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 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
38 JH/02 Support "G" suffix to numbers in ${if comparisons.
40 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
42 NM/01 Bugzilla 1197 - Spec typo
43 Bugzilla 1196 - Spec examples corrections
45 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
50 PP/01 Handle short writes when writing local log-files.
51 In practice, only affects FreeBSD (8 onwards).
52 Bugzilla 1053, with thanks to Dmitry Isaikin.
54 NM/01 Bugzilla 949 - Documentation tweak
56 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
59 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
61 PP/02 Implemented gsasl authenticator.
63 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
65 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
66 `pkg-config foo` for cflags/libs.
68 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
69 with rest of GSASL and with heimdal_gssapi.
71 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
72 `pkg-config foo` for cflags/libs for the TLS implementation.
74 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
75 properties get this fed in as external SSF. A number of robustness
76 and debugging improvements to the cyrus_sasl authenticator.
78 PP/08 cyrus_sasl server now expands the server_realm option.
80 PP/09 Bugzilla 1214 - Log authentication information in reject log.
81 Patch by Jeremy Harris.
83 PP/10 Added dbmjz lookup type.
85 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
87 PP/12 MAIL args handles TAB as well as SP, for better interop with
88 non-compliant senders.
89 Analysis and variant patch by Todd Lyons.
91 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
92 Bug report from Lars Müller <lars@samba.org> (via SUSE),
93 Patch from Dirk Mueller <dmueller@suse.com>
95 PP/13 tls_peerdn now print-escaped for spool files.
96 Observed some $tls_peerdn in wild which contained \n, which resulted
97 in spool file corruption.
99 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
100 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
101 or write after TLS renegotiation, which otherwise led to messages
104 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
105 as a tracking header (ie: a signed header comes before the signature).
106 Patch from Wolfgang Breyha.
108 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
109 comma-sep list; embedded commas doubled.
111 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
113 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
115 Report and patch from Dmitry Banschikov.
117 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
118 Removed SSL_clear() after SSL_new() which led to protocol negotiation
119 failures. We appear to now support TLS1.1+ with Exim.
121 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
122 lets Exim select keys and certificates based upon TLS SNI from client.
123 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
124 before an outbound SMTP session. New log_selector, +tls_sni.
126 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
127 NULL dereference. Report and patch from Alun Jones.
129 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
130 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
131 Not seeing resolver debug output on NetBSD, but suspect this is a
132 resolver implementation change.
134 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
135 Left warnings. Added "eximon gdb" invocation mode.
137 PP/21 Defaulting "accept_8bitmime" to true, not false.
139 PP/22 Added -bw for inetd wait mode support.
141 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
142 locate the relevant includes and libraries. Made this the default.
144 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
145 Bugzilla 1246, report and most of solution from Tomasz Kusy.
147 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
148 This may cause build issues on older platforms.
150 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
151 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
152 gnutls_require_mac & gnutls_require_protocols (no longer supported).
153 Added SNI support via GnuTLS too.
154 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
156 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
158 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
159 Applied second patch from Janne, implementing suggestion to default
160 multiple-strings-in-record handling to match SPF spec.
162 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
164 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
165 read-only, out of scope).
166 Patch from Wolfgang Breyha, report from Stuart Northfield.
168 PP/29 Fix three issues highlighted by clang analyser static analysis.
169 Only crash-plausible issue would require the Cambridge-specific
170 iplookup router and a misconfiguration.
171 Report from Marcin Mirosław.
173 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
175 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
176 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
177 As part of this, removing so much warning spew let me fix some minor
178 real issues in debug logging.
180 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
181 assignment on my part. Fixed.
183 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
184 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
185 Janne Snabb (who went above and beyond: thank you).
187 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
188 string otherwise requires a connection and a bunch more work and it's
189 relatively easy to get wrong. Should also expose TLS library linkage
192 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
193 64-bit ${eval} (JH/03).
195 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
196 GNU libc to support some of the 64-bit stuff, should not lead to
197 conflicts. Defined before os.h is pulled in, so if a given platform
198 needs to override this, it can.
200 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
201 protection layer was required, which is not implemented.
202 Bugzilla 1254, patch from Wolfgang Breyha.
204 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
205 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
206 tls_dhparam take prime identifiers. Also unbreak combination of
207 OpenSSL+DH_params+TLSSNI.
209 PP/39 Disable SSLv2 by default in OpenSSL support.
215 PP/01 Solaris build fix for Oracle's LDAP libraries.
216 Bugzilla 1109, patch from Stephen Usher.
218 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
220 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
223 TF/02 Fix a couple more cases where we did not log the error message
224 when unlink() failed. See also change 4.74-TF/03.
226 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
227 lock up or crash if it happened to be inside a call to libc when it
228 got a SIGUSR1 from exiwhat.
230 The SIGUSR1 handler appends the current process status to the process
231 log which is later printed by exiwhat. It used to use the general
232 purpose logging code to do this, but several functions it calls are
233 not safe for signals.
235 The new output code in the SIGUSR1 handler is specific to the process
236 log, and simple enough that it's easy to inspect for signal safety.
237 Removing some special cases also simplifies the general logging code.
238 Removing the spurious timestamps from the process log simplifies
241 TF/04 Improved ratelimit ACL condition.
243 The /noupdate option has been deprecated in favour of /readonly which
244 has clearer semantics. The /leaky, /strict, and /readonly update modes
245 are mutually exclusive. The update mode is no longer included in the
246 database key; it just determines when the database is updated. (This
247 means that when you upgrde Exim will forget old rate measurements.)
249 Exim now checks that the per_* options are used with an update mode that
250 makes sense for the current ACL. For example, when Exim is processing a
251 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
252 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
253 must specify per_mail/readonly. If you omit the update mode it defaults to
254 /leaky where that makes sense (as before) or /readonly where required.
256 The /noupdate option is now undocumented but still supported for
257 backwards compatibility. It is equivalent to /readonly except that in
258 ACLs where /readonly is required you may specify /leaky/noupdate or
259 /strict/noupdate which are treated the same as /readonly.
261 A useful new feature is the /count= option. This is a generalization
262 of the per_byte option, so that you can measure the throughput of other
263 aggregate values. For example, the per_byte option is now equivalent
264 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
266 The per_rcpt option has been generalized using the /count= mechanism
267 (though it's more complicated than the per_byte equivalence). When it is
268 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
269 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
270 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
271 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
272 Note that using per_rcpt with a non-readonly update mode in more than
273 one ACL will cause the recipients to be double-counted. (The per_mail
274 and per_byte options don't have this problem.)
276 The handling of very low rates has changed slightly. If the computed rate
277 is less than the event's count (usually one) then this event is the first
278 after a long gap. In this case the rate is set to the same as this event's
279 count, so that the first message of a spam run is counted properly.
281 The major new feature is a mechanism for counting the rate of unique
282 events. The new per_addr option counts the number of different
283 recipients that someone has sent messages to in the last time period. It
284 behaves like per_rcpt if all the recipient addresses are different, but
285 duplicate recipient addresses do not increase the measured rate. Like
286 the /count= option this is a general mechanism, so the per_addr option
287 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
288 example, measure the rate that a client uses different sender addresses
289 with the options per_mail/unique=$sender_address. There are further
290 details in the main documentation.
292 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
294 TF/06 Removed a few PCRE remnants.
296 TF/07 Automatically extract Exim's version number from tags in the git
297 repository when doing development or release builds.
299 PP/02 Raise smtp_cmd_buffer_size to 16kB.
300 Bugzilla 879. Patch from Paul Fisher.
302 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
303 Heavily based on revision 40f9a89a from Simon Arlott's tree.
306 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
308 PP/05 Variable $av_failed, true if the AV scanner deferred.
309 Bugzilla 1078. Patch from John Horne.
311 PP/06 Stop make process more reliably on build failure.
312 Bugzilla 1087. Patch from Heiko Schlittermann.
314 PP/07 Make maildir_use_size_file an _expandable_ boolean.
315 Bugzilla 1089. Patch from Heiko Schlittermann.
317 PP/08 Handle ${run} returning more data than OS pipe buffer size.
318 Bugzilla 1131. Patch from Holger Weiß.
320 PP/09 Handle IPv6 addresses with SPF.
321 Bugzilla 860. Patch from Wolfgang Breyha.
323 PP/10 GnuTLS: support TLS 1.2 & 1.1.
325 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
328 PP/11 match_* no longer expand right-hand-side by default.
329 New compile-time build option, EXPAND_LISTMATCH_RHS.
330 New expansion conditions, "inlist", "inlisti".
332 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
334 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
336 PP/14 fix log_write() format string regression from TF/03.
337 Bugzilla 1152. Patch from Dmitry Isaikin.
343 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
345 PP/02 Harmonised TLS library version reporting; only show if debugging.
346 Layout now matches that introduced for other libraries in 4.74 PP/03.
348 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
350 PP/04 New "dns_use_edns0" global option.
352 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
355 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
356 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
358 TK/01 Updated PolarSSL code to 0.14.2.
359 Bugzilla 1097. Patch from Andreas Metzler.
361 PP/07 Catch divide-by-zero in ${eval:...}.
364 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
367 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
368 format-string attack -- SECURITY: remote arbitrary code execution.
370 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
371 time unintentionally subject to list matching rules, letting the header
372 cause arbitrary Exim lookups (of items which can occur in lists, *not*
373 arbitrary string expansion). This allowed for information disclosure.
375 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
376 INT_MIN/-1 -- value coerced to INT_MAX.
382 NM/01 Workround for PCRE version dependency in version reporting
385 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
386 This fixes portability to compilers other than gcc, notably
387 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
389 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
390 makefiles for portability to HP-UX and POSIX correctness.
392 PP/01 Permit LOOKUP_foo enabling on the make command-line.
393 Also via indented variable definition in the Makefile.
394 (Debugging by Oliver Heesakkers).
396 PP/02 Restore caching of spamd results with expanded spamd_address.
397 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
399 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
400 Improves build reliability. Fix from: Frank Elsner
402 NM/02 Fix wide character breakage in the rfc2047 coding
403 Fixes bug 1064. Patch from Andrey N. Oktyabrski
405 NM/03 Allow underscore in dnslist lookups
406 Fixes bug 1026. Patch from Graeme Fowler
408 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
409 Code patches from Adam Ciarcinski of NetBSD.
411 NM/04 Fixed exiqgrep to cope with mailq missing size issue
414 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
415 is logged, to avoid truncation. Patch from John Horne.
417 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
418 Patch from Jakob Hirsch.
420 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
421 SQL string expansion failure details.
422 Patch from Andrey Oktyabrski.
424 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
425 Patch from Simon Arlott.
427 PP/09 New lookups functionality failed to compile on old gcc which rejects
428 extern declarations in function scope.
429 Patch from Oliver Fleischmann
431 PP/10 Use sig_atomic_t for flags set from signal handlers.
432 Check getgroups() return and improve debugging.
433 Fixed developed for diagnosis in bug 927 (which turned out to be
436 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
437 Patch from Mark Zealey.
439 PP/12 Bugzilla 1056: Improved spamd server selection.
440 Patch from Mark Zealey.
442 PP/13 Bugzilla 1086: Deal with maildir quota file races.
443 Based on patch from Heiko Schlittermann.
445 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
446 Patch from Uwe Doering, sign-off by Michael Haardt.
448 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
449 variable declaration deep within a block. Bug and patch from
452 PP/15 lookups-Makefile IRIX compatibilty coercion.
454 PP/16 Make DISABLE_DKIM build knob functional.
456 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
457 Patch by Simon Arlott
459 TF/03 Fix valgrind.h portability to C89 compilers that do not support
460 variable argument macros. Our copy now differs from upstream.
466 TF/01 Failure to get a lock on a hints database can have serious
467 consequences so log it to the panic log.
469 TF/02 Log LMTP confirmation messages in the same way as SMTP,
470 controlled using the smtp_confirmation log selector.
472 TF/03 Include the error message when we fail to unlink a spool file.
474 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
475 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
476 for maintaining out-of-tree patches for some time.
478 PP/01 Bugzilla 139: Documentation and portability issues.
479 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
480 Handle per-OS dynamic-module compilation flags.
482 PP/02 Let /dev/null have normal permissions.
483 The 4.73 fixes were a little too stringent and complained about the
484 permissions on /dev/null. Exempt it from some checks.
485 Reported by Andreas M. Kirchwitz.
487 PP/03 Report version information for many libraries, including
488 Exim version information for dynamically loaded libraries. Created
489 version.h, now support a version extension string for distributors
490 who patch heavily. Dynamic module ABI change.
492 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
493 privilege escalation vulnerability whereby the Exim run-time user
494 can cause root to append content of the attacker's choosing to
497 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
500 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
501 If dropping privileges for untrusted macros, we disabled normal logging
502 on the basis that it would fail; for the Exim run-time user, this is not
503 the case, and it resulted in successful deliveries going unlogged.
504 Fixed. Reported by Andreas Metzler.
510 PP/01 Date: & Message-Id: revert to normally being appended to a message,
511 only prepend for the Resent-* case. Fixes regression introduced in
512 Exim 4.70 by NM/22 for Bugzilla 607.
514 PP/02 Include check_rfc2047_length in configure.default because we're seeing
515 increasing numbers of administrators be bitten by this.
517 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
519 PP/03 Bugzilla 994: added openssl_options main configuration option.
521 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
523 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
525 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
527 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
528 without a peer certificate, leading to a segfault because of an
529 assumption that peers always have certificates. Be a little more
530 paranoid. Problem reported by Martin Tscholak.
532 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
533 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
534 NB: ClamAV planning to remove STREAM in "middle of 2010".
535 CL also introduces -bmalware, various -d+acl logging additions and
536 more caution in buffer sizes.
538 PP/09 Implemented reverse_ip expansion operator.
540 PP/10 Bugzilla 937: provide a "debug" ACL control.
542 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
544 PP/12 Bugzilla 973: Implement --version.
546 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
548 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
550 PP/15 Bugzilla 816: support multiple condition rules on Routers.
552 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
553 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
554 ignore trailing whitespace.
556 JJ/02 prevent non-panic DKIM error from being sent to paniclog
558 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
561 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
562 Notification from Dr Andrew Aitchison.
564 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
565 ExtendedDetectionInfo response format.
566 Notification from John Horne.
568 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
571 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
572 XSL and documented dependency on system catalogs, with examples of how
575 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
578 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
579 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
580 configuration file which is writeable by the Exim user or group.
582 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
583 of configuration files to cover files specified with the -C option if
584 they are going to be used with root privileges, not just the default
587 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
588 option (effectively making it always true).
590 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
591 files to be used while preserving root privileges.
593 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
594 that rogue child processes cannot use them.
596 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
597 run-time user, instead of root.
599 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
600 Exim run-time user without dropping privileges.
602 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
603 result string, instead of calling string_vformat() twice with the same
606 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
607 for other users. Others should always drop root privileges if they use
608 -C on the command line, even for a whitelisted configure file.
610 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
612 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
618 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
619 $data_path, and $header_path variables; fixed documentation bugs and
622 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
623 exipick to access non-standard spools, including the "frozen" queue
626 NM/01 Bugzilla 965: Support mysql stored procedures.
627 Patch from Alain Williams
629 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
631 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
632 Patch from Andreas Metzler
634 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
635 Patch from Kirill Miazine
637 NM/05 Bugzilla 671: Added umask to procmail example.
639 JJ/03 installed exipick 20100323.0, fixing doc bug
641 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
642 directory. Notification and patch from Dan Rosenberg.
644 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
646 TK/02 Improve log output when DKIM signing operation fails.
648 MH/01 Treat the transport option dkim_domain as a colon separated
649 list, not as a single string, and sign the message with each element,
650 omitting multiple occurences of the same signer.
652 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
653 Bugzilla 985, 986. Patch by Simon Arlott
655 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
656 Patch by Simon Arlott
658 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
659 MBX locking. Notification from Dan Rosenberg.
665 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
667 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
669 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
671 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
673 NM/04 Bugzilla 915: Flag broken perl installation during build.
679 TK/01 Added patch by Johannes Berg that expands the main option
680 "spamd_address" if it starts with a dollar sign.
682 TK/02 Write list of recipients to X-Envelope-Sender header when building
683 the mbox-format spool file for content scanning (suggested by Jakob
686 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
687 (http://www.dcc-servers.net/) support via dccifd. Activated by
688 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
690 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
691 by Mark Daniel Reidel <mr@df.eu>.
693 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
694 When building exim an external PCRE library is now needed -
695 PCRE is a system library on the majority of modern systems.
696 See entry on PCRE_LIBS in EDITME file.
698 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
699 conversation. Added nologin parameter to request.
700 Patch contributed by Kirill Miazine.
702 TF/01 Do not log submission mode rewrites if they do not change the address.
704 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
706 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
707 log files in place. Contributed by Roberto Lima.
709 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
711 TF/03 Bugzilla 615: When checking the local_parts router precondition
712 after a local_part_suffix or local_part_prefix option, Exim now
713 does not use the address's named list lookup cache, since this
714 contains cached lookups for the whole local part.
716 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
717 Robert Millan. Documentation is in experimental-spec.txt.
719 TF/04 Bugzilla 668: Fix parallel build (make -j).
721 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
723 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
724 Patch provided by Jan Srzednicki.
726 TF/05 Leading white space used to be stripped from $spam_report which
727 wrecked the formatting. Now it is preserved.
729 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
730 that they are available at delivery time.
732 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
734 TF/08 TLS error reporting now respects the incoming_interface and
735 incoming_port log selectors.
737 TF/09 Produce a more useful error message if an SMTP transport's hosts
738 setting expands to an empty string.
740 NM/06 Bugzilla 744: EXPN did not work under TLS.
741 Patch provided by Phil Pennock.
743 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
744 Patch provided by Richard Godbee.
746 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
747 acl_smtp_notquit, added index entry.
749 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
750 Patch provided by Eugene Bujak.
752 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
753 accept(). Patch provided by Maxim Dounin.
755 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
756 Patch provided by Phil Pennock.
758 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
760 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
761 Patch provided by Brad "anomie" Jorsch.
763 NM/14 Bugzilla 622: Added timeout setting to transport filter.
764 Patch provided by Dean Brooks.
766 TK/05 Add native DKIM support (does not depend on external libraries).
768 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
769 Patch provided by Graeme Fowler.
771 NM/16 Bugzilla 851: Documentation example syntax fix.
773 NM/17 Changed NOTICE file to remove references to embedded PCRE.
775 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
778 NM/19 Bugzilla 745: TLS version reporting.
779 Patch provided by Phil Pennock.
781 NM/20 Bugzilla 167: bool: condition support.
782 Patch provided by Phil Pennock.
784 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
785 clients. Patch provided by Phil Pennock.
787 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
788 Patch provided by Brad "anomie" Jorsch.
790 NM/23 Bugzilla 687: Fix misparses in eximstats.
791 Patch provided by Heiko Schlittermann.
793 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
794 Patch provided by Heiko Schlittermann.
796 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
797 plus update to original patch.
799 NM/26 Bugzilla 799: Documentation correction for ratelimit.
801 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
802 Patch provided by David Brownlee.
804 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
806 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
808 NM/30 Bugzilla 888: TLS documentation bugfixes.
810 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
812 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
813 Unlike the original bugzilla I have changed all shell scripts in src tree.
815 NM/33 Bugzilla 898: Transport filter timeout fix.
816 Patch by Todd Rinaldo.
818 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
819 Patch by Serge Demonchaux.
821 NM/35 Bugzilla 39: Base64 decode bug fixes.
822 Patch by Jakob Hirsch.
824 NM/36 Bugzilla 909: Correct connect() call in dcc code.
826 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
828 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
830 NM/39 Bugzilla 911: Fixed MakeLinks build script.
836 TK/01 Add preliminary DKIM support. Currently requires a forked version of
837 ALT-N's libdkim that I have put here:
838 http://duncanthrax.net/exim-experimental/
840 Note to Michael Haardt: I had to rename some vars in sieve.c. They
841 were called 'true' and it seems that C99 defines that as a reserved
842 keyword to be used with 'bool' variable types. That means you could
843 not include C99-style headers which use bools without triggering
844 build errors in sieve.c.
846 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
847 as mailq or other aliases. Changed the --help handling significantly
848 to do whats expected. exim_usage() emits usage/help information.
850 SC/01 Added the -bylocaldomain option to eximstats.
852 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
854 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
856 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
862 PH/01 Another patch from the Sieve maintainer.
864 PH/02 When an IPv6 address is converted to a string for single-key lookup
865 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
866 dots are used instead of colons so that keys in lsearch files need not
867 contain colons. This was done some time before quoting was made available
868 in lsearch files. However, iplsearch files do require colons in IPv6 keys
869 (notated using the quote facility) so as to distinguish them from IPv4
870 keys. This meant that lookups for IP addresses in host lists did not work
871 for iplsearch lookups.
873 This has been fixed by arranging for IPv6 addresses to be expressed with
874 colons if the lookup type is iplsearch. This is not incompatible, because
875 previously such lookups could never work.
877 The situation is now rather anomolous, since one *can* have colons in
878 ordinary lsearch keys. However, making the change in all cases is
879 incompatible and would probably break a number of configurations.
881 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
884 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
885 conversion specification without a maximum field width, thereby enabling
886 a rogue spamd server to cause a buffer overflow. While nobody in their
887 right mind would setup Exim to query an untrusted spamd server, an
888 attacker that gains access to a server running spamd could potentially
889 exploit this vulnerability to run arbitrary code as the Exim user.
891 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
892 $primary_hostname instead of what libspf2 thinks the hosts name is.
894 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
895 a directory entry by the name of the lookup key. Previously, if a
896 symlink pointed to a non-existing file or a file in a directory that
897 Exim lacked permissions to read, a lookup for a key matching that
898 symlink would fail. Now it is enough that a matching directory entry
899 exists, symlink or not. (Bugzilla 503.)
901 PH/03 The body_linecount and body_zerocount variables are now exported in the
904 PH/04 Added the $dnslist_matched variable.
906 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
907 This means they are set thereafter only if the connection becomes
910 PH/06 Added the client_condition to authenticators so that some can be skipped
911 by clients under certain conditions.
913 PH/07 The error message for a badly-placed control=no_multiline_responses left
914 "_responses" off the end of the name.
916 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
918 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
919 (without spaces) instead of just copying the configuration text.
921 PH/10 Added the /noupdate option to the ratelimit ACL condition.
923 PH/11 Added $max_received_linelength.
925 PH/12 Added +ignore_defer and +include_defer to host lists.
927 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
928 way in which PCRE > 7.0 is built.
930 PH/14 Implemented queue_only_load_latch.
932 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
933 MAIL command. The effect was to mangle the value on 64-bit systems.
935 PH/16 Another patch from the Sieve maintainer.
937 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
939 PH/18 If a system quota error occurred while trying to create the file for
940 a maildir delivery, the message "Mailbox is full" was not appended to the
941 bounce if the delivery eventually timed out. Change 4.67/27 below applied
942 only to a quota excession during the actual writing of the file.
944 PH/19 It seems that peer DN values may contain newlines (and other non-printing
945 characters?) which causes problems in log lines. The DN values are now
946 passed through string_printing() before being added to log lines.
948 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
949 and InterBase are left for another time.)
951 PH/21 Added message_body_newlines option.
953 PH/22 Guard against possible overflow in moan_check_errorcopy().
955 PH/23 POSIX allows open() to be a macro; guard against that.
957 PH/24 If the recipient of an error message contained an @ in the local part
958 (suitably quoted, of course), incorrect values were put in $domain and
959 $local_part during the evaluation of errors_copy.
965 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
966 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
969 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
970 issue a MAIL command.
972 PH/02 In an ACL statement such as
974 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
976 if a client was not listed at all, or was listed with a value other than
977 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
978 the condition was not true (as it should be), so access was not denied.
979 The bug was that the ! inversion was incorrectly passed on to the second
980 item. This has been fixed.
982 PH/03 Added additional dnslists conditions == and =& which are different from
983 = and & when the dns lookup returns more than one IP address.
985 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
986 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
988 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
989 FSYNC, which compiles an option called disable_fsync that allows for
990 bypassing fsync(). The documentation is heavily laced with warnings.
992 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
994 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
995 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
996 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
997 including adding "make clean"; (3) Added -fPIC when compiling the test
998 dynamically loaded module, to get rid of a warning.
1000 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1001 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1002 = 0s. The bug is otherwise harmless.
1004 PH/07 There was a bug in the dovecot authenticator such that the value of
1005 $auth1 could be overwritten, and so not correctly preserved, after a
1006 successful authentication. This usually meant that the value preserved by
1007 the server_setid option was incorrect.
1009 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1011 PH/09 Installed PCRE release 7.0.
1013 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1014 run for batched SMTP input. It is now run at the start of every message
1015 in the batch. While fixing this I discovered that the process information
1016 (output by running exiwhat) was not always getting set for -bs and -bS
1017 input. This is fixed, and it now also says "batched" for BSMTP.
1019 PH/11 Added control=no_pipelining.
1021 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1022 patch, slightly modified), and move the expansion of helo_data till after
1023 the connection is made in the smtp transport (so it can use these
1026 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1028 PH/14 Added log_selector = +pid.
1030 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1032 PH/16 Add ${if forany and ${if forall.
1034 PH/17 Added dsn_from option to vary the From: line in DSNs.
1036 PH/18 Flush SMTP output before performing a callout, unless control =
1037 no_callout_flush is set.
1039 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1040 was true (the default) a successful delivery failed to delete the retry
1041 item, thus causing premature timeout of the address. The bug is now
1044 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1046 PH/21 Long custom messages for fakedefer and fakereject are now split up
1047 into multiline reponses in the same way that messages for "deny" and
1048 other ACL rejections are.
1050 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1051 with slight modification.
1053 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1054 draft, changing the syntax and factoring some duplicate code".
1056 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1057 for deliveries of the second and subsequent messages over the same SMTP
1060 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1061 ${reduce, with only minor "tidies".
1063 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1065 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1066 expansion side effects.
1068 PH/27 When a message times out after an over-quota error from an Exim-imposed
1069 quota, the bounce message says "mailbox is full". This message was not
1070 being given when it was a system quota that was exceeded. It now should
1073 MH/03 Made $recipients available in local_scan(). local_scan() already has
1074 better access to the recipient list through recipients_list[], but
1075 $recipients can be useful in postmaster-provided expansion strings.
1077 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1078 in the case of a MAIL command with additional options following the
1079 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1080 were accidentally chopped off.
1082 PH/29 SMTP synchronization checks are implemented when a command is read -
1083 there is a check that no more input is waiting when there shouldn't be
1084 any. However, for some commands, a delay in an ACL can mean that it is
1085 some time before the response is written. In this time, more input might
1086 arrive, invalidly. So now there are extra checks after an ACL has run for
1087 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1088 pipelining has not been advertised.
1090 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1092 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1093 returned in the SMTP response when smtp_return_error_details was false.
1094 This has been fixed.
1096 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1097 instead of the C I/O that was originally supplied, because problems were
1098 reported on Solaris.
1100 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1101 Exim which did not show up earlier: it was assuming that a call to
1102 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1103 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1104 was a macro that became an assignment, so it seemed to work. This has
1105 changed to a proper function call with a void return, hence the compile
1106 error. Exim's code has been fixed.
1108 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1111 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1113 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1115 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1116 because it turns out to be too big or there is a timeout) there is no
1117 "Completed" line in the log. When some messages of this type were
1118 selected by exigrep, they were listed as "not completed". Others were
1119 picked up by some special patterns. I have improved the selection
1120 criteria to be more general.
1122 PH/38 The host_find_failed option in the manualroute router can now be set
1123 to "ignore", to completely ignore a host whose IP address cannot be
1124 found. If all hosts are ignored, the behaviour is controlled by the new
1125 host_all_ignored option.
1127 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1128 homing or because of multiple MX records with /mx) generated more than
1129 one IP address, and the following item turned out to be the local host,
1130 all the secondary addresses of the first item were incorrectly removed
1131 from the list, along with the local host and any following hosts (which
1132 is what is supposed to happen).
1134 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1135 whoever called Exim into the -H file. In the case of the daemon it was
1136 behaving confusingly. When first started, it used values for whoever
1137 started the daemon, but after a SIGHUP it used the Exim user (because it
1138 calls itself on a restart). I have changed the code so that it now always
1141 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1142 message are rejected with the same error (e.g. no authentication or bad
1143 sender address), and a DATA command is nevertheless sent (as can happen
1144 with PIPELINING or a stupid MUA), the error message that was given to the
1145 RCPT commands is included in the rejection of the DATA command. This is
1146 intended to be helpful for MUAs that show only the final error to their
1149 PH/42 Another patch from the Sieve maintainer.
1151 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1152 Eximstats - Fixed some broken HTML links and added missing column headers
1154 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1155 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1157 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1159 PH/43 Yet another patch from the Sieve maintainer.
1161 PH/44 I found a way to check for a TCP/IP connection going away before sending
1162 the response to the final '.' that terminates a message, but only in the
1163 case where the client has not sent further data following the '.'
1164 (unfortunately, this is allowed). However, in many cases there won't be
1165 any further data because there won't be any more messages to send. A call
1166 to select() can be used: if it shows that the input is "ready", there is
1167 either input waiting, or the socket has been closed. An attempt to read
1168 the next input character can distinguish the two cases. Previously, Exim
1169 would have sent an OK response which the client would never have see.
1170 This could lead to message repetition. This fix should cure that, at
1171 least in a lot of common cases.
1173 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1174 advertised in response to EHLO.
1180 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1181 fixed by 4.65/MH/01 (is this a record?) are fixed:
1183 (i) An empty string was always treated as zero by the numeric comparison
1184 operators. This behaviour has been restored.
1186 (ii) It is documented that the numeric comparison operators always treat
1187 their arguments as decimal numbers. This was broken in that numbers
1188 starting with 0 were being interpreted as octal.
1190 While fixing these problems I realized that there was another issue that
1191 hadn't been noticed. Values of message_size_limit (both the global option
1192 and the transport option) were treated as octal if they started with 0.
1193 The documentation was vague. These values are now always treated as
1194 decimal, and I will make that clear in the documentation.
1200 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1201 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1204 MH/01 Don't check that the operands of numeric comparison operators are
1205 integers when their expansion is in "skipping" mode (fixes bug
1206 introduced by 4.64-PH/07).
1208 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1209 child addresses, Exim now panics and dies. Previously, because the count
1210 is held in a short int, deliveries were likely to be lost. As such a
1211 large number of recipients for a single message is ridiculous
1212 (performance will be very, very poor), I have chosen to impose a limit
1213 rather than extend the field.
1219 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1220 leftover -K file (the existence of which was triggered by #402).
1221 While we were at it, introduced process PID as part of the -K
1222 filename. This should rule out race conditions when creating
1225 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1226 processing considerably. Previous code took too long for large mails,
1227 triggering a timeout which in turn triggers #401.
1229 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1230 in the DK code in transports.c. sendfile() is not really portable,
1231 hence the _LINUX specificness.
1233 TF/01 In the add_headers option to the mail command in an Exim filter,
1234 there was a bug that Exim would claim a syntax error in any
1235 header after the first one which had an odd number of characters
1238 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1239 callout verification, Exim cached a "reject" for the entire domain. This
1240 is correct for most verifications, but it is not correct for a recipient
1241 verification with use_sender or use_postmaster set, because in that case
1242 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1243 case of MAIL FROM:<> rejection from other early rejections (e.g.
1244 rejection of HELO). When verifying a recipient using a non-null MAIL
1245 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1246 Whatever the result of the callout, the value of the domain cache is
1247 left unchanged (for any other kind of callout, getting as far as trying
1248 RCPT means that the domain itself is ok).
1250 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1253 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1254 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1255 session, the daemon ignores ECONNECT errors and logs others; it now
1256 ignores EPIPE as well.
1258 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1259 (quoted-printable decoding).
1261 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1262 later the small subsequent patch to fix an introduced bug.
1264 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1266 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1268 PH/08 An error is now given if message_size_limit is specified negative.
1270 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1271 to be given (somewhat) arbitrary names.
1273 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1276 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1277 miscellaneous code fixes
1279 PH/10 Added the log_reject_target ACL modifier to specify where to log
1282 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1283 hostname. This is wrong, because it relates to the incoming message (and
1284 probably the interface on which it is arriving) and not to the outgoing
1285 callout (which could be using a different interface). This has been
1286 changed to use the value of the helo_data option from the smtp transport
1287 instead - this is what is used when a message is actually being sent. If
1288 there is no remote transport (possible with a router that sets up host
1289 addresses), $smtp_active_hostname is used.
1291 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1292 tweaks were necessary in order to get it to work (see also 21 below):
1293 (a) The code assumed that strncpy() returns a negative number on buffer
1294 overflow, which isn't the case. Replaced with Exim's string_format()
1296 (b) There were several signed/unsigned issues. I just did the minimum
1297 hacking in of casts. There is scope for a larger refactoring.
1298 (c) The code used strcasecmp() which is not a standard C function.
1299 Replaced with Exim's strcmpic() function.
1300 (d) The code set only $1; it now sets $auth1 as well.
1301 (e) A simple test gave the error "authentication client didn't specify
1302 service in request". It would seem that Dovecot has changed its
1303 interface. Fortunately there's a specification; I followed it and
1304 changed what the client sends and it appears to be working now.
1306 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1309 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1310 address A is aliased to B and C, where B exists and C does not. Without
1311 -v the output is "A verified" because verification stops after a
1312 successful redirection if more than one address is generated. However,
1313 with -v the child addresses are also verified. Exim was outputting "A
1314 failed to verify" and then showing the successful verification for C,
1315 with its parentage. It now outputs "B failed to verify", showing B's
1316 parentage before showing the successful verification of C.
1318 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1319 look up a TXT record in a specific list after matching in a combined
1322 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1323 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1324 they consult the DNS. I had assumed they would set it the way they
1325 wanted; and indeed my experiments on Linux seem to show that in some
1326 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1327 To be on the safe side, however, I have now made the interface to
1328 host_find_byname() similar to host_find_bydns(), with an argument
1329 containing the DNS resolver options. The host_find_byname() function now
1330 sets these options at its start, just as host_find_bydns() does. The smtp
1331 transport options dns_qualify_single and dns_search_parents are passed to
1332 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1333 of host_find_byname() use the default settings of RES_DEFNAMES
1334 (qualify_single) but not RES_DNSRCH (search_parents).
1336 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1337 spool_read_header() do less string testing, by means of a preliminary
1338 switch on the second character of optional "-foo" lines. (This is
1339 overdue, caused by the large number of possibilities that now exist.
1340 Originally there were few.) While I was there, I also converted the
1341 str(n)cmp tests so they don't re-test the leading "-" and the first
1342 character, in the hope this might squeeze out yet more improvement.
1344 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1345 flag allowing group syntax was set by the header_syntax check but not
1346 turned off, possible causing trouble later; (2) The flag was not being
1347 set at all for the header_verify test, causing "group"-style headers to
1348 be rejected. I have now set it in this case, and also caused header_
1349 verify to ignore an empty address taken from a group. While doing this, I
1350 came across some other cases where the code for allowing group syntax
1351 while scanning a header line wasn't quite right (mostly, not resetting
1352 the flag correctly in the right place). These bugs could have caused
1353 trouble for malformed header lines. I hope it is now all correct.
1355 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1356 with the "reply" argument non-NULL. The code, however (which originally
1357 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1358 but it didn't always do it. This confused somebody who was copying the
1359 code for some other use. I have removed all the tests.
1361 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1362 feature that was used to support insecure browsers during the U.S. crypto
1363 embargo. It requires special client support, and Exim is probably the
1364 only MTA that supported it -- and would never use it because real RSA is
1365 always available. This code has been removed, because it had the bad
1366 effect of slowing Exim down by computing (never used) parameters for the
1367 RSA_EXPORT functionality.
1369 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1370 authenticator to fail if there's a tab character in the incoming data
1371 (there should never be unless someone is messing about, as it's supposed
1372 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1373 if the connection is using TLS or if the remote IP is the same as the
1374 local IP, and the "valid-client-cert option" if a client certificate has
1377 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1378 authenticators. This can be used for authorization after authentication
1379 succeeds. (In the case of plaintext, it servers for both authentication
1382 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1383 if any retry times were supplied.
1385 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1386 connection, where there is no client IP address to check. In this
1387 situation, the verify now always succeeds.
1389 PH/25 Applied John Jetmore's -Mset patch.
1391 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1393 PH/27 In a string expansion for a processed (not raw) header when multiple
1394 headers of the same name were present, leading whitespace was being
1395 removed from all of them, but trailing whitespace was being removed only
1396 from the last one. Now trailing whitespace is removed from each header
1397 before concatenation. Completely empty headers in a concatenation (as
1398 before) are ignored.
1400 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1401 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1403 PH/29 [Removed. This was a change that I later backed out, and forgot to
1404 correct the ChangeLog entry (that I had efficiently created) before
1405 committing the later change.]
1407 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1408 address errors (4xx response to RCPT) over the same connection as other
1409 messages routed to the same hosts. Such deliveries are always "forced",
1410 so retry times are not inspected. This resulted in far too many retries
1411 for the affected addresses. The effect occurred only when there were more
1412 hosts than the hosts_max_try setting in the smtp transport when it had
1413 the 4xx errors. Those hosts that it had tried were not added to the list
1414 of hosts for which the message was waiting, so if all were tried, there
1415 was no problem. Two fixes have been applied:
1417 (i) If there are any address or message errors in an SMTP delivery, none
1418 of the hosts (tried or untried) are now added to the list of hosts
1419 for which the message is waiting, so the message should not be a
1420 candidate for sending over the same connection that was used for a
1421 successful delivery of some other message. This seems entirely
1422 reasonable: after all the message is NOT "waiting for some host".
1423 This is so "obvious" that I'm not sure why it wasn't done
1424 previously. Hope I haven't missed anything, but it can't do any
1425 harm, as the worst effect is to miss an optimization.
1427 (ii) If, despite (i), such a delivery is accidentally attempted, the
1428 routing retry time is respected, so at least it doesn't keep
1429 hammering the server.
1431 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1432 in ${readsocket because some servers need this prod.
1434 PH/32 Added some extra debug output when updating a wait-xxx database.
1436 PH/33 The hint "could be header name not terminated by colon", which has been
1437 given for certain expansion errors for a long time, was not being given
1438 for the ${if def:h_colon_omitted{... case.
1440 PH/34 The spec says: "With one important exception, whenever a domain list is
1441 being scanned, $domain contains the subject domain." There was at least
1442 one case where this was not true.
1444 PH/35 The error "getsockname() failed: connection reset by peer" was being
1445 written to the panic log as well as the main log, but it isn't really
1446 panic-worthy as it just means the connection died rather early on. I have
1447 removed the panic log writing for the ECONNRESET error when getsockname()
1450 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1451 runs only) independently of the message's sender address. This meant
1452 that, if the 4xx error was in fact related to the sender, a different
1453 message to the same recipient with a different sender could confuse
1454 things. In particualar, this can happen when sending to a greylisting
1455 server, but other circumstances could also provoke similar problems.
1456 I have changed the default so that the retry time for these errors is now
1457 based a combination of the sender and recipient addresses. This change
1458 can be overridden by setting address_retry_include_sender=false in the
1461 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1462 remote server are returned as part of bounce messages. This was not
1463 happening for LMTP over a pipe (the lmtp transport), but now it is the
1464 same for both kinds of LMTP.
1466 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1467 in header lines that were in fact CNAMEs. This is no longer the case.
1469 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1470 and queue runs started by the daemon processed all messages. This has
1471 been fixed so that -R and -S can now usefully be given with -q<time>.
1473 PH/40 Import PCRE release 6.7 (fixes some bugs).
1475 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1477 PH/42 Give an error if -q is specified more than once.
1479 PH/43 Renamed the variables $interface_address and $interface_port as
1480 $received_ip_address and $received_port, to make it clear that these
1481 values apply to message reception, and not to the outgoing interface when
1482 a message is delivered. (The old names remain recognized, of course.)
1484 PH/44 There was no timeout on the connect() call when using a Unix domain
1485 socket in the ${readsocket expansion. There now is.
1487 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1488 be meaningful with "accept".
1490 SC/01 Eximstats V1.43
1491 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1493 SC/02 Eximstats V1.44
1494 Use a glob alias rather than an array ref in the generated
1495 parser. This improves both readability and performance.
1497 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1498 Collect SpamAssassin and rejection statistics.
1499 Don't display local sender or destination tables unless
1500 there is data to show.
1501 Added average volumes into the top table text output.
1503 SC/04 Eximstats V1.46
1504 Collect data on the number of addresses (recipients)
1505 as well as the number of messages.
1507 SC/05 Eximstats V1.47
1508 Added 'Message too big' to the list of mail rejection
1509 reasons (thanks to Marco Gaiarin).
1511 SC/06 Eximstats V1.48
1512 Mainlog lines which have GMT offsets and are too short to
1513 have a flag are now skipped.
1515 SC/07 Eximstats V1.49 (Alain Williams)
1516 Added the -emptyok flag.
1518 SC/08 Eximstats V1.50
1519 Fixes for obtaining the IP address from reject messages.
1521 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1522 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1523 whitesspace changes from 4.64-PH/27
1525 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1528 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1529 are found, allow negative numbers in numeric criteria)
1531 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1533 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1534 to match changes made in 4.64-PH/43
1536 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1538 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1539 "domain" is actually an IP address, and force a failure. This locks out
1540 those revolvers/nameservers that support "A-for-A" lookups, in
1541 contravention of the specifications.
1543 PH/48 When a host name was looked up from an IP address, and the subsequent
1544 forward lookup of the name timed out, the host name was left in
1545 $sender_host_name, contrary to the specification.
1547 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1548 restricted to single-key lookups, Exim was not diagnosing an error if
1549 * or *@ was used with a query-style lookup.
1551 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1553 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1554 long ago, but noone interested enough thought of it. Let's just say that
1555 the "1.1" means that there are some new functions that weren't there at
1556 some point in the past.
1558 PH/51 Error processing for expansion failure of helo_data from an smtp
1559 transport during callout processing was broken.
1561 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1562 tested/used via the -bh/-bhc/-bs options.
1564 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1565 bug, fixed in subsequent PCRE releases).
1567 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1568 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1570 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1576 SC/01 Use a glob alias rather than an array ref in eximstats generated
1577 parser. This improves both readability and performance.
1579 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1580 Don't display local sender or destination tables in eximstats unless
1581 there is data to show.
1582 Added average volumes into the eximstats top table text output.
1584 SC/03 Collect data on the number of addresses (recipients) as well
1585 as the number of messages in eximstats.
1587 TF/01 Correct an error in the documentation for the redirect router. Exim
1588 does (usually) call initgroups() when daemonizing.
1590 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1591 with consistent privilege compared to when running as a daemon.
1593 TF/03 Note in the spec that $authenticated_id is not set for local
1594 submissions from trusted users.
1596 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1597 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1599 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1600 by adding some example configuration directives to the default
1601 configuration file. A little bit of work is required to uncomment the
1602 directives and define how usernames and passwords are checked, but
1603 there is now a framework to start from.
1605 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1606 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1607 without this. I don't know how relevant this is to other LDAP libraries.
1609 PH/02 Add the verb name to the "unknown ACL verb" error.
1611 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1613 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1615 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1616 directory not expanded when it should be if an expanded home directory
1617 was set for the address (which is overridden by the transport).
1619 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1622 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1623 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1624 because it is too late at that time, and has no effect.
1626 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1627 security issue with \' (bugzilla #107). I could not use the
1628 PQescapeStringConn() function, because it needs a PGconn value as one of
1631 PH/08 When testing addresses using -bt, indicate those final addresses that
1632 are duplicates that would not cause an additional delivery. At least one
1633 person was confused, thinking that -bt output corresponded to deliveries.
1634 (Suppressing duplicates isn't a good idea as you lose the information
1635 about possibly different redirections that led to the duplicates.)
1637 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1638 systems where poll() doesn't work, in particular OS X.
1640 PH/10 Added more information to debugging output for retry time not reached.
1642 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1643 operations in malware.c.
1645 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1648 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1649 syslog_duplication set false, log lines that would normally be written
1650 both the the main log and to the reject log were not written to syslog at
1653 PH/14 In the default configuration, change the use of "message" in ACL warn
1654 statements to "add_header".
1656 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1657 not followed by a command (e.g. "seen endif").
1659 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1660 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1663 PH/17 Added extra conditions to the default value of delay_warning_condition
1667 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1668 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1669 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1672 The Auto-Submitted: and various List- headers are standardised, whereas I
1673 don't think Precedence: ever was.
1675 PH/18 Refactored debugging code in route_finduser() to show more information,
1676 in particular, the error code if getpwnam() issues one.
1678 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1679 This is apparently needed in addition to the PH/07 change above to avoid
1680 any possible encoding problems.
1682 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1683 but not after initializing Perl.
1685 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1686 output them only if debugging. By default they are written stderr,
1687 apparently, which is not desirable.
1689 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1692 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1695 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1697 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1698 authenticated or an ident call has been made. Suppress the default
1699 values for $authenticated_id and $authenticated_sender (but permit -oMai
1700 and -oMas) when testing with -bh.
1702 PH/24 Re-jigged the order of the tests in the default configuration so that the
1703 tests for valid domains and recipients precede the DNS black list and CSA
1704 tests, on the grounds that those ones are more expensive.
1706 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1707 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1708 This bug exists in every version of Exim that I still have, right back to
1711 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1712 However, an attempt to turn on case-sensitivity in a regex key by
1713 including (?-i) didn't work because the subject string was already
1714 lowercased, and the effects were non-intuitive. It turns out that a
1715 one-line patch can be used to allow (?-i) to work as expected.
1721 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1722 other effects) broke the use of negated acl sub-conditions.
1724 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1727 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1728 "Deny" causes Exim to reject the incoming connection with a 554 error.
1729 Unfortunately, if there is a major crisis, such as a disk failure,
1730 tcp-wrappers gives "deny", whereas what one would like would be some
1731 kind of temporary error. A kludge has been added to help with this.
1732 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1733 554 error is used if errno is still zero or contains ENOENT (which occurs
1734 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1737 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1739 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1740 errors. Otherwise a message that provokes a temporary error (when other
1741 messages do not) can cause a whole host to time out.
1743 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1744 addresses were routed directly to files or pipes from a redirect router.
1745 File deliveries just didn't batch; pipe deliveries might have suffered
1748 PH/06 A failure to get a lock for a hints database would erroneously always say
1749 "Failed to get write lock", even when it was really a read lock.
1751 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1752 of 0600. This has been changed to use the value of the lockfile_mode
1753 option (which defaults to 0600).
1755 PH/08 Applied small patch from the Sieve maintainer.
1757 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1758 folder from quota calculations, a direct delivery into this folder messed
1759 up the contents of the maildirsize file. This was because the regex was
1760 used only to exclude .Trash (or whatever) when the size of the mailbox
1761 was calculated. There was no check that a delivery was happening into an
1762 excluded directory. This bug has been fixed by ignoring all quota
1763 processing for deliveries into excluded directories.
1765 PH/10 Added the maildirfolder_create_regex option to appendfile.
1771 PH/01 The code for finding all the local interface addresses on a FreeBSD
1772 system running IPv6 was broken. This may well have applied to all BSD
1773 systems, as well as to others that have similar system calls. The broken
1774 code found IPv4 interfaces correctly, but gave incorrect values for the
1775 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1776 that it would not match correctly against @[] and not recognize the IPv6
1779 PH/02 The ipliteral router was not recognizing addresses of the form user@
1780 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1782 PH/03 Added disable_ipv6.
1784 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1785 decoded form, because it is most often used to construct To: headers
1786 lines in autoreplies, and the decoded form may well be syntactically
1787 invalid. However, $reply_address has leading white space removed, and all
1788 newlines turned into spaces so that the autoreply transport does not
1791 PH/05 If group was specified without a user on a router, and no group or user
1792 was specified on a transport, the group from the router was ignored.
1794 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1795 for visible compile-time settings that can be used to change these
1796 numbers, for those that want even more. Backwards compatibility with old
1797 spool files has been maintained. However, going back to a previous Exim
1798 release will lost any variables that are in spool files.
1800 PH/07 Two small changes when running in the test harness: increase delay when
1801 passing a TCP/IP connection to a new process, in case the original
1802 process has to generate a bounce, and remove special handling of
1803 127.0.0.2 (sic), which is no longer necessary.
1805 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1806 be the same on different OS.
1808 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1811 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1812 whether --show-vars was specified or not
1814 JJ/02 exipick: Added support for new ACL variable spool format introduced
1817 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1818 syntactically invalid From: or Reply-to: line, and a filter used this to
1819 generate an autoreply, and therefore failed to obtain an address for the
1820 autoreply, Exim could try to deliver to a non-existent relative file
1821 name, causing unrelated and misleading errors. What now happens is that
1822 it logs this as a hard delivery error, but does not attempt to create a
1825 PH/11 The exinext utility has a -C option for testing purposes, but although
1826 the given file was scanned by exinext itself; it wasn't being passed on
1827 when Exim was called.
1829 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1830 an end-of-file indication when reading a command response.
1832 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1833 compiled. In many other places in Exim, IPv6 addresses are always
1834 recognized, so I have changed this. It also means that IPv4 domain
1835 literals of the form [IPV4:n.n.n.n] are now always recognized.
1837 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1838 used if the router is not running as root, for example, when verifying at
1839 ACL time, or when using -bh. The debugging output from this situation was
1840 non-existent - all you got was a failure to exec. I have made two
1843 (a) Failures to set uid/gid, the current directory, or a process leader
1844 in a subprocess such as that created by queryprogram now generate
1845 suitable debugging ouput when -d is set.
1847 (b) The queryprogram router detects when it is not running as root,
1848 outputs suitable debugging information if -d is set, and then runs
1849 the subprocess without attempting to change uid/gid.
1851 PH/15 Minor change to Makefile for building test_host (undocumented testing
1854 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1855 additional section of a DNS packet that returns MX or SRV records.
1856 Instead, it always explicitly searches for A/AAAA records. This avoids
1857 major problems that occur when a DNS server includes only records of one
1858 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1859 fixed another bug: if SRV records were looked up and the corresponding
1860 address records were *not* found in the additional section, the port
1861 values from the SRV records were lost.
1863 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1864 using the correct key (the original address) when searching the retry
1865 rules in order to find which one to use for generating the retry hint.
1867 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1868 adding the default one. Similarly, if it contains a Reply-To: header, the
1869 errors_reply_to option, if set, is not used.
1871 PH/19 When calculating a retry time, Exim used to measure the "time since
1872 failure" by looking at the "first failed" field in the retry record. Now
1873 it does not use this if it is later than than the arrival time of the
1874 message. Instead it uses the arrival time. This makes for better
1875 behaviour in cases where some deliveries succeed, thus re-setting the
1876 "first failed" field. An example is a quota failure for a huge message
1877 when small messages continue to be delivered. Without this change, the
1878 "time since failure" will always be short, possible causing more frequent
1879 delivery attempts for the huge message than are intended.
1880 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1882 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1883 $1, $2, $3) because the numerical variables can be reset during some
1884 expansion items (e.g. "match"), thereby losing the authentication data.
1886 PH/21 Make -bV show the size of off_t variables so that the test suite can
1887 decide whether to run tests for quotas > 2G.
1889 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1890 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1891 filecount value is greater than 2G or if a quota value is greater than 2G
1892 on a system where the size of off_t is not greater than 4, a panic error
1895 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1896 never match. The debug and -bh output now contains an explicit error
1897 message indicating a malformed IPv4 address or mask.
1899 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1900 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1901 PH/23 above applies.
1903 PH/25 Do not write to syslog when running in the test harness. The only
1904 occasion when this arises is a failure to open the main or panic logs
1905 (for which there is an explicit test).
1907 PH/26 Added the /no_tell option to "control=freeze".
1909 PH/27 If a host name lookup failed very early in a connection, for example, if
1910 the IP address matched host_lookup and the reverse lookup yielded a name
1911 that did not have a forward lookup, an error message of the form "no IP
1912 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1913 could be logged. Now it outputs the IP address instead of "NULL".
1915 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1916 allows the sender and the authenticated sender to be set when
1917 submitting a message from within Exim. Since child_open_exim() is
1918 documented for local_scan(), the new function should be too.
1920 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1921 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1922 results in an empty string is now treated as unset.
1924 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1926 PH/31 Added sender_verify_fail logging option.
1928 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1929 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1930 tidied the source and removed it altogether.
1932 PH/33 When a queue run was abandoned because the load average was too high, a
1933 log line was always written; now it is written only if the queue_run log
1934 selector is set. In addition, the log line for abandonment now contains
1935 information about the queue run such as the pid. This is always present
1936 in "start" and "stop" lines but was omitted from the "abandon" line.
1938 PH/34 Omit spaces between a header name and the colon in the error message that
1939 is given when verify = headers_syntax fails (if there are lots of them,
1940 the message gets confusing).
1942 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1943 names, as there are now some PTR records that contain slashes. This check
1944 is only to protect against broken name servers that fall over on strange
1945 characters, so the fact that it applies to all lookups doesn't matter.
1947 PH/36 Now that the new test suite is complete, we can remove some of the
1948 special code in Exim that was needed for the old test suite. For example,
1949 sorting DNS records because real resolvers return them in an arbitrary
1950 order. The new test suite's fake resolver always returns records in the
1953 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1954 bounces) except when queue_only is set, to avoid logging races between
1955 the different processes.
1957 PH/38 Panic-die if .include specifies a non-absolute path.
1959 PH/39 A tweak to the "H" retry rule from its user.
1961 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1962 a label. They prevented compilation on older perls.
1964 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1965 a warning to be raised on newish perls.
1967 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1968 on queue. Changes to match documented behaviour of showing count of
1969 messages matching specified criteria.
1971 PH/40 Changed the default ident timeout from 30s to 5s.
1973 PH/41 Added support for the use of login_cap features, on those BSD systems
1974 that have them, for controlling the resources used by pipe deliveries.
1976 PH/42 The content-scanning code uses fopen() to create files in which to put
1977 message data. Previously it was not paying any attention to the mode of
1978 the files. Exim runs with umask(0) because the rest of the code creates
1979 files with open(), and sets the required mode explicitly. Thus, these
1980 files were ending up world-writeable. This was not a big issue, because,
1981 being within the spool directory, they were not world-accessible. I have
1982 created a function called modefopen, which takes an additional mode
1983 argument. It sets umask(777), creates the file, chmods it to the required
1984 mode, then resets the umask. All the relevant calls to fopen() in the
1985 content scanning code have been changed to use this function.
1987 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1988 to 24 hours. This avoids potential overflow problems when processing G
1989 and H retry rules. I suspect nobody ever tinkers with this value.
1991 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1993 PH/45 When the plaintext authenticator is running as a client, the server's
1994 challenges are checked to ensure they are valid base64 strings. By
1995 default, the authentication attempt is cancelled if an invalid string is
1996 received. Setting client_ignore_invalid_base64 true ignores these errors.
1997 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1998 they are received. Thus, the responses can be made to depend on the
1999 challenges. If an invalid string is ignored, an empty string is placed in
2002 PH/46 Messages that are created by the autoreply transport now contains a
2003 References: header, in accordance with RFCs 2822 and 3834.
2005 PH/47 Added authenticated_sender_force to the smtp transport.
2007 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2009 PH/49 Installed latest patch from the Sieve maintainer.
2011 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2012 was also set, the appendfile transport was unnecessarily scanning a
2013 directory of message files (e.g. for maildir delivery) to find the count
2014 of files (along with the size), even though it did not need this
2015 information. It now does the scan only if it needs to find either the
2016 size of the count of files.
2018 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2020 PH/52 Two bugs concerned with error handling when the smtp transport is
2023 (i) Exim was not creating retry information for temporary errors given
2024 for individual recipients after the DATA command when the smtp transport
2025 was used in LMTP mode. This meant that they could be retried too
2026 frequently, and not timed out correctly.
2028 (ii) Exim was setting the flag that allows error details to be returned
2029 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2030 recipients that were returned after the DATA command.
2032 PH/53 This is related to PH/52, but is more general: for any failing address,
2033 when detailed error information was permitted to be returned to the
2034 sender, but the error was temporary, then after the final timeout, only
2035 "retry timeout exceeded" was returned. Now it returns the full error as
2036 well as "retry timeout exceeded".
2038 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2039 do this, and (what is worse) MTAs that accept it.
2041 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2042 will now be deprecated.
2044 PH/56 New os.c-cygwin from the Cygwin maintainer.
2046 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2047 formats (previously only available in exim formats via -bpr, -bpru,
2048 and -bpra. Now also available in native and exiqgrep formats)
2050 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2051 with very large, slow to parse queues
2053 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2055 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2057 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2058 responses to authentication challenges, though it was showing the
2059 challenges; (ii) I've removed the CR characters from the debug output for
2062 PH/58 Allow for the insertion of a newline as well as a space when a string
2063 is turned into more than one encoded-word during RFC 2047 encoding. The
2064 Sieve code now uses this.
2066 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2067 data_4xx, lost_connection, tls_required.
2069 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2070 message was being sent as an SMTP response.
2072 PH/61 Add -l and -k options to exicyclog.
2074 PH/62 When verifying, if an address was redirected to one new address, so that
2075 verification continued, and the new address failed or deferred after
2076 having set something in $address_data, the value of $address_data was not
2077 passed back to the ACL. This was different to the case when no
2078 redirection occurred. The value is now passed back in both cases.
2080 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2081 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2082 use login_cap.h, so on its own it isn't the distinguishing feature. The
2083 new name refers directly to the setclassresources() function.
2085 PH/65 Added configuration files for NetBSD3.
2087 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2089 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2090 is preferred over IPv4.
2092 PH/68 The bounce_return_message and bounce_return_body options were not being
2093 honoured for bounces generated during the reception of non-SMTP messages.
2094 In particular, this applied to messages rejected by the ACL. This bug has
2095 been fixed. However, if bounce_return_message is true and bounce_return_
2096 body is false, the headers that are returned for a non-SMTP message
2097 include only those that have been read before the error was detected.
2098 (In the case of an ACL rejection, they have all been read.)
2100 PH/69 The HTML version of the specification is now built in a directory called
2101 spec_html instead of spec.html, because the latter looks like a path with
2102 a MIME-type, and this confuses some software.
2104 PH/70 Catch two compiler warnings in sieve.c.
2106 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2107 function verify_get_ident() calls ip_connect() to connect a socket, but
2108 if the "connect()" function timed out, ip_connect() used to close the
2109 socket. However, verify_get_ident() also closes the socket later, and in
2110 between Exim writes to the log, which may get opened at this point. When
2111 the socket was closed in ip_connect(), the log could get the same file
2112 descriptor number as the socket. This naturally causes chaos. The fix is
2113 not to close the socket in ip_connect(); the socket should be closed by
2114 the function that creates it. There was only one place in the code where
2115 this was missing, in the iplookup router, which I don't think anybody now
2116 uses, but I've fixed it anyway.
2118 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2119 well as to direct DNS lookups. Otherwise the handling of names in host
2120 lists is inconsistent and therefore confusing.
2126 PH/01 Two changes to the default runtime configuration:
2128 (1) Move the checks for relay_from_hosts and authenticated clients from
2129 after to before the (commented out) DNS black list checks.
2131 (2) Add control=submission to the relay_from_hosts and authenticated
2132 clients checks, on the grounds that messages accepted by these
2133 statements are most likely to be submissions.
2135 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2137 (1) Generate an error if the third argument for the ${prvs expansion is
2140 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2143 (3) Reset the variables that are obtained from the first argument of
2144 ${prvscheck and used in the second argument before leaving the code,
2145 because their memory is reclaimed, so using them afterwards may do
2148 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2149 one (it's much easier than Tom thought :-).
2151 (5) Because of (4), we can now allow for the use of $prvscheck_result
2152 inside the third argument.
2154 PH/03 For some reason, the default setting of PATH when running a command from
2155 a pipe transport was just "/usr/bin". I have changed it to
2158 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2159 anything to be listed in the output from -bV.
2161 PH/05 When a filter generated an autoreply, the entire To: header line was
2162 quoted in the delivery log line, like this:
2164 => >A.N.Other <ano@some.domain> <original@ddress> ...
2166 This has been changed so that it extracts the operative address. There
2167 may be more than one such address. If so, they are comma-separated, like
2170 => >ano@some.domain,ona@other.domain <original@ddress> ...
2172 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2173 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2174 not being looked up in the rDNS to get a host name, Exim was showing the
2175 IP address twice in Received: lines, even though the IP addresses were
2176 identical. For example:
2178 Received: from [1.2.3.4] (helo=[1.2.3.4])
2180 However, if the real host name was known, it was omitting the HELO data
2181 if it matched the actual IP address. This has been tidied up so that it
2182 doesn't show the same IP address twice.
2184 PH/07 When both +timestamp and +memory debugging was on, the value given by
2185 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2186 called by the debug printing, thereby overwriting the timestamp buffer.
2187 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2189 PH/08 When the original message was included in an autoreply transport, it
2190 always said "this is a copy of the message, including all the headers",
2191 even if body_only or headers_only was set. It now gives an appropriate
2194 PH/09 Applied a patch from the Sieve maintainer which:
2196 o fixes some comments
2197 o adds the (disabled) notify extension core
2198 o adds some debug output for the result of if/elsif tests
2199 o points to the current vacation draft in the documentation
2200 and documents the missing references header update
2204 o fixes a bug in processing the envelope test (when testing
2205 multiple envelope elements, the last element determinted the
2208 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2209 Electronic Mail") by including:
2211 Auto-submitted: auto-generated
2213 in the messages that it generates (bounce messages and others, such as
2214 warnings). In the case of bounce messages for non-SMTP mesages, there was
2215 also a typo: it was using "Auto_submitted" (underscore instead of
2216 hyphen). Since every message generated by Exim is necessarily in response
2217 to another message, thes have all been changed to:
2219 Auto-Submitted: auto-replied
2221 in accordance with these statements in the RFC:
2223 The auto-replied keyword:
2225 - SHOULD be used on messages sent in direct response to another
2226 message by an automatic process,
2228 - MUST NOT be used on manually-generated messages,
2230 - MAY be used on Delivery Status Notifications (DSNs) and Message
2231 Disposition Notifications (MDNs),
2233 - MUST NOT be used on messages generated by automatic or periodic
2234 processes, except for messages which are automatic responses to
2237 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2238 to the default Received: header definition.
2240 PH/12 Added log selector acl_warn_skipped (default on).
2242 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2243 variables because (a) they are in the wrong storage pool and (b) even if
2244 they were copied, it wouldn't work properly because of the caching.
2246 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2247 checking when decoding. Apparently there are clients that generate
2248 overlong encoded strings. Why am I not surprised?
2250 PH/15 If the first argument of "${if match_address" was not empty, but did not
2251 contain an "@" character, Exim crashed. Now it writes a panic log message
2252 and treats the condition as false.
2254 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2256 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2257 extension "envelope-auth". The code is finished and in agreement with
2258 other implementations, but there is no documentation so far and in fact,
2259 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2260 not changing the active code.
2262 Print executed "if" and "elsif" statements when debugging is used. This
2263 helps a great deal to understand what a filter does.
2265 Document more things not specified clearly in RFC3028. I had all this
2266 sorted out, when out of a sudden new issues came to my mind. Oops."
2268 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2271 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2272 canonical form (as documented). However, after a host name lookup from
2273 the IP address, check_host() was doing a simple string comparison with
2274 addresses acquired from the DNS when checking that the found name did
2275 have the original IP as one of its addresses. Since any found IPv6
2276 addresses are likely to be in abbreviated form, the comparison could
2277 fail. Luckily, there already exists a function for doing the comparison
2278 by converting both addresses to binary, so now that is used instead of
2279 the text comparison.
2281 PH/20 There was another similar case to PH/19, when a complete host name was
2282 given in a host list; looking up its IP address could give an abbreviated
2283 form, whereas the current host's name might or might not be abbreviated.
2284 The same fix has been applied.
2290 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2291 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2294 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2295 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2297 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2299 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2300 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2301 Also, for "Auto-Submitted", treat anything other than "no" as
2302 non-personal, in accordance with RFC 3834. (Previously it treated
2303 anything starting "auto-" as non-personal.)
2305 TF/01 The control=submission/name=... option had a problem with syntax
2306 errors if the name included a slash character. The /name= option
2307 now slurps the rest of the string, so it can include any characters
2308 but it must come last in the list of options (after /sender_retain
2311 PH/05 Some modifications to the interface to the fake nameserver for the new
2319 TK/01 Added the "success_on_redirect" address verification option. See
2320 NewStuff for rationale and an example.
2322 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2324 PH/02 Patch to exigrep to allow it to work on syslog lines.
2326 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2327 fread() to skip over the body file's header line, because in Cygwin the
2328 header line is locked and is inaccessible.
2330 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2331 co-exist for some time) to make it clear that it is the Exim ID that is
2332 referenced, not the Message-ID: header line.
2334 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2335 string_format() function, because snprintf() does not exist on all
2338 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2339 ${stat: expansion item.
2341 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2342 protocol synchronization error", to keep the pedants happy.
2344 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2345 well as for IRIX systems, when gcc is being used. See the host.c source
2348 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2350 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2353 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2354 transport and to the smtp transport in LMTP mode.
2356 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2358 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2359 run a filter in a subprocess. This could lead to confusion in subsequent
2360 lookups in the parent process. There should also be a search_tidyup() at
2361 the end of the subprocess.
2363 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2364 only if the host matched helo_try_verify_hosts, which caused the
2365 verification to occur when the EHLO/HELO command was issued. The ACL just
2366 tested the remembered result. Now, if a previous verification attempt has
2367 not happened, "verify = helo" does it there and then.
2369 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2371 TK/03 Fix log output including CR from clamd.
2373 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2374 exist provoked a memory error which could cause a segfault.
2376 PH/15 Installed PCRE 6.2
2378 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2380 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2381 of the problem. Specifically, suggested +O2 rather than +O1 for the
2384 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2386 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2387 record, the port was not correctly logged when the outgoing_port log
2388 selector was set (it logged the transort's default port).
2390 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2391 fallback_hosts, and "hosts" in the smtp transport.
2393 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2394 host errors such as "Connection refused".
2396 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2397 authentication with radiusclient 0.4.9:
2399 - Error returned from rc_read_config was caught wrongly
2400 - Username/password not passed on to radius server due to wrong length.
2402 The presumption is that some radiusclient API changes for 4.51/PH/17
2403 were not taken care of correctly. The code is still untested by me (my
2404 Linux distribution still has 0.3.2 of radiusclient), but it was
2405 contributed by a Radius user.
2407 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2408 expanding the "port" option of the smtp transport.
2410 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2411 while reading a MIME header. Thanks to Tom Hughes for a patch.
2413 PH/24 Include config.h inside local_scan.h so that configuration settings are
2416 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2417 that in an ACL for RCPT (for example), you can examine exactly what was
2420 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2421 commands, but it was not correctly comparing the address with the actual
2422 client host address. Thus, it would show the EHLO address in Received:
2423 header lines when this was not necessary.
2425 PH/27 Added the % operator to ${eval:}.
2427 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2428 it should be ignoring failures (because with -C, for example, it has lost
2429 privilege). It wasn't ignoring creation failures other than "already
2432 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2435 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2436 because an input line was too long, either on its own, or by virtue of
2437 too many continuations, the temporary file was not being removed, and the
2438 return code was incorrect.
2440 PH/31 Missing "BOOL" in function definition in filtertest.c.
2442 PH/32 Applied Sieve patches from the maintainer.
2444 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2446 PH/33 Added "verify = not_blind".
2448 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2449 Local/Makefile (with some defaults set). These are used in built scripts
2450 such as exicyclog, but they have never been used in the exim_install
2451 script (though there are many overriding facilities there). I have
2452 arranged that the exim_install script now takes note of these two
2455 PH/35 Installed configuration files for Dragonfly.
2457 PH/36 When a locally submitted message by a trusted user did not contain a
2458 From: header, and the sender address was obtained from -f or from an SMTP
2459 MAIL command, and the trusted user did not use -F to supply a sender
2460 name, $originator_name was incorrectly used when constructing a From:
2461 header. Furthermore, $originator_name was used for submission mode
2462 messages from external hosts without From: headers in a similar way,
2463 which is clearly wrong.
2465 PH/37 Added control=suppress_local_fixups.
2467 PH/38 When log_selector = +received_sender was set, and the addition of the
2468 sender made the log line's construction buffer exactly full, or one byte
2469 less than full, an overflow happened when the terminating "\n" was
2472 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2473 when the result of a list match is failure because a DNS lookup failed.
2475 PH/40 RM_COMMAND is now used in the building process.
2477 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2478 the "build-* directories that it finds.
2480 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2481 address in a domain literal was a prefix of an interface address.
2483 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2484 when verifying a sender address, unless rewrite_headers is false.
2486 PH/44 Wrote a long comment about why errors_to addresses are verified as
2487 recipients, not senders.
2489 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2490 the ratelimit ACL was added.
2492 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2494 PH/46 Added extra information about PostgreSQL errors to the error string.
2496 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2497 avoiding the need to install special zones in a real server. This is
2498 backwards compatible; if it can't find the fake resolver, it drops back.
2499 Thus, both old and new test suites can be run.
2501 TF/02 Added util/ratelimit.pl
2503 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2504 clock is set back in time.
2506 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2507 Candler <B.Candler@pobox.com>.
2509 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2510 OK for addresses that are the result of redirections.
2512 PH/48 A number of further additions for the benefit of the new test suite,
2513 including a fake gethostbyname() that interfaces to the fake DNS resolver
2516 TF/06 The fix for widen_domains has also been applied to qualify_single and
2517 search_parents which are the other dnslookup options that can cause
2520 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2523 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2525 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2526 service" if it encountered an MX record with an empty target hostname.
2527 The message is now "an MX or SRV record indicated no SMTP service".
2529 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2530 if the DNS of the sending site is misconfigured. This is quite a
2531 common situation. This change restores the behaviour of treating a
2532 helo verification defer as a failure.
2534 PH/51 If self=fail was set on a router, the bounce message did not include the
2535 actual error message.
2541 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2543 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2544 command itself ended in error, the underlying message about the transport
2545 filter timeout was being overwritten with the pipe command error. Now the
2546 underlying error message should be appended to the second error message.
2548 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2550 PH/02 Reduce the amount of output that "make" produces by default. Full output
2551 can still be requested.
2553 PH/03 The warning log line about a condition test deferring for a "warn" verb
2554 was being output only once per connection, rather than after each
2555 occurrence (because it was using the same function as for successful
2556 "warn" verbs). This seems wrong, so I have changed it.
2558 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2559 it should not have, which might have caused a crash in the right
2560 circumstances, but probably never did.
2562 PH/04 Installed a modified version of Tony Finch's patch to make submission
2563 mode fix the return path as well as the Sender: header line, and to
2564 add a /name= option so that you can make the user's friendly name appear
2567 TF/03 Added the control = fakedefer ACL modifier.
2569 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2570 Mark Lowes for thorough testing.
2572 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2574 TK/03 Merged latest SRS patch from Miles Wilton.
2576 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2577 with the definition in sysexits.h (which is #included earlier).
2578 Fortunately, Exim does not actually use EX_OK. The code used to try to
2579 preserve the sysexits.h value, by assumimg that macro definitions were
2580 scanned for macro replacements. I have been disabused of this notion,
2581 so now the code just undefines EX_OK before #including unistd.h.
2583 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2584 in the smtp transport. When a block could not be written in a single
2585 write() function, the timeout was being re-applied to each part-write.
2586 This seems wrong - if the receiver was accepting one byte at a time it
2587 would take for ever. The timeout is now adjusted when this happens. It
2588 doesn't have to be particularly precise.
2590 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2591 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2593 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2594 without a domain if the check to <postmaster@domain> fails.
2596 SC/01 Eximstats: added -xls and the ability to specify output files
2597 (patch written by Frank Heydlauf).
2599 SC/02 Eximstats: use FileHandles for outputing results.
2601 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2603 SC/04 Eximstats: fixed display of large numbers with -nvr option
2605 SC/05 Eximstats: fixed merging of reports with empty tables.
2607 SC/06 Eximstats: added the -include_original_destination flag
2609 SC/07 Eximstats: removed tabs and trailing whitespace.
2611 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2613 TK/06 MBOX spool code: Add real "From " MBOX separator line
2614 so the .eml file is really in mbox format (even though
2615 most programs do not really care). Patch from Alex Miller.
2617 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2618 The latter is generated from $received_to and is only set if the
2619 message has one envelope recipient. SA can use these headers,
2620 obviously out-of-the-box. Patch from Alex Miller.
2622 PH/08 The ${def test on a variable was returning false if the variable's
2623 value was "0", contrary to what the specification has always said!
2624 The result should be true unless the variable is empty.
2626 PH/09 The syntax error of a character other than { following "${if
2627 def:variable_name" (after optional whitespace) was not being diagnosed.
2628 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2629 accidental colon was present, for example, could give incorrect results.
2631 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2632 result is used (not including appendfile, where other changes are about
2635 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2636 This involved changing a lot of size variables from int to off_t. It
2637 should work with maildirs and everything.
2639 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2640 spamd dying while we are connected to it.
2642 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2645 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2646 function for BDB 4.3.
2648 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2650 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2651 the sources for PCRE have been re-arranged and more files are now
2654 PH/15 The code I had for printing potentially long long variables in PH/11
2655 above was not the best (it lost precision). The length of off_t variables
2656 is now inspected at build time, and an appropriate printing format (%ld
2657 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2658 to be "long long int" or "long int". This is needed for the internal
2659 formatting function string_vformat().
2661 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2662 the configuration file to be ":syslog", then the script "guesses" where
2663 the logs files are, rather than using the compiled in default. In our
2664 case the guess is not the same as the compiled default, so the script
2665 suddenly stopped working when I started to use syslog. The patch checks
2666 to see if log_file_path is "". If so, it attempts to read it from exim
2667 with no configuration file to get the compiled in version, before it
2668 falls back to the previous guessing code."
2670 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2671 implementing BATV in an Exim configuration. See NewStuff for the gory
2674 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2675 Makefile that are specific to HP-UX.
2677 PH/18 If the "use_postmaster" option was set for a recipient callout together
2678 with the "random" option, the postmaster address was used as the MAIL
2679 FROM address for the random test, but not for the subsequent recipient
2680 test. It is now used for both.
2682 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2683 patch removes a few documentation additions to RFC 3028, because the
2684 latest draft now contains them. It adds the new en;ascii-case comparator
2685 and a new error check for 8bit text in MIME parts. Comparator and
2686 require names are now matched exactly. I enabled the subaddress
2687 extension, but it is not well tested yet (read: it works for me)."
2689 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2690 rework some of the code of TK/09 above to avoid the hardwired use of
2691 "%lld" and "long long". Replaced the call to snprintf() with a call to
2694 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2695 records point to non-existent hosts", "retry timeout exceeded", and
2696 "retry time not reached for any host after a long failure period".
2698 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2699 experimental DomainKeys support:
2701 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2702 (2) On an error such as an illegally used "control", the wrong name for
2703 the control was given.
2705 These problems did NOT occur unless DomainKeys support was compiled.
2707 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2709 PH/24 Added ${if match_ip condition.
2711 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2712 databases so that it will be absolutely obvious if a crash occurs in the
2713 DB library. This is a regular occurrence (often caused by mis-matched
2716 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2717 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2718 picked up on a user's system that detects such things. There doesn't seem
2719 to be a gcc warning option for this - only an attribute that has to be
2720 put on the function's prototype. It seems that in Fedora Core 4 they have
2721 set this on a number of new functions. No doubt there will be more in due
2724 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2725 specify a transport. However, if an address that was verified by such a
2726 router was the subject of a callout, Exim crashed because it tried to
2727 read the rcpt_include_affixes from the non-existent transport. Now it
2728 just assumes that the setting of that option is false. This bug was
2729 introduced by 4.51/PH/31.
2731 PH/28 Changed -d+all to exclude +memory, because that information is very
2732 rarely of interest, but it makes the output a lot bigger. People tend to
2733 do -d+all out of habit.
2735 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2736 code in os-type was giving problems when libc.so lives in lib64, like on
2739 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2740 aren't the modern standard, and it seems that some systems' include files
2741 don't always have them. Exim was already checking for some of the newer
2742 ones like T_AAAA, and defining it itself. I've added checks for all the
2743 record types that Exim uses.
2745 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2746 not automatically generating a new one, as it is supposed to. This
2747 prevented TLS from working. If the file did exist, but contained invalid
2748 data, a new version was generated, as expected. It was only the case of a
2749 non-existent file that was broken.
2751 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2752 with a change in libdomainkeys > 0.64.
2754 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2755 from DNS. If the selector record carries the flag, it now has
2756 precedence over the domain-wide flag.
2758 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2760 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2761 the use of a port name that isn't defined in /etc/services) occurred, the
2762 message was deferred as in a normal delivery, and thus remained on the
2763 spool, instead of being failed because of the mua_wrapper setting. This
2764 is now fixed, and I tidied up some of the mua_wrapper messages at the
2767 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2768 the messages in a hash of arrays rather than using individual hashes.
2769 This is a bit cleaner and results in dramatic memory savings, albeit
2770 at a slight CPU cost.
2772 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2773 as requested by Marc Sherman.
2775 SC/10 Eximstats: added histograms for user specified patterns as requested
2778 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2780 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2781 fopen() in the content-scanning modules that did not already have it.
2787 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2788 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2790 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2792 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2794 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2795 to test Sieve filters that use "vacation".
2797 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2798 that changes the way the GnuTLS parameters are stored in the cache file.
2799 The new format can be generated externally. For backward compatibility,
2800 if the data in the cache doesn't make sense, Exim assumes it has read an
2801 old-format file, and it generates new data and writes a new file. This
2802 means that you can't go back to an older release without removing the
2805 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2806 work if there are any delivery delays because "one_time" forces the
2807 parent to be marked "delivered", so its unseen clone is never tried
2808 again. For this reason, Exim now forbids the simultaneous setting of
2811 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2812 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2813 entry if you want to know the details. The fix, however, neglected to
2814 consider the case when local delivery batching is involved. The test for
2815 "previously delivered" was not happening when checking to see if an
2816 address could be batched with a previous (undelivered) one; under
2817 certain circumstances this could lead to multiple deliveries to the same
2820 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2821 in its include files, and this causes problems building Exim.
2823 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2824 header_syntax) but Exim was just ignoring anything given after a slash.
2825 In particular, this caused confusion with an attempt to use "verify =
2826 reverse_host_lookup/defer_ok". An error is now given when options are
2827 supplied for verify items that do not have them. (Maybe reverse_host_
2828 lookup should have a defer_ok option, but that's a different point.)
2830 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2831 defined by RFC 821) to 2048, because there were problems with some AUTH
2832 commands, and RFC 1869 says the size should be increased for extended
2833 SMTP commands that take arguments.
2835 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2838 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2839 "unknown" error; now it says that the functionality isn't in the binary.
2841 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2842 an address' error message when a string expansion fails (syntax or
2843 whatever). Otherwise the password may appear in the log. Following change
2844 PH/42 below, there is no longer a chance of it appearing in a bounce
2847 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2849 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2850 of its addresses was ever tried. (Bugzilla bug #2.)
2852 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2853 the result incorrectly in the debug output. (It correctly added a newline
2854 to what was transported.)
2856 TF/01 Added $received_time.
2858 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2859 commented out examples of how to interface to a virus scanner and to
2860 SpamAssassin. Also added commented examples of av_scanner and
2861 spamd_address settings.
2863 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2864 and controls are allowed in which ACLs. There were a couple of minor
2865 errors. Some of the entries in the conditions table (which is a table of
2866 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2867 negation of where the condition IS allowed.
2869 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2871 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2872 header file does not have a version number, so I've had to invent a new
2873 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2874 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2875 radiusclient), but it was contributed by a Radius user.
2877 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2878 files or directories, for OpenSSL.
2880 PH/19 When an Exim process that is running as root has to create an Exim log
2881 file, it does so in a subprocess that runs as exim:exim so as to get the
2882 ownership right at creation (otherwise, other Exim processes might see
2883 the file with the wrong ownership). There was no test for failure of this
2884 fork() call, which would lead to the process getting stuck as it waited
2885 for a non-existent subprocess. Forks do occasionally fail when resources
2886 run out. I reviewed all the other calls to fork(); they all seem to check
2889 PH/20 When checking for unexpected SMTP input at connect time (before writing
2890 the banner), Exim was not dealing correctly with a non-positive return
2891 from the read() function. If the client had disconnected by this time,
2892 the result was a log entry for a synchronization error with an empty
2893 string after "input=" when read() returned zero. If read() returned -1
2894 (an event I could not check), uninitialized data bytes were printed.
2895 There were reports of junk text (parts of files, etc) appearing after
2898 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2900 PH/22 Added support for macro redefinition, and (re)definition in between
2901 driver and ACL definitions.
2903 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2904 forgetting to use the resulting value; it was using the unexpanded value.
2906 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2907 hadn't been configured. The fix is from Juergen Kreileder, who
2908 understands it better than I do:
2910 "Here's what I see happening with three configured cyrus_sasl
2911 authenticators configured (plain, login, cram-md5):
2913 On startup auth_cyrus_sasl_init() gets called for each of these.
2914 This means three calls to sasl_listmech() without a specified mech_list.
2915 => SASL tests which mechs of all available mechs actually work
2916 => three warnings about OTP not working
2917 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2919 With the patch, sasl_listmech() also gets called three times. But now
2920 SASL's mech_list option is set to the server_mech specified in the the
2921 authenticator. Or in other words, the answer from sasl_listmech()
2922 gets limited to just the mech you're testing for (which is different
2924 => the return list contains just 'plain' or 'login', 'cram-md5' or
2925 nothing depending on the value of ob->server_mech.
2927 I've just tested the patch: Authentication still works fine,
2928 unavailable mechs specified in the exim configuration are still
2929 caught, and the auth.log warnings about OTP are gone."
2931 PH/25 When debugging is enabled, the contents of the command line are added
2932 to the debugging output, even when log_selector=+arguments is not
2935 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2936 answer is "GNU", and only if the return is "GNU/something" is the answer
2939 PH/27 $acl_verify_message is now set immediately after the failure of a
2940 verification in an ACL, and so is available in subsequent modifiers. In
2941 particular, the message can be preserved by coding like this:
2943 warn !verify = sender
2944 set acl_m0 = $acl_verify_message
2946 Previously, $acl_verify_message was set only while expanding "message"
2947 and "log_message" when a very denied access.
2949 PH/28 Modified OS/os.c-Linux with
2951 -#ifndef OS_LOAD_AVERAGE
2952 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2954 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2955 nomenclature these days.)
2957 PH/29 Installed patch from the Sieve maintainer that adds the options
2958 sieve_useraddress and sieve_subaddress to the redirect router.
2960 PH/30 In these circumstances:
2961 . Two addresses routed to the same list of hosts;
2962 . First host does not offer TLS;
2963 . First host accepts first address;
2964 . First host gives temporary error to second address;
2965 . Second host offers TLS and a TLS session is established;
2966 . Second host accepts second address.
2967 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2968 and peerdn, if requested) that were in fact used only for the second
2971 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2972 attention to any local part prefix or suffix that was matched by the
2973 router that accepted the address. It now behaves in the same way as it
2974 does for delivery: the affixes are removed from the local part unless
2975 rcpt_include_affixes is set on the transport.
2977 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2978 timeout during the DATA phase of an incoming message.
2980 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2981 applied a patch sanctioned by the Sieve maintainer.
2983 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2984 the uid or gid is negative. A case of a negative gid caused this to be
2985 noticed. The fix allows for either to be negative.
2987 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2988 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2991 PH/36 The change PH/12 above was broken. Fixed it.
2993 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2994 the grounds that routing the same address twice would always produce the
2995 same answer. This might have been true once, but it is certainly no
2996 longer true now. Routing a child address may depend on the previous
2997 routing that produced that child. Some complicated redirection strategies
2998 went wrong when messages had multiple recipients, and made Exim's
2999 behaviour dependent on the order in which the addresses were given.
3001 I have moved the duplicate checking until after the routing is complete.
3002 Exim scans the addresses that are assigned to local and remote
3003 transports, and removes any duplicates. This means that more work will be
3004 done, as duplicates will always all be routed, but duplicates are
3005 presumably rare, so I don't expect this is of any significance.
3007 For deliveries to pipes, files, and autoreplies, the duplicate checking
3008 still happens during the routing process, since they are not going to be
3011 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3012 It corrects a timeout issue with spamd. This is Ian's comment: "The
3013 background is that sometimes spamd either never reads data from a
3014 connection it has accepted, or it never writes response data. The exiscan
3015 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3016 blindly assumes that writes won't block so it may never time out."
3018 PH/39 Allow G after quota size as well as K and M.
3020 PH/40 The value set for $authenticated_id in an authenticator may not contain
3021 binary zeroes or newlines because the value is written to log lines and
3022 to spool files. There was no check on this. Now the value is run through
3023 the string_printing() function so that such characters are converted to
3024 printable escape sequences.
3026 PH/41 $message_linecount is a new variable that contains the total number of
3027 lines in the message. Compare $body_linecount, which is the count for the
3030 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3031 bounce and delay warning messages, except in certain special cases, which
3034 (a) An SMTP error message from a remote host;
3035 (b) A message specified in a :fail: redirection;
3036 (c) A message specified in a "fail" command in a system filter;
3037 (d) A message specified in a FAIL return from the queryprogram router;
3038 (e) A message specified by the cannot_route_message router option.
3040 In these cases only, Exim does include the error details in bounce and
3041 warning messages. There are also a few cases where bland messages such
3042 as "unrouteable address" or "local delivery error" are given.
3044 PH/43 $value is now also set for the "else" part of a ${run expansion.
3046 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3047 being worked on, but at least Exim now implements the latest version to
3050 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3051 process to complete was treated as a delivery failure, a timeout while
3052 writing the message to the pipe was logged, but erroneously treated as a
3053 successful delivery. Such timeouts include transport filter timeouts. For
3054 consistency with the overall process timeout, these timeouts are now
3055 treated as errors, giving rise to delivery failures by default. However,
3056 there is now a new Boolean option for the pipe transport called
3057 timeout_defer, which, if set TRUE, converts the failures into defers for
3058 both kinds of timeout. A transport filter timeout is now identified in
3061 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3062 systems where "make" and "gmake" are different, calling "gmake" at top
3063 level broke things. I've arranged for the value of $(MAKE) to be passed
3064 from the Makefile to this script so that it can call the same version of
3068 A note about Exim versions 4.44 and 4.50
3069 ----------------------------------------
3071 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3072 changes of various kinds. As a consequence, a big documentation update was
3073 needed. This delayed the release for rather longer than seemed good, especially
3074 in the light of a couple of (minor) security issues. Therefore, the changes
3075 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3076 release. So 4.44 and 4.50 are in effect two different branches that both start
3079 I have left the 4.50 change log unchanged; it contains all the changes since
3080 4.43. The change log for 4.44 is below; many of its items are identical to
3081 those for 4.50. This seems to be the most sensible way to preserve the
3082 historical information.
3088 1. Minor wording change to the doc/README.SIEVE file.
3090 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3091 computation of the current number of files was incorrect.
3093 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3094 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3097 4. Give more explanation in the error message when the command for a transport
3098 filter fails to execute.
3100 5. There are several places where Exim runs a non-Exim command in a
3101 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3102 was being done only for the command run by the queryprogram router. It is
3103 now done for all such subprocesses. The other cases are: ${run, transport
3104 filters, and the commands run by the lmtp and pipe transports.
3106 6. Added CONFIGURE_GROUP build-time option.
3108 7. Some older OS have a limit of 256 on the maximum number of file
3109 descriptors. Exim was using setrlimit() to set 1000 as a large value
3110 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3111 systems. I've change it so that if it can't get 1000, it tries for 256.
3113 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3114 was an oversight, and furthermore, ever since the addition of extra
3115 controls (e.g. 4.43/32), the checks on when to allow different forms of
3116 "control" were broken. There should now be diagnostics for all cases when a
3117 control that does not make sense is encountered.
3119 9. Added the /retain_sender option to "control=submission".
3121 10. $recipients is now available in the predata ACL (oversight).
3123 11. Tidy the search cache before the fork to do a delivery from a message
3124 received from the command line. Otherwise the child will trigger a lookup
3125 failure and thereby defer the delivery if it tries to use (for example) a
3126 cached ldap connection that the parent has called unbind on.
3128 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3129 of $address_data from the recipient verification was clobbered by the
3130 sender verification.
3132 13. The value of address_data from a sender verification is now available in
3133 $sender_address_data in subsequent conditions in the ACL statement.
3135 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3137 15. Added a new option "connect=<time>" to callout options, to set a different
3140 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3141 was its contents. (It was OK if the option was not defined at all.)
3143 17. A "Completed" log line is now written for messages that are removed from
3144 the spool by the -Mrm option.
3146 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3147 information about exactly what failed.
3149 19. Added -dd to debug only the daemon process.
3151 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3152 handles timeouts, both on the server side and network timeouts. Renamed the
3153 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3155 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3156 It is now set to "smtps".
3158 22. $host_address is now set to the target address during the checking of
3159 ignore_target_hosts.
3161 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3162 being passed; this would have caused $sender_host_name to have been used if
3163 matching the list had actually called for a host name (not very likely,
3164 since this list is usually IP addresses). A host name is now passed as
3167 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3168 code that specifies a non-restarting handler (typically sigaction() in
3169 modern systems) in an attempt to fix a rare and obscure crash bug.
3171 25. Narrowed the window for a race in the daemon that could cause it to ignore
3172 SIGCHLD signals. This is not a major problem, because they are used only to
3173 wake it up if nothing else does.
3175 26. A malformed maildirsize file could cause Exim to calculate negative values
3176 for the mailbox size or file count. Odd effects could occur as a result.
3177 The maildirsize information is now recalculated if the size or filecount
3180 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3181 support for a long time. Removed HAVE_SYS_VFS_H.
3183 28. Installed the latest version of exipick from John Jetmore.
3185 29. In an address list, if the pattern was not a regular expression, an empty
3186 subject address (from a bounce message) matched only if the pattern was an
3187 empty string. Non-empty patterns were not even tested. This was the wrong
3188 because it is perfectly reasonable to use an empty address as part of a
3189 database query. An empty address is now tested by patterns that are
3190 lookups. However, all the other forms of pattern expect the subject to
3191 contain a local part and a domain, and therefore, for them, an empty
3192 address still always fails if the pattern is not itself empty.
3194 30. Exim went into a mad DNS loop when attempting to do a callout where the
3195 host was specified on an smtp transport, and looking it up yielded more
3196 than one IP address.
3198 31. Re-factored the code for checking spool and log partition space into a
3199 function that finds that data and another that does the check. The former
3200 is then used to implement four new variables: $spool_space, $log_space,
3201 $spool_inodes, and $log_inodes.
3203 32. The RFC2047 encoding function was originally intended for short strings
3204 such as real names; it was not keeping to the 75-character limit for
3205 encoded words that the RFC imposes. It now respects the limit, and
3206 generates multiple encoded words if necessary. To be on the safe side, I
3207 have increased the buffer size for the ${rfc2047: expansion operator from
3210 33. It is now permitted to omit both strings after an "if" condition; if the
3211 condition is true, the result is "true". As before, when the second string
3212 is omitted, a false condition yields an empty string. This makes it less
3213 cumbersome to write custom ACL and router conditions.
3215 34. Failure to deliver a bounce message always caused it to be frozen, even if
3216 there was an errors_to setting on the router. The errors_to setting is now
3219 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3220 canonical form (fully expanded) before being placed in
3221 $sender_host_address.
3223 36. The table in the code that translates DNS record types into text (T_A to
3224 "A" for instance) was missing entries for NS and CNAME. It is just possible
3225 that this could have caused confusion if both these types were looked up
3226 for the same domain, because the text type is used as part of Exim's
3227 per-process caching. But the chance of anyone hitting this buglet seems
3230 37. The dnsdb lookup has been extended in a number of ways.
3232 (1) There is a new type, "zns", which walks up the domain tree until it
3233 finds some nameserver records. It should be used with care.
3235 (2) There is a new type, "mxh", which is like "mx" except that it returns
3236 just the host names, not the priorities.
3238 (3) It is now possible to give a list of domains (or IP addresses) to be
3239 looked up. The behaviour when one of the lookups defers can be
3240 controlled by a keyword.
3242 (4) It is now possible to specify the separator character for use when
3243 multiple records are returned.
3245 38. The dnslists ACL condition has been extended: it is now possible to supply
3246 a list of IP addresses and/or domains to be looked up in a particular DNS
3249 39. Added log_selector=+queue_time_overall.
3251 40. When running the queue in the test harness, wait just a tad after forking a
3252 delivery process, to get repeatability of debugging output.
3254 41. Include certificate and key file names in error message when GnuTLS fails
3255 to set them up, because the GnuTLS error message doesn't include the name
3256 of the failing file when there is a problem reading it.
3258 42. Allow both -bf and -bF in the same test run.
3260 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3262 44. The "Exiscan patch" is now merged into the mainline Exim source.
3264 45. Sometimes the final signoff response after QUIT could fail to get
3265 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3266 < 0 before doing a fflush(). This bug looks as though it goes back to the
3267 introduction of TLS in release 3.20, but "sometimes" must have been rare
3268 because the tests only now provoked it.
3270 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3271 (this can affect the format of dates).
3273 47. exim_tidydb, when checking for the continued existence of a message for
3274 which it has found a message-specific retry record, was not finding
3275 messages that were in split spool directories. Consequently, it was
3276 deleting retry records that should have stayed in existence.
3278 48. Steve fixed some bugs in eximstats.
3280 49. The SPA authentication driver was not abandoning authentication and moving
3281 on to the next authenticator when an expansion was forced to fail,
3282 contradicting the general specification for all authenticators. Instead it
3283 was generating a temporary error. It now behaves as specified.
3285 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3286 (the order specifies the preference for clients). The order is now AES256,
3287 AES128, 3DES, ARCFOUR128.
3289 51. Small patch to Sieve code - explicitly set From: when generating an
3292 52. Exim crashed if a remote delivery caused a very long error message to be
3293 recorded - for instance if somebody sent an entire SpamAssassin report back
3294 as a large number of 550 error lines. This bug was coincidentally fixed by
3295 increasing the size of one of Exim's internal buffers (big_buffer) that
3296 happened as part of the Exiscan merge. However, to be on the safe side, I
3297 have made the code more robust (and fixed the comments that describe what
3300 53. Now that there can be additional text after "Completed" in log lines (if
3301 the queue_time_overall log selector is set), a one-byte patch to exigrep
3302 was needed to allow it to recognize "Completed" as not the last thing in
3305 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3306 patch that reportedly fixes this has been added. I am not expert enough to
3307 create a test for it. This is what the patch creator wrote:
3309 "I found a little strange behaviour of ldap code when working with
3310 Windows 2003 AD Domain, where users was placed in more than one
3311 Organization Units. When I tried to give exim partial DN, the exit code
3312 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3313 But simultaneously result of request was absolutely normal ldap result,
3314 so I produce this patch..."
3316 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3317 REFERENCE, so I have modified the code to exclude the patch when that macro
3320 55. Some experimental protocols are using DNS PTR records for new purposes. The
3321 keys for these records are domain names, not reversed IP addresses. The
3322 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3323 leaves it alone. Component reversal etc. now happens only for IP addresses.
3326 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3328 57. Double the size of the debug message buffer (to 2048) so that more of very
3329 long debug lines gets shown.
3331 58. The exicyclog utility now does better if the number of log files to keep
3332 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3334 59. Two changes related to the smtp_active_hostname option:
3336 (1) $smtp_active_hostname is now available as a variable.
3337 (2) The default for smtp_banner uses $smtp_active_hostname instead
3338 of $primary_hostname.
3340 60. The host_aton() function is supposed to be passed a string that is known
3341 to be a valid IP address. However, in the case of IPv6 addresses, it was
3342 not checking this. This is a hostage to fortune. Exim now panics and dies
3343 if the condition is not met. A case was found where this could be provoked
3344 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3345 components; fortuitously, this particular loophole had already been fixed
3346 by change 4.50/55 above.
3348 If there are any other similar loopholes, the new check in host_aton()
3349 itself should stop them being exploited. The report I received stated that
3350 data on the command line could provoke the exploit when Exim was running as
3351 exim, but did not say which command line option was involved. All I could
3352 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3353 running as the user.
3356 61. There was a buffer overflow vulnerability in the SPA authentication code
3357 (which came originally from the Samba project). I have added a test to the
3358 spa_base64_to_bits() function which I hope fixes it.
3361 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3362 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3364 63. The daemon start-up calls getloadavg() while still root for those OS that
3365 need the first call to be done as root, but it missed one case: when
3366 deliver_queue_load_max is set with deliver_drop_privilege. This is
3367 necessary for the benefit of the queue runner, because there is no re-exec
3368 when deliver_drop_privilege is set.
3370 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3371 This has been fixed.
3373 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3374 was in use, was not putting the data itself into the right store pool;
3375 consequently, it could be overwritten for a subsequent message in the same
3376 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3379 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3381 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3382 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3383 the function were treating the return as a boolean value, which happened to
3384 work because 0=false and not-0=true, but is not correct code.
3386 68. The host_aton() function was not handling scoped IPv6 addresses (those
3387 with, for example, "%eth0" on the end) correctly.
3389 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3390 negated items (that is, ~something) in unsigned ints. Some compilers
3391 apparently mutter when there is no cast.
3393 70. If an address verification called from an ACL failed, and did not produce a
3394 user-specific message (i.e. there was only a "system" message), nothing was
3395 put in $acl_verify_message. In this situation, it now puts the system
3398 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3399 session; change 4.31/43 added the unwanted input to the log line - except
3400 that it did not do this in the start of session case. It now does.
3402 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3403 This is wrong and can cause the other end to generate a synchronization
3404 error if it is another Exim or anything else that does the synchronization
3405 check. A QUIT command is no longer sent after a timeout.
3407 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3408 during host lookups.
3410 74. The defer_ok option of callout verification was not working if it was used
3411 when verifying addresses in header lines, that is, for this case:
3413 verify = header_sender/callout=defer_ok
3415 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3416 those file descriptors could be used for SMTP connections. If anything
3417 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3418 could be sent to the SMTP client, causing chaos. The daemon now opens
3419 stdin, stdout, and stderr to /dev/null when it puts itself into the
3422 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3423 log by default. The user can override this with suitable Perl magic.
3425 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3426 the log message when discard triggers, was not working for the DATA ACL or
3427 for the non-SMTP ACL.
3429 78. Error message wording change in sieve.c.
3431 79. If smtp_accept_max_per_host was set, the number of connections could be
3432 restricted to fewer than expected, because the daemon was trying to set up
3433 a new connection before checking whether the processes handling previous
3434 connections had finished. The check for completed processes is now done
3435 earlier. On busy systems, this bug wouldn't be noticed because something
3436 else would have woken the daemon, and it would have reaped the completed
3439 80. If a message was submitted locally by a user whose login name contained one
3440 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3441 It caused a spool format error. I have fixed the spool reading code. A
3442 related problem was that the "from" clause in the Received: line became
3443 illegal because of the space(s). It is now covered by ${quote_local_part.
3445 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3448 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3450 83. Patch from Sieve maintainer to fix unterminated string problem in
3451 "vacation" handling.
3453 84. Some minor changes to the Linux configuration files to help with other
3454 OS variants using glibc.
3456 85. One more patch for Sieve to update vacation handling to latest spec.
3459 ----------------------------------------------------
3460 See the note above about the 4.44 and 4.50 releases.
3461 ----------------------------------------------------
3467 1. Change 4.43/35 introduced a bug that caused file counts to be
3468 incorrectly computed when quota_filecount was set in an appendfile
3471 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3472 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3475 3. Give more explanation in the error message when the command for a transport
3476 filter fails to execute.
3478 4. There are several places where Exim runs a non-Exim command in a
3479 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3480 was being done only for the command run by the queryprogram router. It is
3481 now done for all such subprocesses. The other cases are: ${run, transport
3482 filters, and the commands run by the lmtp and pipe transports.
3484 5. Some older OS have a limit of 256 on the maximum number of file
3485 descriptors. Exim was using setrlimit() to set 1000 as a large value
3486 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3487 systems. I've change it so that if it can't get 1000, it tries for 256.
3489 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3490 was an oversight, and furthermore, ever since the addition of extra
3491 controls (e.g. 4.43/32), the checks on when to allow different forms of
3492 "control" were broken. There should now be diagnostics for all cases when a
3493 control that does not make sense is encountered.
3495 7. $recipients is now available in the predata ACL (oversight).
3497 8. Tidy the search cache before the fork to do a delivery from a message
3498 received from the command line. Otherwise the child will trigger a lookup
3499 failure and thereby defer the delivery if it tries to use (for example) a
3500 cached ldap connection that the parent has called unbind on.
3502 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3503 of $address_data from the recipient verification was clobbered by the
3504 sender verification.
3506 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3507 was its contents. (It was OK if the option was not defined at all.)
3509 11. A "Completed" log line is now written for messages that are removed from
3510 the spool by the -Mrm option.
3512 12. $host_address is now set to the target address during the checking of
3513 ignore_target_hosts.
3515 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3516 being passed; this would have caused $sender_host_name to have been used if
3517 matching the list had actually called for a host name (not very likely,
3518 since this list is usually IP addresses). A host name is now passed as
3521 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3522 code that specifies a non-restarting handler (typically sigaction() in
3523 modern systems) in an attempt to fix a rare and obscure crash bug.
3525 15. Narrowed the window for a race in the daemon that could cause it to ignore
3526 SIGCHLD signals. This is not a major problem, because they are used only to
3527 wake it up if nothing else does.
3529 16. A malformed maildirsize file could cause Exim to calculate negative values
3530 for the mailbox size or file count. Odd effects could occur as a result.
3531 The maildirsize information is now recalculated if the size or filecount
3534 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3535 support for a long time. Removed HAVE_SYS_VFS_H.
3537 18. Updated exipick to current release from John Jetmore.
3539 19. Allow an empty sender to be matched against a lookup in an address list.
3540 Previously the only cases considered were a regular expression, or an
3543 20. Exim went into a mad DNS lookup loop when doing a callout where the
3544 host was specified on the transport, if the DNS lookup yielded more than
3547 21. The RFC2047 encoding function was originally intended for short strings
3548 such as real names; it was not keeping to the 75-character limit for
3549 encoded words that the RFC imposes. It now respects the limit, and
3550 generates multiple encoded words if necessary. To be on the safe side, I
3551 have increased the buffer size for the ${rfc2047: expansion operator from
3554 22. Failure to deliver a bounce message always caused it to be frozen, even if
3555 there was an errors_to setting on the router. The errors_to setting is now
3558 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3559 canonical form (fully expanded) before being placed in
3560 $sender_host_address.
3562 24. Updated eximstats to version 1.33
3564 25. Include certificate and key file names in error message when GnuTLS fails
3565 to set them up, because the GnuTLS error message doesn't include the name
3566 of the failing file when there is a problem reading it.
3568 26. Expand error message when OpenSSL has problems setting up cert/key files.
3571 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3572 (this can affect the format of dates).
3574 28. exim_tidydb, when checking for the continued existence of a message for
3575 which it has found a message-specific retry record, was not finding
3576 messages that were in split spool directories. Consequently, it was
3577 deleting retry records that should have stayed in existence.
3579 29. eximstats updated to version 1.35
3580 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3581 1.35 - bugfix such that pie charts by volume are generated correctly
3583 30. The SPA authentication driver was not abandoning authentication and moving
3584 on to the next authenticator when an expansion was forced to fail,
3585 contradicting the general specification for all authenticators. Instead it
3586 was generating a temporary error. It now behaves as specified.
3588 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3589 (the order specifies the preference for clients). The order is now AES256,
3590 AES128, 3DES, ARCFOUR128.
3592 31. Small patch to Sieve code - explicitly set From: when generating an
3595 32. Exim crashed if a remote delivery caused a very long error message to be
3596 recorded - for instance if somebody sent an entire SpamAssassin report back
3597 as a large number of 550 error lines. This bug was coincidentally fixed by
3598 increasing the size of one of Exim's internal buffers (big_buffer) that
3599 happened as part of the Exiscan merge. However, to be on the safe side, I
3600 have made the code more robust (and fixed the comments that describe what
3603 33. Some experimental protocols are using DNS PTR records for new purposes. The
3604 keys for these records are domain names, not reversed IP addresses. The
3605 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3606 leaves it alone. Component reversal etc. now happens only for IP addresses.
3609 34. The host_aton() function is supposed to be passed a string that is known
3610 to be a valid IP address. However, in the case of IPv6 addresses, it was
3611 not checking this. This is a hostage to fortune. Exim now panics and dies
3612 if the condition is not met. A case was found where this could be provoked
3613 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3614 components; fortuitously, this particular loophole had already been fixed
3615 by change 4.50/55 or 4.44/33 above.
3617 If there are any other similar loopholes, the new check in host_aton()
3618 itself should stop them being exploited. The report I received stated that
3619 data on the command line could provoke the exploit when Exim was running as
3620 exim, but did not say which command line option was involved. All I could
3621 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3622 running as the user.
3625 35. There was a buffer overflow vulnerability in the SPA authentication code
3626 (which came originally from the Samba project). I have added a test to the
3627 spa_base64_to_bits() function which I hope fixes it.
3630 36. The daemon start-up calls getloadavg() while still root for those OS that
3631 need the first call to be done as root, but it missed one case: when
3632 deliver_queue_load_max is set with deliver_drop_privilege. This is
3633 necessary for the benefit of the queue runner, because there is no re-exec
3634 when deliver_drop_privilege is set.
3636 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3637 was in use, was not putting the data itself into the right store pool;
3638 consequently, it could be overwritten for a subsequent message in the same
3639 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3642 38. Sometimes the final signoff response after QUIT could fail to get
3643 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3644 < 0 before doing a fflush(). This bug looks as though it goes back to the
3645 introduction of TLS in release 3.20, but "sometimes" must have been rare
3646 because the tests only now provoked it.
3652 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3653 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3654 Now it returns an integer. A number of calls were still expecting a T/F
3655 return. Fortuitously, in all cases, the tests worked in OK situations,
3656 which is the norm. However, things would have gone wrong on any write
3657 failures on the smtp file descriptor. This function is used when sending
3658 messages over SMTP and also when doing verify callouts.
3660 2. When Exim is called to do synchronous delivery of a locally submitted
3661 message (the -odf or -odi options), it no longer closes stderr before doing
3664 3. Implemented the mua_wrapper option.
3666 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3668 5. Implemented the functions header_remove(), header_testname(),
3669 header_add_at_position(), and receive_remove_recipient(), and exported them
3672 6. If an ACL "warn" statement specified the addition of headers, Exim already
3673 inserted X-ACL-Warn: at the start if there was no header name. However, it
3674 was not making this test for the second and subsequent header lines if
3675 there were newlines in the string. This meant that an invalid header could
3676 be inserted if Exim was badly configured.
3678 7. Allow an ACL "warn" statement to add header lines at the start or after all
3679 the Received: headers, as well as at the end.
3681 8. Added the rcpt_4xx retry error code.
3683 9. Added postmaster_mailfrom=xxx to callout verification option.
3685 10. Added mailfrom=xxxx to the callout verification option, for verify=
3688 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3689 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3691 12. Inserted some casts to stop certain compilers warning when using pointer
3692 differences as field lengths or precisions in printf-type calls (mostly
3693 affecting debugging statements).
3695 13. Added optional readline() support for -be (dynamically loaded).
3697 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3698 same clock tick as a message's arrival, so that its received time was the
3699 same as the "first fail" time on the retry record, and that message
3700 remained on the queue past the ultimate address timeout, every queue runner
3701 would try a delivery (because it was past the ultimate address timeout) but
3702 after another failure, the ultimate address timeout, which should have then
3703 bounced the address, did not kick in. This was a "< instead of <=" error;
3704 in most cases the first failure would have been in the next clock tick
3705 after the received time, and all would be well.
3707 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3708 being recognized when the domain list was tested by the match_domain
3709 condition in an expansion string.
3711 16. Added the ${str2b64: operator.
3713 17. Exim was always calling setrlimit() to set a large limit for the number of
3714 processes, without checking whether the existing limit was already
3715 adequate. (It did check for the limit on file descriptors.) Furthermore,
3716 errors from getrlimit() and setrlimit() were being ignored. Now they are
3717 logged to the main and panic logs, but Exim does carry on, to try to do its
3718 job under whatever limits there are.
3720 18. Imported PCRE 5.0.
3722 19. Trivial typo in log message " temporarily refused connection" (the leading
3725 20. If the log selector return_path_on_delivery was set and an address was
3726 redirected to /dev/null, the delivery process crashed because it assumed
3727 that a return path would always be set for a "successful" delivery. In this
3728 case, the whole delivery is bypassed as an optimization, and therefore no
3731 21. Internal re-arrangement: the function for sending a challenge and reading
3732 a response while authentication was assuming a zero-terminated challenge
3733 string. It's now changed to take a pointer and a length, to allow for
3734 binary data in such strings.
3736 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3738 23. Exim was not respecting finduser_retries when seeking the login of the
3739 uid under which it was called; it was always trying 10 times. (The default
3740 setting of finduser_retries is zero.) Also, it was sleeping after the final
3741 failure, which is pointless.
3743 24. Implemented tls_on_connect_ports.
3745 25. Implemented acl_smtp_predata.
3747 26. If the domain in control=submission is set empty, Exim assumes that the
3748 authenticated id is a complete email address when it generates From: or
3749 Sender: header lines.
3751 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3752 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3753 chown and chgrp in /bin and hostname in /usr/bin.
3755 28. Exim was keeping the "process log" file open after each use, just as it
3756 does for the main log. This opens the possibility of it remaining open for
3757 long periods when the USR1 signal hits a daemon. Occasional processlog
3758 errors were reported, that could have been caused by this. Anyway, it seems
3759 much more sensible not to leave this file open at all, so that is what now
3762 29. The long-running daemon process does not normally write to the log once it
3763 has entered its main loop, and it closes the log before doing so. This is
3764 so that log files can straightforwardly be renamed and moved. However,
3765 there are a couple of unusual error situations where the daemon does write
3766 log entries, and I had neglected to close the log afterwards.
3768 30. The text of an SMTP error response that was received during a remote
3769 delivery was being truncated at 512 bytes. This is too short for some of
3770 the long messages that one sometimes sees. I've increased the limit to
3773 31. It is now possible to make retry rules that apply only when a message has a
3774 specific sender, in particular, an empty sender.
3776 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3777 it possible to be selective about when SMTP synchronization is enforced.
3779 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3781 32. Implemented hosts_connection_nolog.
3783 33. Added an ACL for QUIT.
3785 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3788 35. Added mailbox_size and mailbox_filecount to appendfile.
3790 36. Added control = no_multiline_responses to ACLs.
3792 37. There was a bug in the logic of the code that waits for the clock to tick
3793 in the case where the clock went backwards by a substantial amount such
3794 that the microsecond fraction of "now" was more than the microsecond
3795 fraction of "then" (but the whole seconds number was less).
3797 38. Added support for the libradius Radius client library this is found on
3798 FreeBSD (previously only the radiusclient library was supported).
3804 1. When certain lookups returned multiple values in the form name=value, the
3805 quoting of the values was not always being done properly. Specifically:
3806 (a) If the value started with a double quote, but contained no whitespace,
3808 (b) If the value contained whitespace other than a space character (i.e.
3809 tabs or newlines or carriage returns) it was not quoted.
3810 This fix has been applied to the mysql and pgsql lookups by writing a
3811 separate quoting function and calling it from the lookup code. The fix
3812 should probably also be applied to nisplus, ibase and oracle lookups, but
3813 since I cannot test any of those, I have not disturbed their existing code.
3815 2. A hit in the callout cache for a specific address caused a log line with no
3816 reason for rejecting RCPT. Now it says "Previous (cached) callout
3817 verification failure".
3819 3. There was an off-by-one bug in the queryprogram router. An over-long
3820 return line was truncated at 256 instead of 255 characters, thereby
3821 overflowing its buffer with the terminating zero. As well as fixing this, I
3822 have increased the buffer size to 1024 (and made a note to document this).
3824 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3825 when Exim is waiting for an SMTP response from a remote server, Exim
3826 restarts its select() call on the socket, thereby resetting its timeout.
3827 This is not a problem when such interrupts are rare. Somebody set up a cron
3828 job to run exiwhat every 2 minutes, which is less than the normal select()
3829 timeout (5 or 10 minutes). This meant that the select() timeout never
3830 kicked in because it was always reset. I have fixed this by comparing the
3831 time when an interrupt arrives with the time at the start of the first call
3832 to select(). If more time than the timeout has elapsed, the interrupt is
3833 treated as a timeout.
3835 5. Some internal re-factoring in preparation for the addition of Sieve
3836 extensions (by MH). In particular, the "personal" test is moved to a
3837 separate function, and given an option for scanning Cc: and Bcc: (which is
3838 not set for Exim filters).
3840 6. When Exim created an email address using the login of the caller as the
3841 local part (e.g. when creating a From: or Sender: header line), it was not
3842 quoting the local part when it contained special characters such as @.
3844 7. Installed new OpenBSD configuration files.
3846 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3847 try to make them clearer.
3849 9. Callout options, other than the timeout value, were being ignored when
3850 verifying sender addresses in header lines. For example, when using
3852 verify = header_sender/callout=no_cache
3854 the cache was (incorrectly) being used.
3856 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3857 only the Cygwin environment.
3859 11. When return_path_on_delivery was set as a log selector, if different remote
3860 addresses in the same message used different return paths and parallel
3861 remote delivery occurred, the wrong values would sometimes be logged.
3862 (Whenever a remote delivery process finished, the return path value from
3863 the most recently started remote delivery process was logged.)
3865 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3866 lines when AUTH and/or TLS are in use. This is the "received protocol"
3867 field. Exim used to use "asmtp" for authenticated SMTP, without any
3868 indication (in the protocol name) for TLS use. Now it follows the RFC and
3869 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3870 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3871 names appear in log lines as well as in Received: header lines.
3873 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3874 extensions, and comparison tests, and to fix some bugs.
3876 14. Changes to the "personal" filter test:
3878 (1) The test was buggy in that it was just doing the equivalent of
3879 "contains" tests on header lines. For example, if a user's address was
3880 anne@some.where, the "personal" test would incorrectly be true for
3882 To: susanne@some.where
3884 This test is now done by extracting each address from the header in turn,
3885 and checking the entire address. Other tests that are part of "personal"
3886 are now done using regular expressions (for example, to check local parts
3887 of addresses in From: header lines).
3889 (2) The list of non-personal local parts in From: addresses has been
3890 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3891 taken from the Sieve specification recommendations.
3893 (3) If the message contains any header line starting with "List-" it is
3894 treated as non-personal.
3896 (4) The test for "circular" in the Subject: header line has been removed
3897 because it now seems ill-conceived.
3899 15. Minor typos in src/EDITME comments corrected.
3901 16. Installed latest exipick from John Jetmore.
3903 17. If headers_add on a router specified a text string that was too long for
3904 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3905 of string_sprintf() is now avoided.
3907 18. $message_body_size was not set (it was always zero) when running the DATA
3908 ACL and the local_scan() function.
3910 19. For the "mail" command in an Exim filter, no default was being set for
3911 the once_repeat time, causing a random time value to be used if "once" was
3912 specified. (If the value happened to be <= 0, no repeat happened.) The
3913 default is now 0s, meaning "never repeat". The "vacation" command was OK
3914 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3915 (I found it when inspecting the code).
3917 20. There is now an overall timeout for performing a callout verification. It
3918 defaults to 4 times the callout timeout, which applies to individual SMTP
3919 commands during the callout. The overall timeout applies when there is more
3920 than one host that can be tried. The timeout is checked before trying the
3921 next host. This prevents very long delays if there are a large number of
3922 hosts and all are timing out (e.g. when the network connections are timing
3923 out). The value of the overall timeout can be changed by specifying an
3924 additional sub-option for "callout", called "maxwait". For example:
3926 verify = sender/callout=5s,maxwait=20s
3928 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3929 to the end before writing, but this should make it even safer).
3931 22. Exim was forgetting that it had advertised PIPELINING for the second and
3932 subsequent messages on an SMTP connection. It was also not resetting its
3933 memory on STARTTLS and an internal HELO.
3935 23. When Exim logs an SMTP synchronization error within a session, it now
3936 records whether PIPELINING has been advertised or not.
3938 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3939 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3940 rather than long int.
3942 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3944 26. Added the never_mail option to autoreply.
3950 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3951 crash if the getsockname() call failed; this can happen if a connection is
3952 closed very soon after it is established. The problem was simply in the
3953 order in which certain operations were done, causing Exim to try to write
3954 to the SMTP stream before it had set up the file descriptor. The bug has
3955 been fixed by making things happen in the correct order.
3961 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3962 before the connection was closed, thus losing the rejection response.
3964 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3965 some early Solaris releases, but causes trouble in current releases where
3966 socklen_t is defined.
3968 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3971 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3974 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3975 that releasing the top of it at the end releases what was used for sub-
3976 expansions (unless the block got too big). However, discard this block if
3977 the first thing is a variable or header, so that we can use its block when
3978 it is dynamic (useful for very large $message_headers, for example).
3980 6. Lookups now cache *every* query, not just the most recent. A new, separate
3981 store pool is used for this. It can be recovered when all lookup caches are
3982 flushed. Lookups now release memory at the end of their result strings.
3983 This has involved some general refactoring of the lookup sources.
3985 7. Some code has been added to the store_xxx() functions to reduce the amount
3986 of flapping under certain conditions.
3988 8. log_incoming_interface used to affect only the <= reception log lines. Now
3989 it causes the local interface and port to be added to several more SMTP log
3990 lines, for example "SMTP connection from", and rejection lines.
3992 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3994 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3996 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3997 could be overwritten at the end of the current message (or the start of a
3998 new message if it was set in a HELO ACL). The value is now preserved for
3999 the duration of the SMTP connection.
4001 12. If a transport had a headers_rewrite setting, and a matching header line
4002 contained an unqualified address, that address was qualified, even if it
4003 did not match any rewriting rules. The underlying bug was that the values
4004 of the flags that permit the existence of unqualified sender and recipient
4005 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4006 non-local messages, and by -bnq for local messages) were not being
4007 preserved with the message after it was received.
4009 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4010 "next input=" as part of the text comprising the host identity instead of
4011 the correct text. The code was using the same buffer for two different
4012 strings. However, depending on which order the printing function evaluated
4013 its arguments, the bug did not always show up. Under Linux, for example, my
4014 test suite worked just fine.
4016 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4017 doesn't work with some older versions of Perl. It has been changed to "my",
4018 which in any case is probably the better facility to use.
4020 15. A really picky compiler found some instances of statements for creating
4021 error messages that either had too many or two few arguments for the format
4024 16. The size of the buffer for calls to the DNS resolver has been increased
4025 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4026 for addresses that have a lot of PTR records. This alleviates a problem; it
4027 does not fully solve it.
4029 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4030 buffer now truncates the list and logs the incident, which is the same
4031 action as happens when Exim is looking up a host name and its aliases.
4032 Previously in this situation something unpredictable would happen;
4033 sometimes it was "internal error: store_reset failed".
4035 18. If a server dropped the connection unexpectedly when an Exim client was
4036 using GnuTLS and trying to read a response, the client delivery process
4037 crashed while trying to generate an error log message.
4039 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4040 string, for example:
4042 warn message = H1: something\nH2: something
4044 the text was added as a single header line from Exim's point of view
4045 though it ended up OK in the delivered message. However, searching for the
4046 second and subsequent header lines using $h_h2: did not work. This has been
4047 fixed. Similarly, if a system filter added multiple headers in this way,
4048 the routers could not see them.
4050 20. Expanded the error message when iplsearch is called with an invalid key to
4051 suggest using net-iplsearch in a host list.
4053 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4054 ACLs are no longer actually imposed (and a message to that effect is
4057 22. If a "gecos" field in a passwd entry contained escaped characters, in
4058 particular, if it contained a \" sequence, Exim got it wrong when building
4059 a From: or a Sender: header from that name. A second bug also caused
4060 incorrect handling when an unquoted " was present following a character
4061 that needed quoting.
4063 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4064 was not being matched caselessly.
4066 24. Arranged for all hyphens in the exim.8 source to be escaped with
4069 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4070 properly. Recipient callout cache records were still being keyed to include
4071 the sender, even when use_sender was set false. This led to far more
4072 callouts that were necessary. The sender is no longer included in the key
4073 when use_sender is false.
4075 26. Added "control = submission" modifier to ACLs.
4077 27. Added the ${base62d: operator to decode base 62 numbers.
4079 28. dnsdb lookups can now access SRV records.
4081 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4082 the configuration file.
4084 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4085 (-v) mode. This makes the output for a verbose queue run more intelligible.
4087 31. Added a use_postmaster feature to recipient callouts.
4089 32. Added the $body_zerocount variable, containing the number of binary zero
4090 bytes in the message body.
4092 33. The time of last modification of the "new" subdirectory is now used as the
4093 "mailbox time last read" when there is a quota error for a maildir
4096 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4098 35. Added +ignore_unknown as a special item in host lists.
4100 36. Code for decoding IPv6 addresses in host lists is now included, even if
4101 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4102 address was recognized as an IP address, but was then not correctly decoded
4103 into binary, causing unexpected and incorrect effects when compared with
4110 1. Very minor rewording of debugging text in manualroute to say "list of
4111 hosts" instead of "hostlist".
4113 2. If verify=header_syntax was set, and a header line with an unqualified
4114 address (no domain) and a large number of spaces between the end of the
4115 name and the colon was received, the reception process suffered a buffer
4116 overflow, and (when I tested it) crashed. This was caused by some obsolete
4117 code that should have been removed. The fix is to remove it!
4119 3. When running in the test harness, delay a bit after writing a bounce
4120 message to get a bit more predictability in the log output.
4122 4. Added a call to search_tidyup() just before forking a reception process. In
4123 theory, someone could use a lookup in the expansion of smtp_accept_max_
4124 per_host which, without the tidyup, could leave open a database connection.
4126 5. Added the variables $recipient_data and $sender_data which get set from a
4127 lookup success in an ACL "recipients" or "senders" condition, or a router
4128 "senders" option, similar to $domain_data and $local_part_data.
4130 6. Moved the writing of debug_print from before to after the "senders" test
4133 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4134 problems for message scanning, either using a data ACL, or using
4135 local_scan() because the Received: header was not generated till after they
4136 were called (in order to set the time as the time of reception completion).
4137 I have revised the way this works. The header is now generated after the
4138 body is received, but before the ACL or local_scan() are called. After they
4139 are run, the timestamp in the header is updated.
4145 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4146 before starting a queue runner without re-exec. This happened only when
4147 deliver_drop_privilege was set or when the Exim user was set to root. The
4148 effect of the bug was that timeouts during subsequent deliveries caused
4149 crashes instead of being properly handled. The handler is now left at its
4150 default (and expected) setting.
4152 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4153 message, again when deliver_drop_privilege is set or Exim is run as root.
4154 The bug described in (1) was not present in this case, but the tidying up
4155 of the other signals was missing. I have made the two cases consistent.
4157 3. The ignore_target_hosts setting on a manualroute router was being ignored
4158 for hosts that were looked up using the /MX notation.
4160 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4163 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4164 operated on the sender address. After changing the $sender_address to <>
4165 for the sender address verify, Exim was re-instated it as the original
4166 (before rewriting) address, but remembering that it had rewritten it, so it
4167 wasn't rewriting it again. This bug also had the effect of breaking the
4168 sender address verification caching when the sender address was rewritten.
4170 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4171 This has been changed so that if the ip literal address matches
4172 ignore_target_hosts, the router declines.
4174 7. Added expansion conditions match_domain, match_address, and match_local_
4175 part (NOT match_host).
4177 8. The placeholder for the Received: header didn't have a length field set.
4179 9. Added code to Exim itself and to exim_lock to test for a specific race
4180 condition that could lead to file corruption when using MBX delivery. The
4181 issue is with the lockfile that is created in /tmp. If this file is removed
4182 after a process has opened it but before that process has acquired a lock,
4183 there is the potential for a second process to recreate the file and also
4184 acquire a lock. This could lead to two Exim processes writing to the file
4185 at the same time. The added code performs the same test as UW imapd; it
4186 checks after acquiring the lock that its file descriptor still refers to
4187 the same named file.
4189 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4190 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4193 11. Added the smtp_active_hostname option. If used, this will typically be made
4194 to depend on the incoming interface address. Because $interface_address is
4195 not set up until the daemon has forked a reception process, error responses
4196 that can happen earlier (such as "too many connections") no longer contain
4199 12. If an expansion in a condition on a "warn" statement fails because a lookup
4200 defers, the "warn" statement is abandoned, and the next ACL statement is
4201 processed. Previously this caused the whole ACL to be aborted.
4203 13. Added the iplsearch lookup type.
4205 14. Added ident_timeout as a log selector.
4207 15. Added tls_certificate_verified as a log selector.
4209 16. Added a global option tls_require_ciphers (compare the smtp transport
4210 option of the same name). This controls incoming TLS connections.
4212 17. I finally figured out how to make tls_require_ciphers do a similar thing
4213 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4214 before starting the TLS session.
4216 18. Tabs are now shown as \t in -bP output.
4218 19. If the log selector return_path_on_delivery was set, Exim crashed when
4219 bouncing a message because it had too many Received: header lines.
4221 20. If two routers both had headers_remove settings, and the first one included
4222 a superfluous trailing colon, the final name in the first list and the
4223 first name in the second list were incorrectly joined into one item (with a
4224 colon in the middle).
4230 1. Added -C and -D options to the exinext utility, mainly to make it easier
4231 to include in the automated testing, but these could be helpful when
4232 multiple configurations are in use.
4234 2. The exinext utility was not formatting the output nicely when there was
4235 an alternate port involved in the retry record key, nor when there was a
4236 message id as well (for retries that were specific to a specific message
4237 and a specific host). It was also confused by IPv6 addresses, because of
4238 the additional colons they contain. I have fixed the IPv4 problem, and
4239 patched it up to do a reasonable job for IPv6.
4241 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4242 delivery, the log line now contains "pipelined" if PIPELINING was used.
4244 4. An SMTP transport process used to panic and die if the bind() call to set
4245 an explicit outgoing interface failed. This has been changed; it is now
4246 treated in the same way as a connect() failure.
4248 5. A reference to $sender_host_name in the part of a conditional expansion
4249 that was being skipped was still causing a DNS lookup. This no longer
4252 6. The def: expansion condition was not recognizing references to header lines
4253 that used bh_ and bheader_.
4255 7. Added the _cache feature to named lists.
4257 8. The code for checking quota_filecount in the appendfile transport was
4258 allowing one more file than it should have been.
4260 9. For compatibility with Sendmail, the command line option
4268 and sets the incoming protocol and host name (for trusted callers). The
4269 host name and its colon can be omitted when only the protocol is to be set.
4270 Note the Exim already has two private options, -pd and -ps, that refer to
4271 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4272 "s", but I don't think that's a major issue.
4274 10. A number of refactoring changes to the code, none of which should affect
4277 (a) The number of logging options was getting close to filling up the
4278 32-bit word that was used as a bit map. I have split them into two classes:
4279 those that are passed in the argument to log_write(), and those that are
4280 only ever tested independently outside of that function. These are now in
4281 separate 32-bit words, so there is plenty of room for expansion again.
4282 There is no change in the user interface or the logging behaviour.
4284 (b) When building, for example, log lines, the code previously used a
4285 macro that called string_cat() twice, in order to add two strings. This is
4286 not really sufficiently general. Furthermore, there was one instance where
4287 it was actually wrong because one of the argument was used twice, and in
4288 one call a function was used. (As it happened, calling the function twice
4289 did not affect the overall behaviour.) The macro has been replaced by a
4290 function that can join an arbitrary number of extra strings onto a growing
4293 (c) The code for expansion conditions now uses a table and a binary chop
4294 instead of a serial search (which was left over from when there were very
4295 few conditions). Also, it now recognizes conditions like "pam" even when
4296 the relevant support is not compiled in: a suitably worded error message is
4297 given if an attempt is made to use such a condition.
4299 11. Added ${time_interval:xxxxx}.
4301 12. A bug was causing one of the ddress fields not to be passed back correctly
4302 from remote delivery subprocesses. The field in question was not being
4303 subsequently used, so this caused to problems in practice.
4305 13. Added new log selectors queue_time and deliver_time.
4307 14. Might have fixed a bug in maildirsizefile handling that threw up
4308 "unexpected character" debug warnings, and recalculated the data
4309 unnecessarily. In any case, I expanded the warning message to give more
4312 15. Added the message "Restricted characters in address" to the statements in
4313 the default ACL that block characters like @ and % in local parts.
4315 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4316 Three changes have been made:
4318 (a) There was a serious bug; a negative response to MAIL caused the whole
4319 recipient domain to be cached as invalid, thereby blocking all messages
4320 to all local parts at the same domain, from all senders. This bug has
4321 been fixed. The domain is no longer cached after a negative response to
4322 MAIL if the sender used is not empty.
4324 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4327 (c) A new callout option, "use_sender" has been added for people who want
4328 the modified behaviour.
4334 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4337 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4338 indeed breaks things for older releases.
4340 3. Added additional logging to the case where there is a problem reading data
4341 from a filter that is running in a subprocess using a pipe, in order to
4342 try to track down a specific problem.
4344 4. Testing facility fudge: when running in the test harness and attempting
4345 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4346 getting "No route to host". Convert this to a timeout.
4348 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4351 6. Some OS don't have socklen_t but use size_t instead. This affects the
4352 fifth argument of getsockopt() amongst other things. This is now
4353 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4354 can be set for individual OS. I have set it for SunOS5, OSF1, and
4355 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4356 some earlier ones do not.
4358 7. Change 4.30/15 was not doing the test caselessly.
4360 8. The standard form for an IPv6 address literal was being rejected by address
4361 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4362 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4363 this, as well as the form without the "IPv6" on the front (but only when
4364 address literals are enabled, of course).
4366 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4368 10. Exim crashed if a message with an empty sender address specified by -f
4369 encountered a router with an errors_to setting. This could be provoked only
4370 by a command such as
4374 where an empty string was supplied; "<>" did not hit this bug.
4376 11. Installed PCRE release 4.5.
4378 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4379 remained set. It is now erased.
4381 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4382 times from message ids (which are base 36 rather than the normal 62).
4384 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4385 were being counted as actual protocol errors, and logged if the log
4386 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4387 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4388 and DATA following a set of rejected RCPTs do not count as protocol errors.
4389 In other words, Exim assumes they were pipelined, though this may not
4390 actually be the case. Of course, in all cases the client gets an
4391 appropriate error code.
4393 15. If a lookup fails in an ACL condition, a message about the failure may
4394 be available; it is used if testing the ACL cannot continue, because most
4395 such messages specify what the cause of the deferral is. However, some
4396 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4397 that caused an old message to be retained and used if a later statement
4398 caused a defer, replacing the real cause of the deferral.
4400 16. If an IP address had so many PTR records that the DNS lookup buffer
4401 was not large enough to hold them, Exim could crash while trying to process
4402 the truncated data. It now detects and logs this case.
4404 17. Further to 4.21/58, another change has been made: if (and only if) the
4405 first line of a message (the first header line) ends with CRLF, a bare LF
4406 in a subsequent header line has a space inserted after it, so as not to
4407 terminate the header.
4409 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4410 unnecessarily, used atoi() instead of strtol(), and didn't check the
4411 termination when getting file sizes from file names by regex.
4413 19. Completely re-implemented the support for maildirsize files, in the light
4414 of a number of problems with the previous contributed implementation
4415 (4.30/29). In particular:
4417 . If the quota is zero, the maildirsize file is maintained, but no quota is
4420 . If the maildir directory does not exist, it is created before any attempt
4421 to write a maildirsize file.
4423 . The quota value in the file is just a cache; if the quota is changed in
4424 the transport, the new value overrides.
4426 . A regular expression is available for excluding directories from the
4429 20. The autoreply transport checks the characters in options that define the
4430 message's headers; it allows continued headers, but it was checking with
4431 isspace() after an embedded newline instead of explicitly looking for a
4434 21. If all the "regular" hosts to which an address was routed had passed their
4435 expiry times, and had not reached their retry times, the address was
4436 bounced, even if fallback hosts were defined. Now Exim should go on to try
4439 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4440 equivalent code in the SMTP transport. Some hosts send humungous responses
4441 to HELO/EHLO, more than 1024 it seems.
4443 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4444 gives compiler warnings in some environments. I've now done it "properly",
4447 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4448 (because of problems with the built-in one) was declared to return uschar *
4449 instead of char *, causing compiler failure.
4451 25. Fixed a file descriptor leak when processing alias/forward files.
4453 26. Fixed a minor format string issue in dbfn.c.
4455 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4457 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4458 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4461 29. When a sender address is verified, it is cached, to save repeating the test
4462 when there is more than one recipient in a message. However, when the
4463 verification involves a callout, it is possible for different callout
4464 options to be set for different recipients. It is too complicated to keep
4465 track of this in the cache, so now Exim always runs a verification when a
4466 callout is required, relying on the callout cache for the optimization.
4467 The overhead is duplication of the address routing, but this should not be
4470 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4471 to be verified with callout=postmaster, and the main callout worked but the
4472 postmaster check failed, the verification correctly failed. However, if a
4473 subsequent RCPT command asked for sender verification *without* the
4474 postmaster check, incorrect caching caused this verification also to fail,
4477 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4478 it was not caching the DNS options (qualify_single, search_parents) that
4479 were used when the lookup failed. A subsequent lookup with different
4480 options therefore always gave the same answer, though there were cases
4481 where it should not have. (Example: a "domains = !$mx_any" option on a
4482 dnslookup router: the "domains" option is always processed without any
4483 widening, but the router might have qualify_single set.) Now Exim uses the
4484 cached value only when the same options are set.
4486 32. Added John Jetmore's "exipick" utility to the distribution.
4488 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4489 than a timeout (e.g. a certificate is required, and is not provided), an
4490 Exim server now closes the connection immediately. Previously it waited for
4491 the client to close - but if the client is SSL, it seems that they each
4492 wait for each other, leading to a delay before one of them times out.
4494 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4495 maintained 0.8.x compatibility because I don't think many are using it, and
4496 it is clearly obsolete.
4498 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4501 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4502 client certificate was expired. A simple patch fixes this, though I don't
4503 understand the full logic of why the verify callback is called multiple
4506 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4507 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4508 which causes problems with some clients (such as the Certicom SSL Plus
4509 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4510 disables the coutermeasure allowing Eudora to connect."
4512 38. Exim was not checking that a write() to a log file succeeded. This could
4513 lead to Bad Things if a log got too big, in particular if it hit a file
4514 size limit. Exim now panics and dies if it cannot write to a log file, just
4515 as it does if it cannot open a log file.
4517 39. Modified OS/Makefile-Linux so that it now contains
4519 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4521 The two -D definitions ensure that Exim is compiled with large file
4522 support, which makes it possible to handle log files that are bigger than
4525 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4526 instance) a domain was checked against a named list that involved a lookup,
4527 causing $domain_data to be set, then another domain was checked against the
4528 same list, then the first domain was re-checked, the value of $domain_data
4529 after the final check could be wrong. In particular, if the second check
4530 failed, it could be set empty. This bug probably also applied to
4533 41. The strip_trailing_dot option was not being applied to the address given
4534 with the -f command-line option.
4536 42. The code for reading a message's header from the spool was incrementing
4537 $received_count, but never initializing it. This meant that the value was
4538 incorrect (doubled) while delivering a message in the same process in which
4539 it was received. In the most common configuration of Exim, this never
4540 happens - a fresh exec is done - but it can happen when
4541 deliver_drop_privilege is set.
4543 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4544 it now includes up to 150 characters of the unexpected data in the log
4547 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4548 and building data strings. The size of both of these buffers was 10 000
4549 bytes - far larger than anybody would *ever* want, thought I. Needless to
4550 say, somebody hit the limit. I have increased the maximum line length to
4551 20 000 and the maximum data length of concatenated lines to 100 000. I have
4552 also fixed two bugs, because there was no checking on these buffers. Tsk,
4553 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4554 buffer is too small.
4556 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4557 lsearch lookups. Now it does.
4559 46. When parsing a route_list item in a manualroute router, a fixed-length
4560 buffer was used for the list of hosts. I made this 1024 bytes long,
4561 thinking that nobody would ever have a list of hosts that long. Wrong.
4562 Somebody had a whole pile of complicated expansion conditions, and the
4563 string was silently truncated, leading to an expansion error. It turns out
4564 that it is easier to change to an unlimited length (owing to other changes
4565 that have happened since this code was originally written) than to build
4566 structure for giving a limitation error. The length of the item that
4567 expands into the list of hosts is now unlimited.
4569 47. The lsearch lookup could not handle data where the length of text line was
4570 more than 4095 characters. Such lines were truncated, leading to shortened
4571 data being returned. It should now handle lines of any length.
4573 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4574 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4577 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4578 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4579 no longer happens in comment lines. A list of replacements is now placed
4580 at the head of all of the source files, except those whose only change is
4581 to replace PERL_COMMAND in the very first #! line.
4583 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4584 messages on the queue, with a bottom-up merge sort, using code contributed
4585 by Michael Haardt. This should make operations like -bp somewhat faster on
4586 large queues. It won't affect queue runners, except when queue_run_in_order
4589 51. Installed eximstats 1.31 in the distribution.
4591 52. Added support for SRV lookups to the dnslookup router.
4593 53. If an ACL referred to $message_body or $message_body_end, the value was not
4594 reset for any messages that followed in the same SMTP session.
4596 54. The store-handling optimization for building very long strings was not
4597 differentiating between the different store pools. I don't think this
4598 actually made any difference in practice, but I've tidied it.
4600 55. While running the routers to verify a sender address, $sender_address
4601 was still set to the sender address. This is wrong, because when routing to
4602 send a bounce to the sender, it would be empty. Therefore, I have changed
4603 it so that, while verifying a sender address, $sender_address is set to <>.
4604 (There is no change to what happens when verifying a recipient address.)
4606 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4607 target A or AAAA records (if not already returned) without resetting the
4608 qualify_single or search_parents options of the DNS resolver. These are
4609 inappropriate in this case because the targets of MX and SRV records must
4610 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4611 target that, when qualified, matched something in the local domain. These
4612 two options are now turned off when doing these lookups.
4614 57. It seems that at least some releases of Reiserfs (which does not have the
4615 concept of a fixed number of inodes) returns zero and not -1 for the
4616 number of available inodes. This interacted badly with check_spool_inodes,
4617 which assumed that -1 was the "no such thing" setting. What I have done is
4618 to check that the total number of inodes is greater than zero before doing
4619 the test of how many are available.
4621 58. When a "warn" ACL statement has a log_message modifier, the message is
4622 remembered, and not repeated. This is to avoid a lot of repetition when a
4623 message has many recipients that cause the same warning to be written.
4624 Howewer, Exim was preserving the list of already written lines for an
4625 entire SMTP session, which doesn't seem right. The memory is now reset if a
4626 new message is started.
4628 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4629 debugging output unless log_rewrite was also set.
4631 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4632 of (int)(handle) when we know that handle contains (void *)(-1).
4634 61. The Exim daemon panic-logs an error return when it closes the incoming
4635 connection. However "connection reset by peer" seems to be common, and
4636 isn't really an error worthy of noting specially, so that particular error
4639 62. When Exim is trying to find all the local interfaces, it used to panic and
4640 die if the ioctl to get the interface flags failed. However, it seems that
4641 on at least one OS (Solaris 9) it is possible to have an interface that is
4642 included in the list of interfaces, but for which you get a failure error
4643 for this call. This happens when the interface is not "plumbed" into a
4644 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4645 failure of the "get flags" call assumes that the interface is down.
4647 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4648 makes life easier for people who are doing arithmetic on fields extracted
4649 from dates, where you often get leading zeros that should not be
4650 interpreted as octal.
4652 64. Added qualify_domain to the redirect router, to override the global
4655 65. If a pathologically long header line contained very many addresses (the
4656 report of this problem mentioned 10 000) and each of them was rewritten,
4657 Exim could use up a very large amount of memory. (It kept on making new
4658 copies of the header line as it rewrote, and never released the old ones.)
4659 At the expense of a bit more processing, the header rewriting function has
4660 been changed so that it no longer eats memory in this way.
4662 66. The generation of the Received: header has been moved from the time that a
4663 message starts to be received, to the time that it finishes. The timestamp
4664 in the Received: header should now be very close to that of the <= log
4665 line. There are two side-effects of this change:
4667 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4668 logged header lines no longer include the local Received: line, because
4669 it has not yet been created. The same applies to a copy of the message
4670 that is returned to a non-SMTP sender when a message is rejected.
4672 (b) When a filter file is tested using -bf, no additional Received: header
4673 is added to the test message. After some thought, I decided that this
4676 This change does not affect the value of $received_for. It is still set
4677 after address rewriting, but before local_scan() is called.
4679 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4681 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4682 gave an unhelpful panic error message, and a defer error. I have managed to
4683 change this behaviour so that it now rejects any supplied certificate,
4684 which seems right, as the list of acceptable certificates is empty.
4686 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4687 gave an unhelpful defer error. I have not managed to make this reject any
4688 supplied certificates, but the error message it gives is "no certificate
4689 supplied", which is not helpful.
4691 70. exigrep's output now also includes lines that are not associated with any
4692 message, but which match the given pattern. Implemented by a patch from
4693 Martin Sluka, which also tidied up the Perl a bit.
4695 71. Recipient callout verification, like sender verification, was using <> in
4696 the MAIL FROM command. This isn't really the right thing, since the actual
4697 sender may affect whether the remote host accepts the recipient or not. I
4698 have changed it to use the actual sender in the callout; this means that
4699 the cache record is now keyed on a recipient/sender pair, not just the
4700 recipient address. There doesn't seem to be a real danger of callout loops,
4701 since a callout by the remote host to check the sender would use <>.
4702 [SEE ABOVE: changed after hitting problems.]
4704 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4705 temporary errors. However, in the case of such a code being given after
4706 the end of a data transmission (i.e. after ".") Exim was failing to write
4707 a retry record for the message. (Yes, there was some broken host that was
4708 actually sending 8xx at this point.)
4710 73. An unknown lookup type in a host list could cause Exim to panic-die when
4711 the list was checked. (An example that provoked this was putting <; in the
4712 middle of a list instead of at the start.) If this happened during a DATA
4713 ACL check, a -D file could be left lying around. This kind of configuration
4714 error no longer causes Exim to die; instead it causes a defer errror. The
4715 incident is still logged to the main and panic logs.
4717 74. Buglet left over from Exim 3 conversion. The message "too many messages
4718 in one connection" was written to the rejectlog but not the mainlog, except
4719 when address rewriting (yes!) was being logged.
4721 75. Added write_rejectlog option.
4723 76. When a system filter was run not as root (that is, when system_filter_user
4724 was set), the values of the $n variables were not being returned to the
4725 main process; thus, they were not subsequently available in the $sn
4728 77. Added +return_path_on_delivery log selector.
4730 78. A connection timeout was being treated differently from recipients deferred
4731 when testing hosts_max_try with a message that was older than the host's
4732 retry timeout. (The host should not be counted, thus allowing all hosts to
4733 be tried at least once before bouncing.) This may have been the cause of an
4734 occasionally reported bug whereby a message would remain on the queue
4735 longer than the retry timeout, but would be bounced if a delivery was
4736 forced. I say "may" because I never totally pinned down the problem;
4737 setting up timeout/retry tests is difficult. See also the next item.
4739 79. The ultimate address timeout was not being applied to errors that involved
4740 a combination of host plus message (for example, a timeout on a MAIL
4741 command). When an address resolved to a number of possible hosts, and they
4742 were not all tried for each delivery (e.g. because of hosts_max_try), a
4743 message could remain on the queue longer than the retry timeout.
4745 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4748 81. Fixed an obscure SMTP outgoing bug which required at least the following
4749 conditions: (a) there was another message waiting for the same server;
4750 (b) the server returned 5xx to all RCPT commands in the first message so
4751 that the message was not completed; (c) the server dropped the connection
4752 or gave a negative response to the RSET that Exim sends to abort the
4753 transaction. The observed case was a dropped connection after DATA that had
4754 been sent in pipelining mode. That is, the server had advertised PIPELINING
4755 but was not implementing it correctly. The effect of the bug was incorrect
4756 behaviour, such as trying another host, and this could lead to a crash.
4762 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4763 and daemon.c were passed as pointers to ints; they should have been
4764 pointers to socklen_t variables (which are typically unsigned ints).
4766 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4769 3. Fixed a really odd bug that affected only the testing scheme; patching a
4770 certain fixed string in the binary changed the value of another string that
4771 happened to be identical to the end of the original first string.
4773 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4774 name", it returns that address as the IP address. On some operating
4775 systems (e.g. Solaris), it also passes back the IP address string as the
4776 "host name". However, on others (e.g. Linux), it passes back an empty
4777 string. Exim wasn't checking for this, and was changing the host name to an
4778 empty string, assuming it had been canonicized.
4780 5. Although rare, it is permitted to have more than one PTR record for a given
4781 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4782 all the names associated with an address, because they do in Solaris.
4783 However, it seems that they do not in Linux for data that comes from the
4784 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4785 I found this out when I moved to a new Linux workstation and tried to run
4786 the Exim test suite.
4788 To get round this problem I have changed the code so that it now does its
4789 own call to the DNS to look up PTR records when searching for a host name.
4790 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4791 addresses that are only in /etc/hosts are still found.
4793 This behaviour is, however, controlled by an option called host_lookup_
4794 order, which defaults to "bydns:byaddr". If people want to use the other
4795 order, or indeed, just use one or the other means of lookup, they can
4796 specify it in this variable.
4798 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4799 some operating systems, this comes back from gethostbyaddr() as an empty
4800 string, and this is what Exim used to test for. However, it seems that in
4801 other systems, "." is yielded. Exim now tests for this case too.
4803 7. The values of check_spool_space and check_log_space are now held internally
4804 as a number of kilobytes instead of an absolute number of bytes. If a
4805 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4806 kilobyte. This means that much larger values can be stored.
4808 8. Exim monitor: an attempt to get the action menu when not actually pointing
4809 at a message produces an empty menu entitled "No message selected". This
4810 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4811 no entries in it ("Shell widget menu has zero width and/or height"). So I
4812 have added a single, blank menu entry in this case.
4814 9. Added ${quote_local_part.
4816 10. MIME decoding is now applied to the contents of Subject: header lines when
4819 11. Now that a reference to $sender_host_address automatically causes a reverse
4820 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4821 host lookup before query-style lookups in lists that might use this
4822 variable. This has therefore been abolished, and the "net-" prefix is no
4823 longer necessary for query-style lookups.
4825 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4826 have been a typo for LFLAGS, so it has been changed.
4828 13. The install script calls Exim with "-C /dev/null" in order to find the
4829 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4830 to be output. Howeve, since Exim outputs its version number before the
4831 error, it didn't break the script. It just looked ugly. I fixed this by
4832 always allowing "-C /dev/null" if the caller is root.
4834 14. Ignore overlarge ACL variable number when reading spool file - insurance
4835 against a later release with more variables having written the file.
4837 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4838 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4839 this, as well as the form without the "IPv6" on the front.
4841 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4842 OS/Makefile-Darwin file.
4844 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4845 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4847 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4848 variable. However, some OS use ints for inodes. I've added cast to long int
4849 to get rid of the compiler warning.
4851 19. I had forgotten to lock out "/../" in configuration file names when
4852 ALT_CONFIG_PREFIX was set.
4854 20. Routers used for verification do not need to specify transports. However,
4855 if such a router generated a host list, and callout was configured, Exim
4856 crashed, because it could not find a port number from the (non-existent)
4857 transport. It now assumes port 25 in this circumstance.
4859 21. Added the -t option to exigrep.
4861 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4862 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4863 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4864 them is set, LOOKUP_LSEARCH is forced.
4866 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4868 24. Added sender and host information to the "rejected by local_scan()" log
4869 line; previously there was no indication of these.
4871 25. Added .include_if_exists.
4873 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4874 Linux. It turns out that not all file systems support this. Apparently some
4875 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4876 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4877 not supported on the file descriptor, is now ignored when Exim is trying to
4878 sync a directory. This applies only to Linux.
4880 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4882 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4883 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4886 29. Added support for maildirsize files from supplied patch (modified a bit).
4888 30. The use of :fail: followed by an empty string could lead Exim to respond to
4889 sender verification failures with (e.g.):
4891 550 Verification failed for <xxx>
4892 550 Sender verify failed
4894 where the first response line was missing the '-' that indicates it is not
4895 the final line of the response.
4897 31. The loop for finding the name of the user that called Exim had a hardwired
4898 limit of 10; it now uses the value of finduser_retries, which is used for
4899 all other user lookups.
4901 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4904 33. Exim was neglecting to zero errno before one call of strtol() when
4905 expanding a string and expecting an integer value. On some systems this
4906 resulted in spurious "integer overflow" errors. Also, it was casting the
4907 result into an int without checking.
4909 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4910 did not work. The code looks as if it has *never* worked, though it appears
4911 to have been documented since at least releast 1.62. I have made it work.
4913 35. The "timeout_DNS" error in retry rules, also documented since at least
4914 1.62, also never worked. As it isn't clear exactly what this means, and
4915 clearly it isn't a major issue, I have abolished the feature by treating it
4916 as "timeout", and writing a warning to the main and panic logs.
4918 36. The display of retry rules for -brt wasn't always showing the error code
4921 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4922 timeout_connect_A, timeout_connect_MX.
4924 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4925 to the empty sender.
4927 39. The daemon was not analysing the content of -oX till after it had closed
4928 stderr and disconnected from the controlling terminal. This meant that any
4929 syntax errors were only noted on the panic log, and the return code from
4930 the command was 0. By re-arranging the code a little, I've made the
4931 decoding happen first, so such errors now appear on stderr, and the return
4932 code is 1. However, the actual setting up of the sockets still happens in
4933 the disconnected process, so errors there are still only recorded on the
4936 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4937 connections (as happens on some IP stacks) was logged at start up time as
4938 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4939 it from "IPv6 and IPv4", which means that two separate sockets are being
4942 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4943 says whether AF_INET or AF_INET6 was passed as an argument.
4945 42. Exiwhat output was messed up when time zones were included in log
4948 43. Exiwhat now gives more information about the daemon's listening ports,
4949 and whether -tls-on-connect was used.
4951 44. The "port" option of the smtp transport is now expanded.
4953 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4954 silently ignored. Now an error message is written to the main and panic
4957 46. There's a new ACL modifier called "logwrite" which writes to a log file
4958 as soon as it is encountered.
4960 47. Added $local_user_uid and $local_user_gid at routing time.
4962 48. Exim crashed when trying to verify a sender address that was being
4965 49. Exim was recognizing only a space character after ".include". It now also
4966 recognizes a tab character.
4968 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4969 extracting the relevant information from the specification. The man page no
4970 longer contains scrambled data for the -d option, and I've added a section
4971 at the front about calling Exim under different names.
4973 51. Added "extra_headers" argument to the "mail" command in filter files.
4975 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4978 53. Installed eximstats 1.29.
4980 54. Added transport_filter_timeout as a generic transport option.
4982 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4983 Cc: header lines. This was required by RFC 822, but it not required by RFC
4986 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4987 incoming messages that did not have them. Now it does so only if the
4988 message originates locally, that is, if there is no associated remote host
4989 address. When Resent- header lines are present, this applies to the Resent-
4990 lines rather than the non-Resent- lines.
4992 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4993 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4995 58. Messages for configuration errors now include the name of the main
4996 configuration file - useful now that there may be more than one file in a
4997 list (.included file names were always shown).
4999 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5000 for those rare installations that do not start the daemon as root or run it
5001 setuid root. I've cut out the call to initgroups() if the daemon is not
5004 60. The Exim user and group can now be bound into the binary as text strings
5005 that are looked up at the start of Exim's processing.
5007 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5009 62. Added $mailstore_basename variable.
5011 63. Installed patch to sieve.c from Michael Haardt.
5013 64. When Exim failed to open the panic log after failing to open the main log,
5014 the original message it was trying to log was written to stderr and debug
5015 output, but if they were not available (the usual case in production), it
5016 was lost. Now it is written to syslog before the two lines that record the
5017 failures to open the logs.
5019 65. Users' Exim filters run in subprocesses under the user's uid. It is
5020 possible for a "deliver" command or an alias in a "personal" command to
5021 provoke an address rewrite. If logging of address rewriting is configured,
5022 this fails because the process is not running as root or exim. There may be
5023 a better way of dealing with this, but for the moment (because 4.30 needs
5024 to be released), I have disabled address rewrite logging when running a
5025 filter in a non-root, non-exim process.
5031 1. The buildconfig auxiliary program wasn't quoting the value set for
5032 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5033 not defined. This bug was masked in 4.22 by the effect that was fixed in
5036 2. Some messages that were rejected after a message id was allocated were
5037 shown as "incomplete" by exigrep. It no longer does this for messages that
5038 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5040 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5041 have allow_domain_literals set, the ID did not get logged in the <= line.
5042 Domain literals are now always recognized in Message-ID: header lines.
5044 4. The first argument for a ${extract expansion item is the key name or field
5045 number. Leading and trailing spaces in this item were not being ignored,
5046 causing some misleading effects.
5048 5. When deliver_drop_privilege was set, single queue runner processes started
5049 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5050 same command in the process it spins off) were not dropping privilege.
5052 6. When the daemon running as "exim" started a queue runner, it always
5053 re-executed Exim in the spun-off process. This is a waste of effort when
5054 deliver_drop_privilege is set. The new process now just calls the
5055 queue-runner function directly.
5061 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5064 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5065 ignored. Though the use of -oP was forcing the writing of a pid file, it
5066 was always written to the default place.
5068 3. If the message "no IP address found for host xxxx" is generated during
5069 incoming verification, it is now followed by identification of the incoming
5070 connection (so you can more easily find what provoked it).
5072 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5074 5. Added some features to "harden" Exim a bit more against certain attacks:
5076 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5077 be put in Local/Makefile. This is like the never_users runtime option,
5078 but it cannot be overridden. The default setting is "root".
5080 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5081 prefix string with which any file named in a -C command line option
5084 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5085 is retained for -C and -D only if the caller of Exim is root. Without
5086 it, the exim user may also use -C and -D and retain privilege.
5088 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5089 command line option is disabled.
5091 6. Macro names set by the -D option must start with an upper case letter, just
5092 like macro names defined in the configuration file.
5094 7. Added "dereference=" facility to LDAP.
5096 8. Two instances of the typo "uknown" in the source files are fixed.
5098 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5099 the Configure-Makefile script screwed up while processing it.
5101 10. Incorporated PCRE 4.4.
5103 11. The SMTP synchronization check was not operating right at the start of an
5104 SMTP session. For example, it could not catch a HELO sent before the client
5105 waited for the greeting. There is now a check for outstanding input at the
5106 point when the greeting is written. Because of the duplex, asynchronous
5107 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5108 way, but not yet received, when the check is performed.
5110 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5111 on TCP/IP sockets, because this apparently causes some broken clients to
5114 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5115 unchanged) from the Cygwin maintainer.
5117 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5118 was supported instead of testing for mbx. Effectively a typo.
5120 15. The spa authenticator server code was not checking that the input it
5121 received was valid base64.
5123 16. The debug output line for the "set" modifier in ACLs was not showing the
5124 name of the variable that was being set.
5126 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5128 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5129 lookup on its own. Something else was needed to trigger the lookup. For
5130 example, a match in host_lookup or the need for a host name in a host list.
5131 Now, if $sender_host_name is referenced and the host name has not yet been
5132 looked up, a lookup is performed. If the lookup fails, the variable remains
5133 empty, and $host_lookup_failed is set to "1".
5135 19. Added "eqi" as a case-independent comparison operator.
5137 20. The saslauthd authentication condition could segfault if neither service
5138 nor realm was specified.
5140 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5141 error message that was logged was misleading, and incoming SMTP
5142 connections were dropped. The message is now more accurate, and temporary
5143 errors are given to SMTP connections.
5145 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5146 (except RSET) with a 421 error, until QUIT is received. However, it was
5147 failing to send a response to QUIT.
5149 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5150 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5151 ACL is now run after the helo_try_verify_hosts code.
5153 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5154 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5155 case-independent, so other case variants are also recognized). Apparently
5156 some systems use these upper case variants.
5158 25. If more than two messages were waiting for the same host, and a transport
5159 filter was specified for the transport, Exim sent two messages over the
5160 same TCP/IP connection, and then failed with "socket operation on non-
5161 socket" when it tried to send the third.
5163 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5165 27. The extern definition of crypt16() in expand.c was not being excluded when
5166 the OS had its own crypt16() function.
5168 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5169 as a preferred synonym for return_size_limit, both as an option and as an
5172 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5174 30. Changed the default configuration ACL to relax the local part checking rule
5175 for addresses that are not in any local domains. For these addresses,
5176 slashes and pipe symbols are allowed within local parts, but the sequence
5177 /../ is explicitly forbidden.
5179 31. SPA server authentication was not clearing the challenge buffer before
5182 32. log_message in a "warn" ACL statement was writing to the reject log as
5183 well as to the main log, which contradicts the documentation and doesn't
5184 seem right (because no rejection is happening). So I have stopped it.
5186 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5187 However, I am unable to do any testing of this.
5189 34. Fixed an infelicity in the appendfile transport. When checking directories
5190 for a mailbox, to see if any needed to be created, it was accidentally
5191 using path names with one or more superfluous leading slashes; tracing
5192 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5194 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5195 contents are added to the log line that is written for every discarded
5196 recipient. (Previously a log_message setting was ignored.)
5198 36. The ${quote: operator now quotes the string if it is empty.
5200 37. The install script runs exim in order to find its version number. If for
5201 some reason other than non-existence or emptiness, which it checks, it
5202 could not run './exim', it was installing it with an empty version number,
5203 i.e. as "exim-". This error state is now caught, and the installation is
5206 38. An argument was missing from the function that creates an error message
5207 when Exim fails to connect to the socket for saslauthd authentication.
5208 This could cause Exim to crash, or give a corrupted message.
5210 39. Added isip, isip4, and isip6 to ${if conditions.
5212 40. The ACL variables $acl_xx are now saved with the message, and can be
5213 accessed later in routers, transports, and filters.
5215 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5216 strings in the file are not string-expanded.
5218 42. If a MAIL command specified a SIZE value that was too large to fit into an
5219 int variable, the check against message_size_limit failed. Such values are
5220 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5221 day this will have to be increased, but I don't think I want to be around
5222 when emails are that large.
5229 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5230 iconv() is not standard in FreeBSD.
5232 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5233 IPv6 enabled. The observed symptom was a segmentation fault on return from
5234 the function os_common_find_running_interfaces() in src/os.c.
5236 3. In the check_special_case() function in daemon.c I had used "errno" as an
5237 argument name, which causes warnings on some systems. This was basically a
5238 typo, since it was named "eno" in the comments!
5240 4. The code that waits for the clock to tick (at a resolution of some fraction
5241 of a second) so as to ensure message-id uniqueness was always waiting for
5242 at least one whole tick, when it could have waited for less. [This is
5243 almost certainly not relevant at current processor speeds, where it is
5244 unlikely to ever wait at all. But we try to future-proof.]
5246 5. The function that sleeps for a time interval that includes fractions of a
5247 second contained a race. It did not block SIGALRM between setting the
5248 timer, and suspending (a couple of lines later). If the interval was short
5249 and the sigsuspend() was delayed until after it had expired, the suspension
5250 never ended. On busy systems this could lead to processes getting stuck for
5253 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5254 process, before it forks any delivery processes. The open lookup caching
5255 mechanism meant that the open file or database connection was passed into
5256 the delivery process. The problem was that delivery processes always tidy
5257 up cached lookup data. This could cause a problem for the next delivery
5258 process started by the queue runner, because the external queue runner
5259 process does not know about the closure. So the next delivery process
5260 still has data in the lookup cache. In the case of a file lookup, there was
5261 no problem because closing a file descriptor in a subprocess doesn't affect
5262 the parent. However, if the lookup was caching a connection to a database,
5263 the connection was closed, and the second delivery process was likely to
5264 see errors such as "PGSQL: query failed: server closed the connection
5265 unexpectedly". The problem has been fixed by closing all cached lookups
5266 in a queue runner before running a delivery process.
5268 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5269 seem to have the "const" qualifier which it has on other OS. I've
5272 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5273 *of the same type* (client or server) with the same public name that an
5274 error should be diagnosed.
5276 9. When Exim looked up a host name for an IP address, but failed to find the
5277 original IP address when looking up the host name (a safety check), it
5278 output the message "<ip address> does not match any IP for NULL", which was
5279 confusing, to say the least. The bug was that the host name should have
5280 appeared instead of "NULL".
5282 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5283 user that is built into the binary, and the -C or -D options is used, root
5284 privilege is dropped before the configuration file is read. In addition,
5285 logging is switched to stderr instead of the normal log files. If the
5286 configuration then re-defines the Exim user, the unprivileged environment
5287 is probably not what is expected, so Exim logs a panic warning message (but
5290 However, if deliver_drop_privilege is set, the unprivileged state may well
5291 be exactly what is intended, so the warning has been cut out in that case,
5292 and Exim is allowed to try to write to its normal log files.
5298 1. smtp_return_error_details was not giving details for temporary sender
5299 or receiver verification errors.
5301 2. Diagnose a configuration error if two authenticators have the same public
5304 3. Exim used not to create the message log file for a message until the first
5305 delivery attempt. This could be confusing when incoming messages were held
5306 for policy or load reasons. The message log file is now created at the time
5307 the message is received, and an initial "Received" line is written to it.
5309 4. The automatically generated man page for command line options had a minor
5310 bug that caused no ill effects; however, a more serious problem was that
5311 the procedure for building the man page automatically didn't always
5312 operate. Consequently, release 4.20 contains an out-of-date version. This
5313 shouldn't happen again.
5315 5. When building Exim with embedded Perl support, the script that builds the
5316 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5317 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5319 6. The freeze_tell option was not being used for messages that were frozen on
5320 arrival, either by an ACL or by local_scan().
5322 7. Added the smtp_incomplete_transaction log selector.
5324 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5325 was accepting AUTH without a new EHLO.
5327 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5328 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5331 10. Logging of TCP/IP connections (when configured) now happens in the main
5332 daemon process instead of the child process, so that the TCP/IP connection
5333 count is more accurate (but it can never be perfect).
5335 11. The use of "drop" in a nested ACL was not being handled correctly in the
5336 outer ACL. Now, if condition failure induced by the nested "drop" causes
5337 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5338 or "require"), the connection is dropped.
5340 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5341 that yield "discard" can now be used with an "accept" or a "discard" verb,
5342 but an error is generated for any others (because I can't see a useful way
5343 to define what should happen).
5345 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5346 are now processed in the same way as lines in the Exim configuration file.
5347 In particular, continuation lines are supported.
5349 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5351 15. Added -ti meaning -t -i.
5353 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5354 domains, and warn by logging if others are found.
5356 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5357 returned from GIFCONF when Exim is trying to find the list of interfaces on
5358 a host. The code in os.c has been modified to copy each ifreq to an aligned
5359 structure in all cases.
5361 Also, in some cases, the returned ifreq's were being copied to a 'struct
5362 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5363 means the last couple of bytes of an IPv6 address could be chopped if the
5364 ifreq contained only a normal sockaddr (14 bytes storage).
5366 18. Named domain lists were not supported in the hosts_treat_as_local option.
5367 An entry such as +xxxx was not recognized, and was treated as a literal
5370 19. Ensure that header lines added by a DATA ACL are included in the reject log
5371 if the ACL subsequently rejects the message.
5373 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5374 MD5 (which is deprecated).
5376 21. When testing a filter file using -bf, Exim was writing a message when it
5377 took the sender from a "From " line in the message, but it was not doing so
5378 when it took $return_path from a Return-Path: header line. It now does.
5380 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5381 with a valid header line field name (a series of printing characters
5382 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5384 23. Changed "disc" in the source to "disk" to conform to the documentation and
5385 the book and for uniformity.
5387 24. Ignore Sendmail's -Ooption=value command line item.
5389 25. When execve() failed while trying to run a command in a pipe transport,
5390 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5391 could be confused with a return value of 69 from the command itself. This
5392 has been changed to 127, the value the shell returns if it is asked to run
5393 a non-existent command. The wording for the related log line suggests a
5394 non-existent command as the problem.
5396 26. If received_header_text expands to an empty string, do not add a Received:
5397 header line to the message. (Well, it adds a token one on the spool, but
5398 marks it "old" so that it doesn't get used or transmitted.)
5400 27. Installed eximstats 1.28 (addition of -nt option).
5402 28. There was no check for failure on the call to getsockname() in the daemon
5403 code. This can fail if there is a shortage of resources on the system, with
5404 ENOMEM, for example. A temporary error is now given on failure.
5406 29. Contrary to the C standard, it seems that in some environments, the
5407 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5408 program. Exim now does this explicitly; it affects the formatting of
5409 timestamps using strftime().
5411 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5412 complaints. I've now initialized all the variables, to avoid this.
5414 32. Header lines added by a system filter were not being "seen" during
5415 transport-time rewrites.
5417 33. The info_callback() function passed to OpenSSL is set up with type void
5418 (*)(SSL *, int, int), as described somewhere. However, when calling the
5419 function (actually a macro) that sets it up, the type void(*)() is
5420 expected. I've put in a cast to prevent warnings from picky compilers.
5422 34. If a DNS black list lookup found a CNAME record, but there were no A
5423 records associated with the domain it pointed at, Exim crashed.
5425 35. If a DNS black list lookup returned more than one A record, Exim ignored
5426 all but the first. It now scans all returned addresses if a particular IP
5427 value is being sought. In this situation, the contents of the
5428 $dnslist_value variable are a list of all the addresses, separated by a
5431 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5432 to accept a host name and all its aliases if the forward lookup for any of
5433 them yielded the IP address of the incoming connection. Now it accepts only
5434 those names whose forward lookup yields the correct IP address. Any other
5435 names are discarded. This closes a loophole whereby a rogue DNS
5436 administrator could create reverse DNS records to break through a
5437 wildcarded host restriction in an ACL.
5439 37. If a user filter or a system filter that ran in a subprocess used any of
5440 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5441 the wrong values were passed to the pipe command ($thisaddress had the
5442 value of $0, $0 had the value of $1, etc). This bug was introduced by
5443 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5445 38. Improved the line breaking for long SMTP error messages from ACLs.
5446 Previously, if there was no break point between 40 and 75 characters, Exim
5447 left the rest of the message alone. Two changes have been made: (a) I've
5448 reduced the minimum length to 35 characters; (b) if it can't find a break
5449 point between 35 and 75 characters, it looks ahead and uses the first one
5450 that it finds. This may give the occasional overlong line, but at least the
5451 remaining text gets split now.
5453 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5454 file descriptors might be low, and that setting 1000 would always raise it.
5455 It turns out that in some environments, the limit is already over 1000 and
5456 that lowering it causes trouble. So now Exim takes care not to decrease it.
5458 40. When delivering a message, the value of $return_path is set to $sender_
5459 address at the start of routing (routers may change the value). By an
5460 oversight, this default was not being set up when an address was tested by
5461 -bt or -bv, which affected the outcome if any router or filter referred to
5464 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5465 log only when "message" or "log_message" are set. However, if one of the
5466 conditions was an address verification, or a call to a nested ACL, the
5467 messages generated by the underlying test were being passed through. This
5468 no longer happens. The underlying message is available in $acl_verify_
5469 message for both "message" and "log_message" expansions, so it can be
5470 passed through if needed.
5472 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5473 new expansion $bh_ to give the encoded byte string without charset
5474 translation. Translation happens only if iconv() is available; HAVE_ICONV
5475 indicates this at build time. HEADERS_CHARSET gives the charset to
5476 translate to; headers_charset can change it in the configuration, and
5477 "headers charset" can change it in an individual filter file.
5479 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5480 that creates RFC 2047 encoded "words" labels them as that charset instead
5481 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5482 expansion operator; (ii) when Exim creates a From: line for a local
5483 message; (iii) when a header line is rewritten to include a "phrase" part.
5485 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5486 buggy, causing it to skip the first lines of messages whose message ID
5487 ended in 'D'. This would not have bitten before Exim release 4.14, because
5488 message IDs were unlikely to end in 'D' before then. The effect was to have
5489 incorrect size information for certain domains.
5491 45. #include "config.h" was missing at the start of the crypt16.c module. This
5492 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5495 46. If there was a timeout during a "random" callout check, Exim treated it as
5496 a failure of the random address, and carried on sending RSET and the real
5497 address. If the delay was just some slowness somewhere, the response to the
5498 original RCPT would be taken as a response to RSET and so on, causing
5499 mayhem of various kinds.
5501 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5502 when I implemented it. It didn't allow for the fact that some option values
5503 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5504 the right test for positive values.
5506 48. Domain names in DNS records are case-independent. Exim always looks them up
5507 in lower case. Some resolvers return domain names in exactly the case they
5508 appear in the zone file, that is, they may contain uppercase letters. Not
5509 all resolvers do this - some return always lower case. Exim was treating a
5510 change of case by a resolver as a change of domain, similar to a widening
5511 of a domain abbreviation. This triggered its re-routing code and so it was
5512 trying to route what was effectively the same domain again. This normally
5513 caused routing to fail (because the router wouldn't handle the domain
5514 twice). Now Exim checks for this case specially, and just changes the
5515 casing of the domain that it ultimately uses when it transmits the message
5518 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5521 50. If a filter generated a file delivery with a non-absolute name (possible if
5522 no home directory exists for the router), the forbid_file option was not
5525 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5526 the existing equality matching.
5528 52. Exim was using ints instead of ino_t variables in some places where it was
5529 dealing with inode numbers.
5531 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5532 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5533 called TMPDIR, and if it finds it is different, it changes its value.
5535 54. The smtp_printf() function is now made available to local_scan() so
5536 additional output lines can be written before returning. There is also an
5537 smtp_fflush() function to enable the detection of a dropped connection.
5538 The variables smtp_input and smtp_batched_input are exported to
5541 55. Changed the default runtime configuration: the message "Unknown user"
5542 has been removed from the ACL, and instead placed on the localuser router,
5543 using the cannot_route_message feature. This means that any verification
5544 failures that generate their own messages won't get overridden. Similarly,
5545 the "Unrouteable address" message that was in the ACL for unverifiable
5546 relay addresses has also been removed.
5548 56. Added hosts_avoid_esmtp to the smtp transport.
5550 57. The exicyclog script was not checking for the esoteric option
5551 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5552 will work only if exicyclog is run under the appropriate euid.
5554 58. Following a discussion on the list, the rules by which Exim recognises line
5555 endings on incoming messages have been changed. The -dropcr and drop_cr
5556 options are now no-ops, retained only for backwards compatibility. The
5557 following line terminators are recognized: LF CRLF CR. However, special
5558 processing applies to CR:
5560 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5561 nor a local message in the state where . is a terminator.
5563 (ii) If a bare CR is encountered in a header line, an extra space is added
5564 after the line terminator so as not to end the header. The reasoning
5565 behind this is that bare CRs in header lines are most likely either
5566 to be mistakes, or people trying to play silly games.
5568 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5569 was being incorrectly given as 18 bytes larger than it should have been.
5570 This is a VOB (very old bug).
5572 60. This may never have affected anything current, but just in case it has:
5573 When the local host is found other than at the start of a list of hosts,
5574 the local host, those with the same MX, and any that follow, are discarded.
5575 When the list in question was part of a longer list of hosts, the following
5576 hosts (not currently being processed) were also being discarded. This no
5577 longer happens. I'm not sure if this situation could ever has previously
5580 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5583 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5584 header. This is something that is recommended in a new Internet Draft, and
5585 is something that is documented as being done by Sendmail. There are two
5586 possible values. For messages generated by the autoreply transport, Exim
5589 Auto-Submitted: auto-replied
5591 whereas for all other generated messages (e.g. bounces) it adds
5593 Auto-Submitted: auto-generated
5595 63. The "personal" condition in filters now includes a test for the
5596 Auto-Submitted: header. If it contains the string "auto-" the message it
5597 not considered personal.
5599 64. Added rcpt_include_affixes as a generic transport option.
5601 65. Added queue_only_override (default true).
5603 66. Added the syslog_duplication option.
5605 67. If what should have been the first header line of a message consisted of
5606 a space followed by a colon, Exim was mis-interpreting it as a header line.
5607 It isn't of course - it is syntactically invalid and should therefore be
5608 treated as the start of the message body. The misbehaviour could have
5609 caused a number of strange effects, including loss of data in subsequent
5610 header lines, and spool format errors.
5612 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5613 client host had authenticated. This control can now be exercised by an ACL
5614 for more flexibility.
5616 69. By default, callouts do not happen when testing with -bh. There is now a
5617 variant, -bhc, which does actually run the callout code, including
5618 consulting and updating the callout cache.
5620 70. Added support for saslauthd authentication, courtesy of Alexander
5623 71. If statvfs() failed on the spool or log directories while checking their
5624 size for availability, Exim confusingly gave the error "space shortage".
5625 Furthermore, in debugging mode it crashed with a floating point exception.
5626 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5627 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5628 serious problem, Exim now writes to the main and panic logs when this
5629 happens, with details of the failure. It then refuses to accept the
5630 incoming message, giving the message "spool directory problem" or "log
5631 directory problem" with a 421 code for SMTP messages.
5633 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5634 0, 1, and 2 exist, because some OS complain for execs without them (see
5635 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5636 descriptors. However, the code omitted to check that the open succeeded,
5637 causing mysterious errors if for some reason the permissions on /dev/null
5638 got screwed. Now Exim writes a message to the main and panic logs, and
5639 bombs out if it can't open /dev/null.
5641 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5642 interact so that it is all more flexible. It is supposed to remain
5643 backwards compatible. Also added extra_local_interfaces.
5645 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5646 to bomb out with an assertion failure - to the client this appears as a
5647 connection drop. This problem occurs in the part of the code that was taken
5648 from the Samba project. Fortunately, the assertion is in a very simple
5649 function, so I have fixed this by reproducing the function inline in the
5650 one place where it is called, and arranging for authentication to fail
5651 instead of killing the process with assert().
5653 75. The SPA client code was not working when the server requested OEM rather
5654 than Unicode encoding.
5656 76. Added code to make require_files with a specific uid setting more usable in
5657 the case where statting the file as root fails - usually a non-root-mounted
5658 NFS file system. When this happens and the failure is EACCES, Exim now
5659 forks a subprocess and does the per-uid checking as the relevant uid.
5661 77. Added process_log_path.
5663 78. If log_file_path was not explicitly set, a setting of check_log_space or
5664 check_log_inodes was ignored.
5666 79. If a space check for the spool or log partitions fails, the incident is now
5667 logged. Of course, in the latter case the data may get lost...
5669 80. Added the %p formatting code to string_format() so that it can be used to
5670 print addresses in debug_print(). Adjusted all the address printing in the
5671 debugging in store.c to use %p rather than %d.
5673 81. There was a concern that a line of code in smtp_in.c could overflow a
5674 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5675 initially expressed, the concern was not well-founded, because trailing
5676 spaces are removed early. However, if the trailing spaces were followed by
5677 a NULL, they did not get removed, so the overflow was possible. Two fixes
5680 (a) I re-wrote the offending code in a cleaner fashion.
5681 (b) If an incoming SMTP command contains a NULL character, it is rejected
5684 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5685 runs initgroups(), so that if the Exim user is in any additional groups,
5686 they will be used during message reception.
5692 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob