compiler found some instances of statements for creating
4045 error messages that either had too many or two few arguments for the format
4048 16. The size of the buffer for calls to the DNS resolver has been increased
4049 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4050 for addresses that have a lot of PTR records. This alleviates a problem; it
4051 does not fully solve it.
4053 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4054 buffer now truncates the list and logs the incident, which is the same
4055 action as happens when Exim is looking up a host name and its aliases.
4056 Previously in this situation something unpredictable would happen;
4057 sometimes it was "internal error: store_reset failed".
4059 18. If a server dropped the connection unexpectedly when an Exim client was
4060 using GnuTLS and trying to read a response, the client delivery process
4061 crashed while trying to generate an error log message.
4063 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4064 string, for example:
4066 warn message = H1: something\nH2: something
4068 the text was added as a single header line from Exim's point of view
4069 though it ended up OK in the delivered message. However, searching for the
4070 second and subsequent header lines using $h_h2: did not work. This has been
4071 fixed. Similarly, if a system filter added multiple headers in this way,
4072 the routers could not see them.
4074 20. Expanded the error message when iplsearch is called with an invalid key to
4075 suggest using net-iplsearch in a host list.
4077 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4078 ACLs are no longer actually imposed (and a message to that effect is
4081 22. If a "gecos" field in a passwd entry contained escaped characters, in
4082 particular, if it contained a \" sequence, Exim got it wrong when building
4083 a From: or a Sender: header from that name. A second bug also caused
4084 incorrect handling when an unquoted " was present following a character
4085 that needed quoting.
4087 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4088 was not being matched caselessly.
4090 24. Arranged for all hyphens in the exim.8 source to be escaped with
4093 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4094 properly. Recipient callout cache records were still being keyed to include
4095 the sender, even when use_sender was set false. This led to far more
4096 callouts that were necessary. The sender is no longer included in the key
4097 when use_sender is false.
4099 26. Added "control = submission" modifier to ACLs.
4101 27. Added the ${base62d: operator to decode base 62 numbers.
4103 28. dnsdb lookups can now access SRV records.
4105 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4106 the configuration file.
4108 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4109 (-v) mode. This makes the output for a verbose queue run more intelligible.
4111 31. Added a use_postmaster feature to recipient callouts.
4113 32. Added the $body_zerocount variable, containing the number of binary zero
4114 bytes in the message body.
4116 33. The time of last modification of the "new" subdirectory is now used as the
4117 "mailbox time last read" when there is a quota error for a maildir
4120 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4122 35. Added +ignore_unknown as a special item in host lists.
4124 36. Code for decoding IPv6 addresses in host lists is now included, even if
4125 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4126 address was recognized as an IP address, but was then not correctly decoded
4127 into binary, causing unexpected and incorrect effects when compared with
4134 1. Very minor rewording of debugging text in manualroute to say "list of
4135 hosts" instead of "hostlist".
4137 2. If verify=header_syntax was set, and a header line with an unqualified
4138 address (no domain) and a large number of spaces between the end of the
4139 name and the colon was received, the reception process suffered a buffer
4140 overflow, and (when I tested it) crashed. This was caused by some obsolete
4141 code that should have been removed. The fix is to remove it!
4143 3. When running in the test harness, delay a bit after writing a bounce
4144 message to get a bit more predictability in the log output.
4146 4. Added a call to search_tidyup() just before forking a reception process. In
4147 theory, someone could use a lookup in the expansion of smtp_accept_max_
4148 per_host which, without the tidyup, could leave open a database connection.
4150 5. Added the variables $recipient_data and $sender_data which get set from a
4151 lookup success in an ACL "recipients" or "senders" condition, or a router
4152 "senders" option, similar to $domain_data and $local_part_data.
4154 6. Moved the writing of debug_print from before to after the "senders" test
4157 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4158 problems for message scanning, either using a data ACL, or using
4159 local_scan() because the Received: header was not generated till after they
4160 were called (in order to set the time as the time of reception completion).
4161 I have revised the way this works. The header is now generated after the
4162 body is received, but before the ACL or local_scan() are called. After they
4163 are run, the timestamp in the header is updated.
4169 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4170 before starting a queue runner without re-exec. This happened only when
4171 deliver_drop_privilege was set or when the Exim user was set to root. The
4172 effect of the bug was that timeouts during subsequent deliveries caused
4173 crashes instead of being properly handled. The handler is now left at its
4174 default (and expected) setting.
4176 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4177 message, again when deliver_drop_privilege is set or Exim is run as root.
4178 The bug described in (1) was not present in this case, but the tidying up
4179 of the other signals was missing. I have made the two cases consistent.
4181 3. The ignore_target_hosts setting on a manualroute router was being ignored
4182 for hosts that were looked up using the /MX notation.
4184 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4187 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4188 operated on the sender address. After changing the $sender_address to <>
4189 for the sender address verify, Exim was re-instated it as the original
4190 (before rewriting) address, but remembering that it had rewritten it, so it
4191 wasn't rewriting it again. This bug also had the effect of breaking the
4192 sender address verification caching when the sender address was rewritten.
4194 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4195 This has been changed so that if the ip literal address matches
4196 ignore_target_hosts, the router declines.
4198 7. Added expansion conditions match_domain, match_address, and match_local_
4199 part (NOT match_host).
4201 8. The placeholder for the Received: header didn't have a length field set.
4203 9. Added code to Exim itself and to exim_lock to test for a specific race
4204 condition that could lead to file corruption when using MBX delivery. The
4205 issue is with the lockfile that is created in /tmp. If this file is removed
4206 after a process has opened it but before that process has acquired a lock,
4207 there is the potential for a second process to recreate the file and also
4208 acquire a lock. This could lead to two Exim processes writing to the file
4209 at the same time. The added code performs the same test as UW imapd; it
4210 checks after acquiring the lock that its file descriptor still refers to
4211 the same named file.
4213 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4214 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4217 11. Added the smtp_active_hostname option. If used, this will typically be made
4218 to depend on the incoming interface address. Because $interface_address is
4219 not set up until the daemon has forked a reception process, error responses
4220 that can happen earlier (such as "too many connections") no longer contain
4223 12. If an expansion in a condition on a "warn" statement fails because a lookup
4224 defers, the "warn" statement is abandoned, and the next ACL statement is
4225 processed. Previously this caused the whole ACL to be aborted.
4227 13. Added the iplsearch lookup type.
4229 14. Added ident_timeout as a log selector.
4231 15. Added tls_certificate_verified as a log selector.
4233 16. Added a global option tls_require_ciphers (compare the smtp transport
4234 option of the same name). This controls incoming TLS connections.
4236 17. I finally figured out how to make tls_require_ciphers do a similar thing
4237 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4238 before starting the TLS session.
4240 18. Tabs are now shown as \t in -bP output.
4242 19. If the log selector return_path_on_delivery was set, Exim crashed when
4243 bouncing a message because it had too many Received: header lines.
4245 20. If two routers both had headers_remove settings, and the first one included
4246 a superfluous trailing colon, the final name in the first list and the
4247 first name in the second list were incorrectly joined into one item (with a
4248 colon in the middle).
4254 1. Added -C and -D options to the exinext utility, mainly to make it easier
4255 to include in the automated testing, but these could be helpful when
4256 multiple configurations are in use.
4258 2. The exinext utility was not formatting the output nicely when there was
4259 an alternate port involved in the retry record key, nor when there was a
4260 message id as well (for retries that were specific to a specific message
4261 and a specific host). It was also confused by IPv6 addresses, because of
4262 the additional colons they contain. I have fixed the IPv4 problem, and
4263 patched it up to do a reasonable job for IPv6.
4265 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4266 delivery, the log line now contains "pipelined" if PIPELINING was used.
4268 4. An SMTP transport process used to panic and die if the bind() call to set
4269 an explicit outgoing interface failed. This has been changed; it is now
4270 treated in the same way as a connect() failure.
4272 5. A reference to $sender_host_name in the part of a conditional expansion
4273 that was being skipped was still causing a DNS lookup. This no longer
4276 6. The def: expansion condition was not recognizing references to header lines
4277 that used bh_ and bheader_.
4279 7. Added the _cache feature to named lists.
4281 8. The code for checking quota_filecount in the appendfile transport was
4282 allowing one more file than it should have been.
4284 9. For compatibility with Sendmail, the command line option
4292 and sets the incoming protocol and host name (for trusted callers). The
4293 host name and its colon can be omitted when only the protocol is to be set.
4294 Note the Exim already has two private options, -pd and -ps, that refer to
4295 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4296 "s", but I don't think that's a major issue.
4298 10. A number of refactoring changes to the code, none of which should affect
4301 (a) The number of logging options was getting close to filling up the
4302 32-bit word that was used as a bit map. I have split them into two classes:
4303 those that are passed in the argument to log_write(), and those that are
4304 only ever tested independently outside of that function. These are now in
4305 separate 32-bit words, so there is plenty of room for expansion again.
4306 There is no change in the user interface or the logging behaviour.
4308 (b) When building, for example, log lines, the code previously used a
4309 macro that called string_cat() twice, in order to add two strings. This is
4310 not really sufficiently general. Furthermore, there was one instance where
4311 it was actually wrong because one of the argument was used twice, and in
4312 one call a function was used. (As it happened, calling the function twice
4313 did not affect the overall behaviour.) The macro has been replaced by a
4314 function that can join an arbitrary number of extra strings onto a growing
4317 (c) The code for expansion conditions now uses a table and a binary chop
4318 instead of a serial search (which was left over from when there were very
4319 few conditions). Also, it now recognizes conditions like "pam" even when
4320 the relevant support is not compiled in: a suitably worded error message is
4321 given if an attempt is made to use such a condition.
4323 11. Added ${time_interval:xxxxx}.
4325 12. A bug was causing one of the ddress fields not to be passed back correctly
4326 from remote delivery subprocesses. The field in question was not being
4327 subsequently used, so this caused to problems in practice.
4329 13. Added new log selectors queue_time and deliver_time.
4331 14. Might have fixed a bug in maildirsizefile handling that threw up
4332 "unexpected character" debug warnings, and recalculated the data
4333 unnecessarily. In any case, I expanded the warning message to give more
4336 15. Added the message "Restricted characters in address" to the statements in
4337 the default ACL that block characters like @ and % in local parts.
4339 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4340 Three changes have been made:
4342 (a) There was a serious bug; a negative response to MAIL caused the whole
4343 recipient domain to be cached as invalid, thereby blocking all messages
4344 to all local parts at the same domain, from all senders. This bug has
4345 been fixed. The domain is no longer cached after a negative response to
4346 MAIL if the sender used is not empty.
4348 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4351 (c) A new callout option, "use_sender" has been added for people who want
4352 the modified behaviour.
4358 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4361 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4362 indeed breaks things for older releases.
4364 3. Added additional logging to the case where there is a problem reading data
4365 from a filter that is running in a subprocess using a pipe, in order to
4366 try to track down a specific problem.
4368 4. Testing facility fudge: when running in the test harness and attempting
4369 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4370 getting "No route to host". Convert this to a timeout.
4372 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4375 6. Some OS don't have socklen_t but use size_t instead. This affects the
4376 fifth argument of getsockopt() amongst other things. This is now
4377 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4378 can be set for individual OS. I have set it for SunOS5, OSF1, and
4379 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4380 some earlier ones do not.
4382 7. Change 4.30/15 was not doing the test caselessly.
4384 8. The standard form for an IPv6 address literal was being rejected by address
4385 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4386 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4387 this, as well as the form without the "IPv6" on the front (but only when
4388 address literals are enabled, of course).
4390 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4392 10. Exim crashed if a message with an empty sender address specified by -f
4393 encountered a router with an errors_to setting. This could be provoked only
4394 by a command such as
4398 where an empty string was supplied; "<>" did not hit this bug.
4400 11. Installed PCRE release 4.5.
4402 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4403 remained set. It is now erased.
4405 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4406 times from message ids (which are base 36 rather than the normal 62).
4408 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4409 were being counted as actual protocol errors, and logged if the log
4410 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4411 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4412 and DATA following a set of rejected RCPTs do not count as protocol errors.
4413 In other words, Exim assumes they were pipelined, though this may not
4414 actually be the case. Of course, in all cases the client gets an
4415 appropriate error code.
4417 15. If a lookup fails in an ACL condition, a message about the failure may
4418 be available; it is used if testing the ACL cannot continue, because most
4419 such messages specify what the cause of the deferral is. However, some
4420 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4421 that caused an old message to be retained and used if a later statement
4422 caused a defer, replacing the real cause of the deferral.
4424 16. If an IP address had so many PTR records that the DNS lookup buffer
4425 was not large enough to hold them, Exim could crash while trying to process
4426 the truncated data. It now detects and logs this case.
4428 17. Further to 4.21/58, another change has been made: if (and only if) the
4429 first line of a message (the first header line) ends with CRLF, a bare LF
4430 in a subsequent header line has a space inserted after it, so as not to
4431 terminate the header.
4433 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4434 unnecessarily, used atoi() instead of strtol(), and didn't check the
4435 termination when getting file sizes from file names by regex.
4437 19. Completely re-implemented the support for maildirsize files, in the light
4438 of a number of problems with the previous contributed implementation
4439 (4.30/29). In particular:
4441 . If the quota is zero, the maildirsize file is maintained, but no quota is
4444 . If the maildir directory does not exist, it is created before any attempt
4445 to write a maildirsize file.
4447 . The quota value in the file is just a cache; if the quota is changed in
4448 the transport, the new value overrides.
4450 . A regular expression is available for excluding directories from the
4453 20. The autoreply transport checks the characters in options that define the
4454 message's headers; it allows continued headers, but it was checking with
4455 isspace() after an embedded newline instead of explicitly looking for a
4458 21. If all the "regular" hosts to which an address was routed had passed their
4459 expiry times, and had not reached their retry times, the address was
4460 bounced, even if fallback hosts were defined. Now Exim should go on to try
4463 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4464 equivalent code in the SMTP transport. Some hosts send humungous responses
4465 to HELO/EHLO, more than 1024 it seems.
4467 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4468 gives compiler warnings in some environments. I've now done it "properly",
4471 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4472 (because of problems with the built-in one) was declared to return uschar *
4473 instead of char *, causing compiler failure.
4475 25. Fixed a file descriptor leak when processing alias/forward files.
4477 26. Fixed a minor format string issue in dbfn.c.
4479 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4481 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4482 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4485 29. When a sender address is verified, it is cached, to save repeating the test
4486 when there is more than one recipient in a message. However, when the
4487 verification involves a callout, it is possible for different callout
4488 options to be set for different recipients. It is too complicated to keep
4489 track of this in the cache, so now Exim always runs a verification when a
4490 callout is required, relying on the callout cache for the optimization.
4491 The overhead is duplication of the address routing, but this should not be
4494 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4495 to be verified with callout=postmaster, and the main callout worked but the
4496 postmaster check failed, the verification correctly failed. However, if a
4497 subsequent RCPT command asked for sender verification *without* the
4498 postmaster check, incorrect caching caused this verification also to fail,
4501 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4502 it was not caching the DNS options (qualify_single, search_parents) that
4503 were used when the lookup failed. A subsequent lookup with different
4504 options therefore always gave the same answer, though there were cases
4505 where it should not have. (Example: a "domains = !$mx_any" option on a
4506 dnslookup router: the "domains" option is always processed without any
4507 widening, but the router might have qualify_single set.) Now Exim uses the
4508 cached value only when the same options are set.
4510 32. Added John Jetmore's "exipick" utility to the distribution.
4512 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4513 than a timeout (e.g. a certificate is required, and is not provided), an
4514 Exim server now closes the connection immediately. Previously it waited for
4515 the client to close - but if the client is SSL, it seems that they each
4516 wait for each other, leading to a delay before one of them times out.
4518 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4519 maintained 0.8.x compatibility because I don't think many are using it, and
4520 it is clearly obsolete.
4522 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4525 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4526 client certificate was expired. A simple patch fixes this, though I don't
4527 understand the full logic of why the verify callback is called multiple
4530 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4531 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4532 which causes problems with some clients (such as the Certicom SSL Plus
4533 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4534 disables the coutermeasure allowing Eudora to connect."
4536 38. Exim was not checking that a write() to a log file succeeded. This could
4537 lead to Bad Things if a log got too big, in particular if it hit a file
4538 size limit. Exim now panics and dies if it cannot write to a log file, just
4539 as it does if it cannot open a log file.
4541 39. Modified OS/Makefile-Linux so that it now contains
4543 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4545 The two -D definitions ensure that Exim is compiled with large file
4546 support, which makes it possible to handle log files that are bigger than
4549 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4550 instance) a domain was checked against a named list that involved a lookup,
4551 causing $domain_data to be set, then another domain was checked against the
4552 same list, then the first domain was re-checked, the value of $domain_data
4553 after the final check could be wrong. In particular, if the second check
4554 failed, it could be set empty. This bug probably also applied to
4557 41. The strip_trailing_dot option was not being applied to the address given
4558 with the -f command-line option.
4560 42. The code for reading a message's header from the spool was incrementing
4561 $received_count, but never initializing it. This meant that the value was
4562 incorrect (doubled) while delivering a message in the same process in which
4563 it was received. In the most common configuration of Exim, this never
4564 happens - a fresh exec is done - but it can happen when
4565 deliver_drop_privilege is set.
4567 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4568 it now includes up to 150 characters of the unexpected data in the log
4571 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4572 and building data strings. The size of both of these buffers was 10 000
4573 bytes - far larger than anybody would *ever* want, thought I. Needless to
4574 say, somebody hit the limit. I have increased the maximum line length to
4575 20 000 and the maximum data length of concatenated lines to 100 000. I have
4576 also fixed two bugs, because there was no checking on these buffers. Tsk,
4577 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4578 buffer is too small.
4580 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4581 lsearch lookups. Now it does.
4583 46. When parsing a route_list item in a manualroute router, a fixed-length
4584 buffer was used for the list of hosts. I made this 1024 bytes long,
4585 thinking that nobody would ever have a list of hosts that long. Wrong.
4586 Somebody had a whole pile of complicated expansion conditions, and the
4587 string was silently truncated, leading to an expansion error. It turns out
4588 that it is easier to change to an unlimited length (owing to other changes
4589 that have happened since this code was originally written) than to build
4590 structure for giving a limitation error. The length of the item that
4591 expands into the list of hosts is now unlimited.
4593 47. The lsearch lookup could not handle data where the length of text line was
4594 more than 4095 characters. Such lines were truncated, leading to shortened
4595 data being returned. It should now handle lines of any length.
4597 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4598 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4601 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4602 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4603 no longer happens in comment lines. A list of replacements is now placed
4604 at the head of all of the source files, except those whose only change is
4605 to replace PERL_COMMAND in the very first #! line.
4607 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4608 messages on the queue, with a bottom-up merge sort, using code contributed
4609 by Michael Haardt. This should make operations like -bp somewhat faster on
4610 large queues. It won't affect queue runners, except when queue_run_in_order
4613 51. Installed eximstats 1.31 in the distribution.
4615 52. Added support for SRV lookups to the dnslookup router.
4617 53. If an ACL referred to $message_body or $message_body_end, the value was not
4618 reset for any messages that followed in the same SMTP session.
4620 54. The store-handling optimization for building very long strings was not
4621 differentiating between the different store pools. I don't think this
4622 actually made any difference in practice, but I've tidied it.
4624 55. While running the routers to verify a sender address, $sender_address
4625 was still set to the sender address. This is wrong, because when routing to
4626 send a bounce to the sender, it would be empty. Therefore, I have changed
4627 it so that, while verifying a sender address, $sender_address is set to <>.
4628 (There is no change to what happens when verifying a recipient address.)
4630 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4631 target A or AAAA records (if not already returned) without resetting the
4632 qualify_single or search_parents options of the DNS resolver. These are
4633 inappropriate in this case because the targets of MX and SRV records must
4634 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4635 target that, when qualified, matched something in the local domain. These
4636 two options are now turned off when doing these lookups.
4638 57. It seems that at least some releases of Reiserfs (which does not have the
4639 concept of a fixed number of inodes) returns zero and not -1 for the
4640 number of available inodes. This interacted badly with check_spool_inodes,
4641 which assumed that -1 was the "no such thing" setting. What I have done is
4642 to check that the total number of inodes is greater than zero before doing
4643 the test of how many are available.
4645 58. When a "warn" ACL statement has a log_message modifier, the message is
4646 remembered, and not repeated. This is to avoid a lot of repetition when a
4647 message has many recipients that cause the same warning to be written.
4648 Howewer, Exim was preserving the list of already written lines for an
4649 entire SMTP session, which doesn't seem right. The memory is now reset if a
4650 new message is started.
4652 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4653 debugging output unless log_rewrite was also set.
4655 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4656 of (int)(handle) when we know that handle contains (void *)(-1).
4658 61. The Exim daemon panic-logs an error return when it closes the incoming
4659 connection. However "connection reset by peer" seems to be common, and
4660 isn't really an error worthy of noting specially, so that particular error
4663 62. When Exim is trying to find all the local interfaces, it used to panic and
4664 die if the ioctl to get the interface flags failed. However, it seems that
4665 on at least one OS (Solaris 9) it is possible to have an interface that is
4666 included in the list of interfaces, but for which you get a failure error
4667 for this call. This happens when the interface is not "plumbed" into a
4668 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4669 failure of the "get flags" call assumes that the interface is down.
4671 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4672 makes life easier for people who are doing arithmetic on fields extracted
4673 from dates, where you often get leading zeros that should not be
4674 interpreted as octal.
4676 64. Added qualify_domain to the redirect router, to override the global
4679 65. If a pathologically long header line contained very many addresses (the
4680 report of this problem mentioned 10 000) and each of them was rewritten,
4681 Exim could use up a very large amount of memory. (It kept on making new
4682 copies of the header line as it rewrote, and never released the old ones.)
4683 At the expense of a bit more processing, the header rewriting function has
4684 been changed so that it no longer eats memory in this way.
4686 66. The generation of the Received: header has been moved from the time that a
4687 message starts to be received, to the time that it finishes. The timestamp
4688 in the Received: header should now be very close to that of the <= log
4689 line. There are two side-effects of this change:
4691 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4692 logged header lines no longer include the local Received: line, because
4693 it has not yet been created. The same applies to a copy of the message
4694 that is returned to a non-SMTP sender when a message is rejected.
4696 (b) When a filter file is tested using -bf, no additional Received: header
4697 is added to the test message. After some thought, I decided that this
4700 This change does not affect the value of $received_for. It is still set
4701 after address rewriting, but before local_scan() is called.
4703 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4705 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4706 gave an unhelpful panic error message, and a defer error. I have managed to
4707 change this behaviour so that it now rejects any supplied certificate,
4708 which seems right, as the list of acceptable certificates is empty.
4710 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4711 gave an unhelpful defer error. I have not managed to make this reject any
4712 supplied certificates, but the error message it gives is "no certificate
4713 supplied", which is not helpful.
4715 70. exigrep's output now also includes lines that are not associated with any
4716 message, but which match the given pattern. Implemented by a patch from
4717 Martin Sluka, which also tidied up the Perl a bit.
4719 71. Recipient callout verification, like sender verification, was using <> in
4720 the MAIL FROM command. This isn't really the right thing, since the actual
4721 sender may affect whether the remote host accepts the recipient or not. I
4722 have changed it to use the actual sender in the callout; this means that
4723 the cache record is now keyed on a recipient/sender pair, not just the
4724 recipient address. There doesn't seem to be a real danger of callout loops,
4725 since a callout by the remote host to check the sender would use <>.
4726 [SEE ABOVE: changed after hitting problems.]
4728 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4729 temporary errors. However, in the case of such a code being given after
4730 the end of a data transmission (i.e. after ".") Exim was failing to write
4731 a retry record for the message. (Yes, there was some broken host that was
4732 actually sending 8xx at this point.)
4734 73. An unknown lookup type in a host list could cause Exim to panic-die when
4735 the list was checked. (An example that provoked this was putting <; in the
4736 middle of a list instead of at the start.) If this happened during a DATA
4737 ACL check, a -D file could be left lying around. This kind of configuration
4738 error no longer causes Exim to die; instead it causes a defer errror. The
4739 incident is still logged to the main and panic logs.
4741 74. Buglet left over from Exim 3 conversion. The message "too many messages
4742 in one connection" was written to the rejectlog but not the mainlog, except
4743 when address rewriting (yes!) was being logged.
4745 75. Added write_rejectlog option.
4747 76. When a system filter was run not as root (that is, when system_filter_user
4748 was set), the values of the $n variables were not being returned to the
4749 main process; thus, they were not subsequently available in the $sn
4752 77. Added +return_path_on_delivery log selector.
4754 78. A connection timeout was being treated differently from recipients deferred
4755 when testing hosts_max_try with a message that was older than the host's
4756 retry timeout. (The host should not be counted, thus allowing all hosts to
4757 be tried at least once before bouncing.) This may have been the cause of an
4758 occasionally reported bug whereby a message would remain on the queue
4759 longer than the retry timeout, but would be bounced if a delivery was
4760 forced. I say "may" because I never totally pinned down the problem;
4761 setting up timeout/retry tests is difficult. See also the next item.
4763 79. The ultimate address timeout was not being applied to errors that involved
4764 a combination of host plus message (for example, a timeout on a MAIL
4765 command). When an address resolved to a number of possible hosts, and they
4766 were not all tried for each delivery (e.g. because of hosts_max_try), a
4767 message could remain on the queue longer than the retry timeout.
4769 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4772 81. Fixed an obscure SMTP outgoing bug which required at least the following
4773 conditions: (a) there was another message waiting for the same server;
4774 (b) the server returned 5xx to all RCPT commands in the first message so
4775 that the message was not completed; (c) the server dropped the connection
4776 or gave a negative response to the RSET that Exim sends to abort the
4777 transaction. The observed case was a dropped connection after DATA that had
4778 been sent in pipelining mode. That is, the server had advertised PIPELINING
4779 but was not implementing it correctly. The effect of the bug was incorrect
4780 behaviour, such as trying another host, and this could lead to a crash.
4786 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4787 and daemon.c were passed as pointers to ints; they should have been
4788 pointers to socklen_t variables (which are typically unsigned ints).
4790 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4793 3. Fixed a really odd bug that affected only the testing scheme; patching a
4794 certain fixed string in the binary changed the value of another string that
4795 happened to be identical to the end of the original first string.
4797 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4798 name", it returns that address as the IP address. On some operating
4799 systems (e.g. Solaris), it also passes back the IP address string as the
4800 "host name". However, on others (e.g. Linux), it passes back an empty
4801 string. Exim wasn't checking for this, and was changing the host name to an
4802 empty string, assuming it had been canonicized.
4804 5. Although rare, it is permitted to have more than one PTR record for a given
4805 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4806 all the names associated with an address, because they do in Solaris.
4807 However, it seems that they do not in Linux for data that comes from the
4808 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4809 I found this out when I moved to a new Linux workstation and tried to run
4810 the Exim test suite.
4812 To get round this problem I have changed the code so that it now does its
4813 own call to the DNS to look up PTR records when searching for a host name.
4814 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4815 addresses that are only in /etc/hosts are still found.
4817 This behaviour is, however, controlled by an option called host_lookup_
4818 order, which defaults to "bydns:byaddr". If people want to use the other
4819 order, or indeed, just use one or the other means of lookup, they can
4820 specify it in this variable.
4822 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4823 some operating systems, this comes back from gethostbyaddr() as an empty
4824 string, and this is what Exim used to test for. However, it seems that in
4825 other systems, "." is yielded. Exim now tests for this case too.
4827 7. The values of check_spool_space and check_log_space are now held internally
4828 as a number of kilobytes instead of an absolute number of bytes. If a
4829 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4830 kilobyte. This means that much larger values can be stored.
4832 8. Exim monitor: an attempt to get the action menu when not actually pointing
4833 at a message produces an empty menu entitled "No message selected". This
4834 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4835 no entries in it ("Shell widget menu has zero width and/or height"). So I
4836 have added a single, blank menu entry in this case.
4838 9. Added ${quote_local_part.
4840 10. MIME decoding is now applied to the contents of Subject: header lines when
4843 11. Now that a reference to $sender_host_address automatically causes a reverse
4844 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4845 host lookup before query-style lookups in lists that might use this
4846 variable. This has therefore been abolished, and the "net-" prefix is no
4847 longer necessary for query-style lookups.
4849 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4850 have been a typo for LFLAGS, so it has been changed.
4852 13. The install script calls Exim with "-C /dev/null" in order to find the
4853 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4854 to be output. Howeve, since Exim outputs its version number before the
4855 error, it didn't break the script. It just looked ugly. I fixed this by
4856 always allowing "-C /dev/null" if the caller is root.
4858 14. Ignore overlarge ACL variable number when reading spool file - insurance
4859 against a later release with more variables having written the file.
4861 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4862 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4863 this, as well as the form without the "IPv6" on the front.
4865 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4866 OS/Makefile-Darwin file.
4868 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4869 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4871 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4872 variable. However, some OS use ints for inodes. I've added cast to long int
4873 to get rid of the compiler warning.
4875 19. I had forgotten to lock out "/../" in configuration file names when
4876 ALT_CONFIG_PREFIX was set.
4878 20. Routers used for verification do not need to specify transports. However,
4879 if such a router generated a host list, and callout was configured, Exim
4880 crashed, because it could not find a port number from the (non-existent)
4881 transport. It now assumes port 25 in this circumstance.
4883 21. Added the -t option to exigrep.
4885 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4886 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4887 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4888 them is set, LOOKUP_LSEARCH is forced.
4890 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4892 24. Added sender and host information to the "rejected by local_scan()" log
4893 line; previously there was no indication of these.
4895 25. Added .include_if_exists.
4897 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4898 Linux. It turns out that not all file systems support this. Apparently some
4899 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4900 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4901 not supported on the file descriptor, is now ignored when Exim is trying to
4902 sync a directory. This applies only to Linux.
4904 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4906 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4907 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4910 29. Added support for maildirsize files from supplied patch (modified a bit).
4912 30. The use of :fail: followed by an empty string could lead Exim to respond to
4913 sender verification failures with (e.g.):
4915 550 Verification failed for <xxx>
4916 550 Sender verify failed
4918 where the first response line was missing the '-' that indicates it is not
4919 the final line of the response.
4921 31. The loop for finding the name of the user that called Exim had a hardwired
4922 limit of 10; it now uses the value of finduser_retries, which is used for
4923 all other user lookups.
4925 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4928 33. Exim was neglecting to zero errno before one call of strtol() when
4929 expanding a string and expecting an integer value. On some systems this
4930 resulted in spurious "integer overflow" errors. Also, it was casting the
4931 result into an int without checking.
4933 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4934 did not work. The code looks as if it has *never* worked, though it appears
4935 to have been documented since at least releast 1.62. I have made it work.
4937 35. The "timeout_DNS" error in retry rules, also documented since at least
4938 1.62, also never worked. As it isn't clear exactly what this means, and
4939 clearly it isn't a major issue, I have abolished the feature by treating it
4940 as "timeout", and writing a warning to the main and panic logs.
4942 36. The display of retry rules for -brt wasn't always showing the error code
4945 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4946 timeout_connect_A, timeout_connect_MX.
4948 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4949 to the empty sender.
4951 39. The daemon was not analysing the content of -oX till after it had closed
4952 stderr and disconnected from the controlling terminal. This meant that any
4953 syntax errors were only noted on the panic log, and the return code from
4954 the command was 0. By re-arranging the code a little, I've made the
4955 decoding happen first, so such errors now appear on stderr, and the return
4956 code is 1. However, the actual setting up of the sockets still happens in
4957 the disconnected process, so errors there are still only recorded on the
4960 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4961 connections (as happens on some IP stacks) was logged at start up time as
4962 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4963 it from "IPv6 and IPv4", which means that two separate sockets are being
4966 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4967 says whether AF_INET or AF_INET6 was passed as an argument.
4969 42. Exiwhat output was messed up when time zones were included in log
4972 43. Exiwhat now gives more information about the daemon's listening ports,
4973 and whether -tls-on-connect was used.
4975 44. The "port" option of the smtp transport is now expanded.
4977 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4978 silently ignored. Now an error message is written to the main and panic
4981 46. There's a new ACL modifier called "logwrite" which writes to a log file
4982 as soon as it is encountered.
4984 47. Added $local_user_uid and $local_user_gid at routing time.
4986 48. Exim crashed when trying to verify a sender address that was being
4989 49. Exim was recognizing only a space character after ".include". It now also
4990 recognizes a tab character.
4992 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4993 extracting the relevant information from the specification. The man page no
4994 longer contains scrambled data for the -d option, and I've added a section
4995 at the front about calling Exim under different names.
4997 51. Added "extra_headers" argument to the "mail" command in filter files.
4999 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5002 53. Installed eximstats 1.29.
5004 54. Added transport_filter_timeout as a generic transport option.
5006 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5007 Cc: header lines. This was required by RFC 822, but it not required by RFC
5010 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5011 incoming messages that did not have them. Now it does so only if the
5012 message originates locally, that is, if there is no associated remote host
5013 address. When Resent- header lines are present, this applies to the Resent-
5014 lines rather than the non-Resent- lines.
5016 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5017 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5019 58. Messages for configuration errors now include the name of the main
5020 configuration file - useful now that there may be more than one file in a
5021 list (.included file names were always shown).
5023 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5024 for those rare installations that do not start the daemon as root or run it
5025 setuid root. I've cut out the call to initgroups() if the daemon is not
5028 60. The Exim user and group can now be bound into the binary as text strings
5029 that are looked up at the start of Exim's processing.
5031 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5033 62. Added $mailstore_basename variable.
5035 63. Installed patch to sieve.c from Michael Haardt.
5037 64. When Exim failed to open the panic log after failing to open the main log,
5038 the original message it was trying to log was written to stderr and debug
5039 output, but if they were not available (the usual case in production), it
5040 was lost. Now it is written to syslog before the two lines that record the
5041 failures to open the logs.
5043 65. Users' Exim filters run in subprocesses under the user's uid. It is
5044 possible for a "deliver" command or an alias in a "personal" command to
5045 provoke an address rewrite. If logging of address rewriting is configured,
5046 this fails because the process is not running as root or exim. There may be
5047 a better way of dealing with this, but for the moment (because 4.30 needs
5048 to be released), I have disabled address rewrite logging when running a
5049 filter in a non-root, non-exim process.
5055 1. The buildconfig auxiliary program wasn't quoting the value set for
5056 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5057 not defined. This bug was masked in 4.22 by the effect that was fixed in
5060 2. Some messages that were rejected after a message id was allocated were
5061 shown as "incomplete" by exigrep. It no longer does this for messages that
5062 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5064 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5065 have allow_domain_literals set, the ID did not get logged in the <= line.
5066 Domain literals are now always recognized in Message-ID: header lines.
5068 4. The first argument for a ${extract expansion item is the key name or field
5069 number. Leading and trailing spaces in this item were not being ignored,
5070 causing some misleading effects.
5072 5. When deliver_drop_privilege was set, single queue runner processes started
5073 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5074 same command in the process it spins off) were not dropping privilege.
5076 6. When the daemon running as "exim" started a queue runner, it always
5077 re-executed Exim in the spun-off process. This is a waste of effort when
5078 deliver_drop_privilege is set. The new process now just calls the
5079 queue-runner function directly.
5085 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5088 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5089 ignored. Though the use of -oP was forcing the writing of a pid file, it
5090 was always written to the default place.
5092 3. If the message "no IP address found for host xxxx" is generated during
5093 incoming verification, it is now followed by identification of the incoming
5094 connection (so you can more easily find what provoked it).
5096 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5098 5. Added some features to "harden" Exim a bit more against certain attacks:
5100 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5101 be put in Local/Makefile. This is like the never_users runtime option,
5102 but it cannot be overridden. The default setting is "root".
5104 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5105 prefix string with which any file named in a -C command line option
5108 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5109 is retained for -C and -D only if the caller of Exim is root. Without
5110 it, the exim user may also use -C and -D and retain privilege.
5112 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5113 command line option is disabled.
5115 6. Macro names set by the -D option must start with an upper case letter, just
5116 like macro names defined in the configuration file.
5118 7. Added "dereference=" facility to LDAP.
5120 8. Two instances of the typo "uknown" in the source files are fixed.
5122 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5123 the Configure-Makefile script screwed up while processing it.
5125 10. Incorporated PCRE 4.4.
5127 11. The SMTP synchronization check was not operating right at the start of an
5128 SMTP session. For example, it could not catch a HELO sent before the client
5129 waited for the greeting. There is now a check for outstanding input at the
5130 point when the greeting is written. Because of the duplex, asynchronous
5131 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5132 way, but not yet received, when the check is performed.
5134 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5135 on TCP/IP sockets, because this apparently causes some broken clients to
5138 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5139 unchanged) from the Cygwin maintainer.
5141 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5142 was supported instead of testing for mbx. Effectively a typo.
5144 15. The spa authenticator server code was not checking that the input it
5145 received was valid base64.
5147 16. The debug output line for the "set" modifier in ACLs was not showing the
5148 name of the variable that was being set.
5150 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5152 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5153 lookup on its own. Something else was needed to trigger the lookup. For
5154 example, a match in host_lookup or the need for a host name in a host list.
5155 Now, if $sender_host_name is referenced and the host name has not yet been
5156 looked up, a lookup is performed. If the lookup fails, the variable remains
5157 empty, and $host_lookup_failed is set to "1".
5159 19. Added "eqi" as a case-independent comparison operator.
5161 20. The saslauthd authentication condition could segfault if neither service
5162 nor realm was specified.
5164 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5165 error message that was logged was misleading, and incoming SMTP
5166 connections were dropped. The message is now more accurate, and temporary
5167 errors are given to SMTP connections.
5169 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5170 (except RSET) with a 421 error, until QUIT is received. However, it was
5171 failing to send a response to QUIT.
5173 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5174 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5175 ACL is now run after the helo_try_verify_hosts code.
5177 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5178 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5179 case-independent, so other case variants are also recognized). Apparently
5180 some systems use these upper case variants.
5182 25. If more than two messages were waiting for the same host, and a transport
5183 filter was specified for the transport, Exim sent two messages over the
5184 same TCP/IP connection, and then failed with "socket operation on non-
5185 socket" when it tried to send the third.
5187 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5189 27. The extern definition of crypt16() in expand.c was not being excluded when
5190 the OS had its own crypt16() function.
5192 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5193 as a preferred synonym for return_size_limit, both as an option and as an
5196 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5198 30. Changed the default configuration ACL to relax the local part checking rule
5199 for addresses that are not in any local domains. For these addresses,
5200 slashes and pipe symbols are allowed within local parts, but the sequence
5201 /../ is explicitly forbidden.
5203 31. SPA server authentication was not clearing the challenge buffer before
5206 32. log_message in a "warn" ACL statement was writing to the reject log as
5207 well as to the main log, which contradicts the documentation and doesn't
5208 seem right (because no rejection is happening). So I have stopped it.
5210 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5211 However, I am unable to do any testing of this.
5213 34. Fixed an infelicity in the appendfile transport. When checking directories
5214 for a mailbox, to see if any needed to be created, it was accidentally
5215 using path names with one or more superfluous leading slashes; tracing
5216 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5218 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5219 contents are added to the log line that is written for every discarded
5220 recipient. (Previously a log_message setting was ignored.)
5222 36. The ${quote: operator now quotes the string if it is empty.
5224 37. The install script runs exim in order to find its version number. If for
5225 some reason other than non-existence or emptiness, which it checks, it
5226 could not run './exim', it was installing it with an empty version number,
5227 i.e. as "exim-". This error state is now caught, and the installation is
5230 38. An argument was missing from the function that creates an error message
5231 when Exim fails to connect to the socket for saslauthd authentication.
5232 This could cause Exim to crash, or give a corrupted message.
5234 39. Added isip, isip4, and isip6 to ${if conditions.
5236 40. The ACL variables $acl_xx are now saved with the message, and can be
5237 accessed later in routers, transports, and filters.
5239 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5240 strings in the file are not string-expanded.
5242 42. If a MAIL command specified a SIZE value that was too large to fit into an
5243 int variable, the check against message_size_limit failed. Such values are
5244 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5245 day this will have to be increased, but I don't think I want to be around
5246 when emails are that large.
5253 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5254 iconv() is not standard in FreeBSD.
5256 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5257 IPv6 enabled. The observed symptom was a segmentation fault on return from
5258 the function os_common_find_running_interfaces() in src/os.c.
5260 3. In the check_special_case() function in daemon.c I had used "errno" as an
5261 argument name, which causes warnings on some systems. This was basically a
5262 typo, since it was named "eno" in the comments!
5264 4. The code that waits for the clock to tick (at a resolution of some fraction
5265 of a second) so as to ensure message-id uniqueness was always waiting for
5266 at least one whole tick, when it could have waited for less. [This is
5267 almost certainly not relevant at current processor speeds, where it is
5268 unlikely to ever wait at all. But we try to future-proof.]
5270 5. The function that sleeps for a time interval that includes fractions of a
5271 second contained a race. It did not block SIGALRM between setting the
5272 timer, and suspending (a couple of lines later). If the interval was short
5273 and the sigsuspend() was delayed until after it had expired, the suspension
5274 never ended. On busy systems this could lead to processes getting stuck for
5277 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5278 process, before it forks any delivery processes. The open lookup caching
5279 mechanism meant that the open file or database connection was passed into
5280 the delivery process. The problem was that delivery processes always tidy
5281 up cached lookup data. This could cause a problem for the next delivery
5282 process started by the queue runner, because the external queue runner
5283 process does not know about the closure. So the next delivery process
5284 still has data in the lookup cache. In the case of a file lookup, there was
5285 no problem because closing a file descriptor in a subprocess doesn't affect
5286 the parent. However, if the lookup was caching a connection to a database,
5287 the connection was closed, and the second delivery process was likely to
5288 see errors such as "PGSQL: query failed: server closed the connection
5289 unexpectedly". The problem has been fixed by closing all cached lookups
5290 in a queue runner before running a delivery process.
5292 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5293 seem to have the "const" qualifier which it has on other OS. I've
5296 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5297 *of the same type* (client or server) with the same public name that an
5298 error should be diagnosed.
5300 9. When Exim looked up a host name for an IP address, but failed to find the
5301 original IP address when looking up the host name (a safety check), it
5302 output the message "<ip address> does not match any IP for NULL", which was
5303 confusing, to say the least. The bug was that the host name should have
5304 appeared instead of "NULL".
5306 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5307 user that is built into the binary, and the -C or -D options is used, root
5308 privilege is dropped before the configuration file is read. In addition,
5309 logging is switched to stderr instead of the normal log files. If the
5310 configuration then re-defines the Exim user, the unprivileged environment
5311 is probably not what is expected, so Exim logs a panic warning message (but
5314 However, if deliver_drop_privilege is set, the unprivileged state may well
5315 be exactly what is intended, so the warning has been cut out in that case,
5316 and Exim is allowed to try to write to its normal log files.
5322 1. smtp_return_error_details was not giving details for temporary sender
5323 or receiver verification errors.
5325 2. Diagnose a configuration error if two authenticators have the same public
5328 3. Exim used not to create the message log file for a message until the first
5329 delivery attempt. This could be confusing when incoming messages were held
5330 for policy or load reasons. The message log file is now created at the time
5331 the message is received, and an initial "Received" line is written to it.
5333 4. The automatically generated man page for command line options had a minor
5334 bug that caused no ill effects; however, a more serious problem was that
5335 the procedure for building the man page automatically didn't always
5336 operate. Consequently, release 4.20 contains an out-of-date version. This
5337 shouldn't happen again.
5339 5. When building Exim with embedded Perl support, the script that builds the
5340 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5341 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5343 6. The freeze_tell option was not being used for messages that were frozen on
5344 arrival, either by an ACL or by local_scan().
5346 7. Added the smtp_incomplete_transaction log selector.
5348 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5349 was accepting AUTH without a new EHLO.
5351 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5352 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5355 10. Logging of TCP/IP connections (when configured) now happens in the main
5356 daemon process instead of the child process, so that the TCP/IP connection
5357 count is more accurate (but it can never be perfect).
5359 11. The use of "drop" in a nested ACL was not being handled correctly in the
5360 outer ACL. Now, if condition failure induced by the nested "drop" causes
5361 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5362 or "require"), the connection is dropped.
5364 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5365 that yield "discard" can now be used with an "accept" or a "discard" verb,
5366 but an error is generated for any others (because I can't see a useful way
5367 to define what should happen).
5369 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5370 are now processed in the same way as lines in the Exim configuration file.
5371 In particular, continuation lines are supported.
5373 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5375 15. Added -ti meaning -t -i.
5377 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5378 domains, and warn by logging if others are found.
5380 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5381 returned from GIFCONF when Exim is trying to find the list of interfaces on
5382 a host. The code in os.c has been modified to copy each ifreq to an aligned
5383 structure in all cases.
5385 Also, in some cases, the returned ifreq's were being copied to a 'struct
5386 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5387 means the last couple of bytes of an IPv6 address could be chopped if the
5388 ifreq contained only a normal sockaddr (14 bytes storage).
5390 18. Named domain lists were not supported in the hosts_treat_as_local option.
5391 An entry such as +xxxx was not recognized, and was treated as a literal
5394 19. Ensure that header lines added by a DATA ACL are included in the reject log
5395 if the ACL subsequently rejects the message.
5397 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5398 MD5 (which is deprecated).
5400 21. When testing a filter file using -bf, Exim was writing a message when it
5401 took the sender from a "From " line in the message, but it was not doing so
5402 when it took $return_path from a Return-Path: header line. It now does.
5404 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5405 with a valid header line field name (a series of printing characters
5406 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5408 23. Changed "disc" in the source to "disk" to conform to the documentation and
5409 the book and for uniformity.
5411 24. Ignore Sendmail's -Ooption=value command line item.
5413 25. When execve() failed while trying to run a command in a pipe transport,
5414 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5415 could be confused with a return value of 69 from the command itself. This
5416 has been changed to 127, the value the shell returns if it is asked to run
5417 a non-existent command. The wording for the related log line suggests a
5418 non-existent command as the problem.
5420 26. If received_header_text expands to an empty string, do not add a Received:
5421 header line to the message. (Well, it adds a token one on the spool, but
5422 marks it "old" so that it doesn't get used or transmitted.)
5424 27. Installed eximstats 1.28 (addition of -nt option).
5426 28. There was no check for failure on the call to getsockname() in the daemon
5427 code. This can fail if there is a shortage of resources on the system, with
5428 ENOMEM, for example. A temporary error is now given on failure.
5430 29. Contrary to the C standard, it seems that in some environments, the
5431 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5432 program. Exim now does this explicitly; it affects the formatting of
5433 timestamps using strftime().
5435 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5436 complaints. I've now initialized all the variables, to avoid this.
5438 32. Header lines added by a system filter were not being "seen" during
5439 transport-time rewrites.
5441 33. The info_callback() function passed to OpenSSL is set up with type void
5442 (*)(SSL *, int, int), as described somewhere. However, when calling the
5443 function (actually a macro) that sets it up, the type void(*)() is
5444 expected. I've put in a cast to prevent warnings from picky compilers.
5446 34. If a DNS black list lookup found a CNAME record, but there were no A
5447 records associated with the domain it pointed at, Exim crashed.
5449 35. If a DNS black list lookup returned more than one A record, Exim ignored
5450 all but the first. It now scans all returned addresses if a particular IP
5451 value is being sought. In this situation, the contents of the
5452 $dnslist_value variable are a list of all the addresses, separated by a
5455 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5456 to accept a host name and all its aliases if the forward lookup for any of
5457 them yielded the IP address of the incoming connection. Now it accepts only
5458 those names whose forward lookup yields the correct IP address. Any other
5459 names are discarded. This closes a loophole whereby a rogue DNS
5460 administrator could create reverse DNS records to break through a
5461 wildcarded host restriction in an ACL.
5463 37. If a user filter or a system filter that ran in a subprocess used any of
5464 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5465 the wrong values were passed to the pipe command ($thisaddress had the
5466 value of $0, $0 had the value of $1, etc). This bug was introduced by
5467 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5469 38. Improved the line breaking for long SMTP error messages from ACLs.
5470 Previously, if there was no break point between 40 and 75 characters, Exim
5471 left the rest of the message alone. Two changes have been made: (a) I've
5472 reduced the minimum length to 35 characters; (b) if it can't find a break
5473 point between 35 and 75 characters, it looks ahead and uses the first one
5474 that it finds. This may give the occasional overlong line, but at least the
5475 remaining text gets split now.
5477 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5478 file descriptors might be low, and that setting 1000 would always raise it.
5479 It turns out that in some environments, the limit is already over 1000 and
5480 that lowering it causes trouble. So now Exim takes care not to decrease it.
5482 40. When delivering a message, the value of $return_path is set to $sender_
5483 address at the start of routing (routers may change the value). By an
5484 oversight, this default was not being set up when an address was tested by
5485 -bt or -bv, which affected the outcome if any router or filter referred to
5488 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5489 log only when "message" or "log_message" are set. However, if one of the
5490 conditions was an address verification, or a call to a nested ACL, the
5491 messages generated by the underlying test were being passed through. This
5492 no longer happens. The underlying message is available in $acl_verify_
5493 message for both "message" and "log_message" expansions, so it can be
5494 passed through if needed.
5496 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5497 new expansion $bh_ to give the encoded byte string without charset
5498 translation. Translation happens only if iconv() is available; HAVE_ICONV
5499 indicates this at build time. HEADERS_CHARSET gives the charset to
5500 translate to; headers_charset can change it in the configuration, and
5501 "headers charset" can change it in an individual filter file.
5503 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5504 that creates RFC 2047 encoded "words" labels them as that charset instead
5505 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5506 expansion operator; (ii) when Exim creates a From: line for a local
5507 message; (iii) when a header line is rewritten to include a "phrase" part.
5509 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5510 buggy, causing it to skip the first lines of messages whose message ID
5511 ended in 'D'. This would not have bitten before Exim release 4.14, because
5512 message IDs were unlikely to end in 'D' before then. The effect was to have
5513 incorrect size information for certain domains.
5515 45. #include "config.h" was missing at the start of the crypt16.c module. This
5516 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5519 46. If there was a timeout during a "random" callout check, Exim treated it as
5520 a failure of the random address, and carried on sending RSET and the real
5521 address. If the delay was just some slowness somewhere, the response to the
5522 original RCPT would be taken as a response to RSET and so on, causing
5523 mayhem of various kinds.
5525 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5526 when I implemented it. It didn't allow for the fact that some option values
5527 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5528 the right test for positive values.
5530 48. Domain names in DNS records are case-independent. Exim always looks them up
5531 in lower case. Some resolvers return domain names in exactly the case they
5532 appear in the zone file, that is, they may contain uppercase letters. Not
5533 all resolvers do this - some return always lower case. Exim was treating a
5534 change of case by a resolver as a change of domain, similar to a widening
5535 of a domain abbreviation. This triggered its re-routing code and so it was
5536 trying to route what was effectively the same domain again. This normally
5537 caused routing to fail (because the router wouldn't handle the domain
5538 twice). Now Exim checks for this case specially, and just changes the
5539 casing of the domain that it ultimately uses when it transmits the message
5542 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5545 50. If a filter generated a file delivery with a non-absolute name (possible if
5546 no home directory exists for the router), the forbid_file option was not
5549 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5550 the existing equality matching.
5552 52. Exim was using ints instead of ino_t variables in some places where it was
5553 dealing with inode numbers.
5555 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5556 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5557 called TMPDIR, and if it finds it is different, it changes its value.
5559 54. The smtp_printf() function is now made available to local_scan() so
5560 additional output lines can be written before returning. There is also an
5561 smtp_fflush() function to enable the detection of a dropped connection.
5562 The variables smtp_input and smtp_batched_input are exported to
5565 55. Changed the default runtime configuration: the message "Unknown user"
5566 has been removed from the ACL, and instead placed on the localuser router,
5567 using the cannot_route_message feature. This means that any verification
5568 failures that generate their own messages won't get overridden. Similarly,
5569 the "Unrouteable address" message that was in the ACL for unverifiable
5570 relay addresses has also been removed.
5572 56. Added hosts_avoid_esmtp to the smtp transport.
5574 57. The exicyclog script was not checking for the esoteric option
5575 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5576 will work only if exicyclog is run under the appropriate euid.
5578 58. Following a discussion on the list, the rules by which Exim recognises line
5579 endings on incoming messages have been changed. The -dropcr and drop_cr
5580 options are now no-ops, retained only for backwards compatibility. The
5581 following line terminators are recognized: LF CRLF CR. However, special
5582 processing applies to CR:
5584 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5585 nor a local message in the state where . is a terminator.
5587 (ii) If a bare CR is encountered in a header line, an extra space is added
5588 after the line terminator so as not to end the header. The reasoning
5589 behind this is that bare CRs in header lines are most likely either
5590 to be mistakes, or people trying to play silly games.
5592 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5593 was being incorrectly given as 18 bytes larger than it should have been.
5594 This is a VOB (very old bug).
5596 60. This may never have affected anything current, but just in case it has:
5597 When the local host is found other than at the start of a list of hosts,
5598 the local host, those with the same MX, and any that follow, are discarded.
5599 When the list in question was part of a longer list of hosts, the following
5600 hosts (not currently being processed) were also being discarded. This no
5601 longer happens. I'm not sure if this situation could ever has previously
5604 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5607 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5608 header. This is something that is recommended in a new Internet Draft, and
5609 is something that is documented as being done by Sendmail. There are two
5610 possible values. For messages generated by the autoreply transport, Exim
5613 Auto-Submitted: auto-replied
5615 whereas for all other generated messages (e.g. bounces) it adds
5617 Auto-Submitted: auto-generated
5619 63. The "personal" condition in filters now includes a test for the
5620 Auto-Submitted: header. If it contains the string "auto-" the message it
5621 not considered personal.
5623 64. Added rcpt_include_affixes as a generic transport option.
5625 65. Added queue_only_override (default true).
5627 66. Added the syslog_duplication option.
5629 67. If what should have been the first header line of a message consisted of
5630 a space followed by a colon, Exim was mis-interpreting it as a header line.
5631 It isn't of course - it is syntactically invalid and should therefore be
5632 treated as the start of the message body. The misbehaviour could have
5633 caused a number of strange effects, including loss of data in subsequent
5634 header lines, and spool format errors.
5636 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5637 client host had authenticated. This control can now be exercised by an ACL
5638 for more flexibility.
5640 69. By default, callouts do not happen when testing with -bh. There is now a
5641 variant, -bhc, which does actually run the callout code, including
5642 consulting and updating the callout cache.
5644 70. Added support for saslauthd authentication, courtesy of Alexander
5647 71. If statvfs() failed on the spool or log directories while checking their
5648 size for availability, Exim confusingly gave the error "space shortage".
5649 Furthermore, in debugging mode it crashed with a floating point exception.
5650 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5651 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5652 serious problem, Exim now writes to the main and panic logs when this
5653 happens, with details of the failure. It then refuses to accept the
5654 incoming message, giving the message "spool directory problem" or "log
5655 directory problem" with a 421 code for SMTP messages.
5657 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5658 0, 1, and 2 exist, because some OS complain for execs without them (see
5659 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5660 descriptors. However, the code omitted to check that the open succeeded,
5661 causing mysterious errors if for some reason the permissions on /dev/null
5662 got screwed. Now Exim writes a message to the main and panic logs, and
5663 bombs out if it can't open /dev/null.
5665 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5666 interact so that it is all more flexible. It is supposed to remain
5667 backwards compatible. Also added extra_local_interfaces.
5669 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5670 to bomb out with an assertion failure - to the client this appears as a
5671 connection drop. This problem occurs in the part of the code that was taken
5672 from the Samba project. Fortunately, the assertion is in a very simple
5673 function, so I have fixed this by reproducing the function inline in the
5674 one place where it is called, and arranging for authentication to fail
5675 instead of killing the process with assert().
5677 75. The SPA client code was not working when the server requested OEM rather
5678 than Unicode encoding.
5680 76. Added code to make require_files with a specific uid setting more usable in
5681 the case where statting the file as root fails - usually a non-root-mounted
5682 NFS file system. When this happens and the failure is EACCES, Exim now
5683 forks a subprocess and does the per-uid checking as the relevant uid.
5685 77. Added process_log_path.
5687 78. If log_file_path was not explicitly set, a setting of check_log_space or
5688 check_log_inodes was ignored.
5690 79. If a space check for the spool or log partitions fails, the incident is now
5691 logged. Of course, in the latter case the data may get lost...
5693 80. Added the %p formatting code to string_format() so that it can be used to
5694 print addresses in debug_print(). Adjusted all the address printing in the
5695 debugging in store.c to use %p rather than %d.
5697 81. There was a concern that a line of code in smtp_in.c could overflow a
5698 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5699 initially expressed, the concern was not well-founded, because trailing
5700 spaces are removed early. However, if the trailing spaces were followed by
5701 a NULL, they did not get removed, so the overflow was possible. Two fixes
5704 (a) I re-wrote the offending code in a cleaner fashion.
5705 (b) If an incoming SMTP command contains a NULL character, it is rejected
5708 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5709 runs initgroups(), so that if the Exim user is in any additional groups,
5710 they will be used during message reception.
5716 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob