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}
47 PP/09 Add gnutls_enable_pkcs11 option.
49 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
50 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
52 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
53 "acl {{name}{arg}...}", and optional args on acl condition
59 PP/01 Handle short writes when writing local log-files.
60 In practice, only affects FreeBSD (8 onwards).
61 Bugzilla 1053, with thanks to Dmitry Isaikin.
63 NM/01 Bugzilla 949 - Documentation tweak
65 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
68 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
70 PP/02 Implemented gsasl authenticator.
72 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
74 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
75 `pkg-config foo` for cflags/libs.
77 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
78 with rest of GSASL and with heimdal_gssapi.
80 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
81 `pkg-config foo` for cflags/libs for the TLS implementation.
83 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
84 properties get this fed in as external SSF. A number of robustness
85 and debugging improvements to the cyrus_sasl authenticator.
87 PP/08 cyrus_sasl server now expands the server_realm option.
89 PP/09 Bugzilla 1214 - Log authentication information in reject log.
90 Patch by Jeremy Harris.
92 PP/10 Added dbmjz lookup type.
94 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
96 PP/12 MAIL args handles TAB as well as SP, for better interop with
97 non-compliant senders.
98 Analysis and variant patch by Todd Lyons.
100 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
101 Bug report from Lars Müller <lars@samba.org> (via SUSE),
102 Patch from Dirk Mueller <dmueller@suse.com>
104 PP/13 tls_peerdn now print-escaped for spool files.
105 Observed some $tls_peerdn in wild which contained \n, which resulted
106 in spool file corruption.
108 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
109 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
110 or write after TLS renegotiation, which otherwise led to messages
113 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
114 as a tracking header (ie: a signed header comes before the signature).
115 Patch from Wolfgang Breyha.
117 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
118 comma-sep list; embedded commas doubled.
120 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
122 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
124 Report and patch from Dmitry Banschikov.
126 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
127 Removed SSL_clear() after SSL_new() which led to protocol negotiation
128 failures. We appear to now support TLS1.1+ with Exim.
130 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
131 lets Exim select keys and certificates based upon TLS SNI from client.
132 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
133 before an outbound SMTP session. New log_selector, +tls_sni.
135 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
136 NULL dereference. Report and patch from Alun Jones.
138 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
139 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
140 Not seeing resolver debug output on NetBSD, but suspect this is a
141 resolver implementation change.
143 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
144 Left warnings. Added "eximon gdb" invocation mode.
146 PP/21 Defaulting "accept_8bitmime" to true, not false.
148 PP/22 Added -bw for inetd wait mode support.
150 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
151 locate the relevant includes and libraries. Made this the default.
153 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
154 Bugzilla 1246, report and most of solution from Tomasz Kusy.
156 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
157 This may cause build issues on older platforms.
159 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
160 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
161 gnutls_require_mac & gnutls_require_protocols (no longer supported).
162 Added SNI support via GnuTLS too.
163 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
165 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
167 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
168 Applied second patch from Janne, implementing suggestion to default
169 multiple-strings-in-record handling to match SPF spec.
171 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
173 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
174 read-only, out of scope).
175 Patch from Wolfgang Breyha, report from Stuart Northfield.
177 PP/29 Fix three issues highlighted by clang analyser static analysis.
178 Only crash-plausible issue would require the Cambridge-specific
179 iplookup router and a misconfiguration.
180 Report from Marcin Mirosław.
182 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
184 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
185 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
186 As part of this, removing so much warning spew let me fix some minor
187 real issues in debug logging.
189 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
190 assignment on my part. Fixed.
192 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
193 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
194 Janne Snabb (who went above and beyond: thank you).
196 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
197 string otherwise requires a connection and a bunch more work and it's
198 relatively easy to get wrong. Should also expose TLS library linkage
201 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
202 64-bit ${eval} (JH/03).
204 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
205 GNU libc to support some of the 64-bit stuff, should not lead to
206 conflicts. Defined before os.h is pulled in, so if a given platform
207 needs to override this, it can.
209 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
210 protection layer was required, which is not implemented.
211 Bugzilla 1254, patch from Wolfgang Breyha.
213 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
214 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
215 tls_dhparam take prime identifiers. Also unbreak combination of
216 OpenSSL+DH_params+TLSSNI.
218 PP/39 Disable SSLv2 by default in OpenSSL support.
224 PP/01 Solaris build fix for Oracle's LDAP libraries.
225 Bugzilla 1109, patch from Stephen Usher.
227 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
229 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
232 TF/02 Fix a couple more cases where we did not log the error message
233 when unlink() failed. See also change 4.74-TF/03.
235 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
236 lock up or crash if it happened to be inside a call to libc when it
237 got a SIGUSR1 from exiwhat.
239 The SIGUSR1 handler appends the current process status to the process
240 log which is later printed by exiwhat. It used to use the general
241 purpose logging code to do this, but several functions it calls are
242 not safe for signals.
244 The new output code in the SIGUSR1 handler is specific to the process
245 log, and simple enough that it's easy to inspect for signal safety.
246 Removing some special cases also simplifies the general logging code.
247 Removing the spurious timestamps from the process log simplifies
250 TF/04 Improved ratelimit ACL condition.
252 The /noupdate option has been deprecated in favour of /readonly which
253 has clearer semantics. The /leaky, /strict, and /readonly update modes
254 are mutually exclusive. The update mode is no longer included in the
255 database key; it just determines when the database is updated. (This
256 means that when you upgrde Exim will forget old rate measurements.)
258 Exim now checks that the per_* options are used with an update mode that
259 makes sense for the current ACL. For example, when Exim is processing a
260 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
261 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
262 must specify per_mail/readonly. If you omit the update mode it defaults to
263 /leaky where that makes sense (as before) or /readonly where required.
265 The /noupdate option is now undocumented but still supported for
266 backwards compatibility. It is equivalent to /readonly except that in
267 ACLs where /readonly is required you may specify /leaky/noupdate or
268 /strict/noupdate which are treated the same as /readonly.
270 A useful new feature is the /count= option. This is a generalization
271 of the per_byte option, so that you can measure the throughput of other
272 aggregate values. For example, the per_byte option is now equivalent
273 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
275 The per_rcpt option has been generalized using the /count= mechanism
276 (though it's more complicated than the per_byte equivalence). When it is
277 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
278 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
279 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
280 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
281 Note that using per_rcpt with a non-readonly update mode in more than
282 one ACL will cause the recipients to be double-counted. (The per_mail
283 and per_byte options don't have this problem.)
285 The handling of very low rates has changed slightly. If the computed rate
286 is less than the event's count (usually one) then this event is the first
287 after a long gap. In this case the rate is set to the same as this event's
288 count, so that the first message of a spam run is counted properly.
290 The major new feature is a mechanism for counting the rate of unique
291 events. The new per_addr option counts the number of different
292 recipients that someone has sent messages to in the last time period. It
293 behaves like per_rcpt if all the recipient addresses are different, but
294 duplicate recipient addresses do not increase the measured rate. Like
295 the /count= option this is a general mechanism, so the per_addr option
296 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
297 example, measure the rate that a client uses different sender addresses
298 with the options per_mail/unique=$sender_address. There are further
299 details in the main documentation.
301 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
303 TF/06 Removed a few PCRE remnants.
305 TF/07 Automatically extract Exim's version number from tags in the git
306 repository when doing development or release builds.
308 PP/02 Raise smtp_cmd_buffer_size to 16kB.
309 Bugzilla 879. Patch from Paul Fisher.
311 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
312 Heavily based on revision 40f9a89a from Simon Arlott's tree.
315 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
317 PP/05 Variable $av_failed, true if the AV scanner deferred.
318 Bugzilla 1078. Patch from John Horne.
320 PP/06 Stop make process more reliably on build failure.
321 Bugzilla 1087. Patch from Heiko Schlittermann.
323 PP/07 Make maildir_use_size_file an _expandable_ boolean.
324 Bugzilla 1089. Patch from Heiko Schlittermann.
326 PP/08 Handle ${run} returning more data than OS pipe buffer size.
327 Bugzilla 1131. Patch from Holger Weiß.
329 PP/09 Handle IPv6 addresses with SPF.
330 Bugzilla 860. Patch from Wolfgang Breyha.
332 PP/10 GnuTLS: support TLS 1.2 & 1.1.
334 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
337 PP/11 match_* no longer expand right-hand-side by default.
338 New compile-time build option, EXPAND_LISTMATCH_RHS.
339 New expansion conditions, "inlist", "inlisti".
341 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
343 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
345 PP/14 fix log_write() format string regression from TF/03.
346 Bugzilla 1152. Patch from Dmitry Isaikin.
352 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
354 PP/02 Harmonised TLS library version reporting; only show if debugging.
355 Layout now matches that introduced for other libraries in 4.74 PP/03.
357 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
359 PP/04 New "dns_use_edns0" global option.
361 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
364 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
365 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
367 TK/01 Updated PolarSSL code to 0.14.2.
368 Bugzilla 1097. Patch from Andreas Metzler.
370 PP/07 Catch divide-by-zero in ${eval:...}.
373 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
376 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
377 format-string attack -- SECURITY: remote arbitrary code execution.
379 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
380 time unintentionally subject to list matching rules, letting the header
381 cause arbitrary Exim lookups (of items which can occur in lists, *not*
382 arbitrary string expansion). This allowed for information disclosure.
384 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
385 INT_MIN/-1 -- value coerced to INT_MAX.
391 NM/01 Workround for PCRE version dependency in version reporting
394 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
395 This fixes portability to compilers other than gcc, notably
396 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
398 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
399 makefiles for portability to HP-UX and POSIX correctness.
401 PP/01 Permit LOOKUP_foo enabling on the make command-line.
402 Also via indented variable definition in the Makefile.
403 (Debugging by Oliver Heesakkers).
405 PP/02 Restore caching of spamd results with expanded spamd_address.
406 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
408 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
409 Improves build reliability. Fix from: Frank Elsner
411 NM/02 Fix wide character breakage in the rfc2047 coding
412 Fixes bug 1064. Patch from Andrey N. Oktyabrski
414 NM/03 Allow underscore in dnslist lookups
415 Fixes bug 1026. Patch from Graeme Fowler
417 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
418 Code patches from Adam Ciarcinski of NetBSD.
420 NM/04 Fixed exiqgrep to cope with mailq missing size issue
423 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
424 is logged, to avoid truncation. Patch from John Horne.
426 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
427 Patch from Jakob Hirsch.
429 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
430 SQL string expansion failure details.
431 Patch from Andrey Oktyabrski.
433 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
434 Patch from Simon Arlott.
436 PP/09 New lookups functionality failed to compile on old gcc which rejects
437 extern declarations in function scope.
438 Patch from Oliver Fleischmann
440 PP/10 Use sig_atomic_t for flags set from signal handlers.
441 Check getgroups() return and improve debugging.
442 Fixed developed for diagnosis in bug 927 (which turned out to be
445 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
446 Patch from Mark Zealey.
448 PP/12 Bugzilla 1056: Improved spamd server selection.
449 Patch from Mark Zealey.
451 PP/13 Bugzilla 1086: Deal with maildir quota file races.
452 Based on patch from Heiko Schlittermann.
454 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
455 Patch from Uwe Doering, sign-off by Michael Haardt.
457 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
458 variable declaration deep within a block. Bug and patch from
461 PP/15 lookups-Makefile IRIX compatibilty coercion.
463 PP/16 Make DISABLE_DKIM build knob functional.
465 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
466 Patch by Simon Arlott
468 TF/03 Fix valgrind.h portability to C89 compilers that do not support
469 variable argument macros. Our copy now differs from upstream.
475 TF/01 Failure to get a lock on a hints database can have serious
476 consequences so log it to the panic log.
478 TF/02 Log LMTP confirmation messages in the same way as SMTP,
479 controlled using the smtp_confirmation log selector.
481 TF/03 Include the error message when we fail to unlink a spool file.
483 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
484 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
485 for maintaining out-of-tree patches for some time.
487 PP/01 Bugzilla 139: Documentation and portability issues.
488 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
489 Handle per-OS dynamic-module compilation flags.
491 PP/02 Let /dev/null have normal permissions.
492 The 4.73 fixes were a little too stringent and complained about the
493 permissions on /dev/null. Exempt it from some checks.
494 Reported by Andreas M. Kirchwitz.
496 PP/03 Report version information for many libraries, including
497 Exim version information for dynamically loaded libraries. Created
498 version.h, now support a version extension string for distributors
499 who patch heavily. Dynamic module ABI change.
501 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
502 privilege escalation vulnerability whereby the Exim run-time user
503 can cause root to append content of the attacker's choosing to
506 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
509 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
510 If dropping privileges for untrusted macros, we disabled normal logging
511 on the basis that it would fail; for the Exim run-time user, this is not
512 the case, and it resulted in successful deliveries going unlogged.
513 Fixed. Reported by Andreas Metzler.
519 PP/01 Date: & Message-Id: revert to normally being appended to a message,
520 only prepend for the Resent-* case. Fixes regression introduced in
521 Exim 4.70 by NM/22 for Bugzilla 607.
523 PP/02 Include check_rfc2047_length in configure.default because we're seeing
524 increasing numbers of administrators be bitten by this.
526 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
528 PP/03 Bugzilla 994: added openssl_options main configuration option.
530 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
532 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
534 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
536 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
537 without a peer certificate, leading to a segfault because of an
538 assumption that peers always have certificates. Be a little more
539 paranoid. Problem reported by Martin Tscholak.
541 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
542 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
543 NB: ClamAV planning to remove STREAM in "middle of 2010".
544 CL also introduces -bmalware, various -d+acl logging additions and
545 more caution in buffer sizes.
547 PP/09 Implemented reverse_ip expansion operator.
549 PP/10 Bugzilla 937: provide a "debug" ACL control.
551 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
553 PP/12 Bugzilla 973: Implement --version.
555 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
557 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
559 PP/15 Bugzilla 816: support multiple condition rules on Routers.
561 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
562 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
563 ignore trailing whitespace.
565 JJ/02 prevent non-panic DKIM error from being sent to paniclog
567 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
570 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
571 Notification from Dr Andrew Aitchison.
573 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
574 ExtendedDetectionInfo response format.
575 Notification from John Horne.
577 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
580 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
581 XSL and documented dependency on system catalogs, with examples of how
584 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
587 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
588 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
589 configuration file which is writeable by the Exim user or group.
591 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
592 of configuration files to cover files specified with the -C option if
593 they are going to be used with root privileges, not just the default
596 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
597 option (effectively making it always true).
599 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
600 files to be used while preserving root privileges.
602 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
603 that rogue child processes cannot use them.
605 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
606 run-time user, instead of root.
608 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
609 Exim run-time user without dropping privileges.
611 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
612 result string, instead of calling string_vformat() twice with the same
615 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
616 for other users. Others should always drop root privileges if they use
617 -C on the command line, even for a whitelisted configure file.
619 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
621 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
627 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
628 $data_path, and $header_path variables; fixed documentation bugs and
631 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
632 exipick to access non-standard spools, including the "frozen" queue
635 NM/01 Bugzilla 965: Support mysql stored procedures.
636 Patch from Alain Williams
638 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
640 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
641 Patch from Andreas Metzler
643 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
644 Patch from Kirill Miazine
646 NM/05 Bugzilla 671: Added umask to procmail example.
648 JJ/03 installed exipick 20100323.0, fixing doc bug
650 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
651 directory. Notification and patch from Dan Rosenberg.
653 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
655 TK/02 Improve log output when DKIM signing operation fails.
657 MH/01 Treat the transport option dkim_domain as a colon separated
658 list, not as a single string, and sign the message with each element,
659 omitting multiple occurences of the same signer.
661 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
662 Bugzilla 985, 986. Patch by Simon Arlott
664 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
665 Patch by Simon Arlott
667 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
668 MBX locking. Notification from Dan Rosenberg.
674 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
676 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
678 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
680 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
682 NM/04 Bugzilla 915: Flag broken perl installation during build.
688 TK/01 Added patch by Johannes Berg that expands the main option
689 "spamd_address" if it starts with a dollar sign.
691 TK/02 Write list of recipients to X-Envelope-Sender header when building
692 the mbox-format spool file for content scanning (suggested by Jakob
695 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
696 (http://www.dcc-servers.net/) support via dccifd. Activated by
697 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
699 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
700 by Mark Daniel Reidel <mr@df.eu>.
702 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
703 When building exim an external PCRE library is now needed -
704 PCRE is a system library on the majority of modern systems.
705 See entry on PCRE_LIBS in EDITME file.
707 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
708 conversation. Added nologin parameter to request.
709 Patch contributed by Kirill Miazine.
711 TF/01 Do not log submission mode rewrites if they do not change the address.
713 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
715 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
716 log files in place. Contributed by Roberto Lima.
718 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
720 TF/03 Bugzilla 615: When checking the local_parts router precondition
721 after a local_part_suffix or local_part_prefix option, Exim now
722 does not use the address's named list lookup cache, since this
723 contains cached lookups for the whole local part.
725 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
726 Robert Millan. Documentation is in experimental-spec.txt.
728 TF/04 Bugzilla 668: Fix parallel build (make -j).
730 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
732 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
733 Patch provided by Jan Srzednicki.
735 TF/05 Leading white space used to be stripped from $spam_report which
736 wrecked the formatting. Now it is preserved.
738 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
739 that they are available at delivery time.
741 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
743 TF/08 TLS error reporting now respects the incoming_interface and
744 incoming_port log selectors.
746 TF/09 Produce a more useful error message if an SMTP transport's hosts
747 setting expands to an empty string.
749 NM/06 Bugzilla 744: EXPN did not work under TLS.
750 Patch provided by Phil Pennock.
752 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
753 Patch provided by Richard Godbee.
755 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
756 acl_smtp_notquit, added index entry.
758 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
759 Patch provided by Eugene Bujak.
761 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
762 accept(). Patch provided by Maxim Dounin.
764 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
765 Patch provided by Phil Pennock.
767 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
769 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
770 Patch provided by Brad "anomie" Jorsch.
772 NM/14 Bugzilla 622: Added timeout setting to transport filter.
773 Patch provided by Dean Brooks.
775 TK/05 Add native DKIM support (does not depend on external libraries).
777 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
778 Patch provided by Graeme Fowler.
780 NM/16 Bugzilla 851: Documentation example syntax fix.
782 NM/17 Changed NOTICE file to remove references to embedded PCRE.
784 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
787 NM/19 Bugzilla 745: TLS version reporting.
788 Patch provided by Phil Pennock.
790 NM/20 Bugzilla 167: bool: condition support.
791 Patch provided by Phil Pennock.
793 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
794 clients. Patch provided by Phil Pennock.
796 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
797 Patch provided by Brad "anomie" Jorsch.
799 NM/23 Bugzilla 687: Fix misparses in eximstats.
800 Patch provided by Heiko Schlittermann.
802 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
803 Patch provided by Heiko Schlittermann.
805 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
806 plus update to original patch.
808 NM/26 Bugzilla 799: Documentation correction for ratelimit.
810 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
811 Patch provided by David Brownlee.
813 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
815 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
817 NM/30 Bugzilla 888: TLS documentation bugfixes.
819 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
821 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
822 Unlike the original bugzilla I have changed all shell scripts in src tree.
824 NM/33 Bugzilla 898: Transport filter timeout fix.
825 Patch by Todd Rinaldo.
827 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
828 Patch by Serge Demonchaux.
830 NM/35 Bugzilla 39: Base64 decode bug fixes.
831 Patch by Jakob Hirsch.
833 NM/36 Bugzilla 909: Correct connect() call in dcc code.
835 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
837 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
839 NM/39 Bugzilla 911: Fixed MakeLinks build script.
845 TK/01 Add preliminary DKIM support. Currently requires a forked version of
846 ALT-N's libdkim that I have put here:
847 http://duncanthrax.net/exim-experimental/
849 Note to Michael Haardt: I had to rename some vars in sieve.c. They
850 were called 'true' and it seems that C99 defines that as a reserved
851 keyword to be used with 'bool' variable types. That means you could
852 not include C99-style headers which use bools without triggering
853 build errors in sieve.c.
855 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
856 as mailq or other aliases. Changed the --help handling significantly
857 to do whats expected. exim_usage() emits usage/help information.
859 SC/01 Added the -bylocaldomain option to eximstats.
861 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
863 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
865 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
871 PH/01 Another patch from the Sieve maintainer.
873 PH/02 When an IPv6 address is converted to a string for single-key lookup
874 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
875 dots are used instead of colons so that keys in lsearch files need not
876 contain colons. This was done some time before quoting was made available
877 in lsearch files. However, iplsearch files do require colons in IPv6 keys
878 (notated using the quote facility) so as to distinguish them from IPv4
879 keys. This meant that lookups for IP addresses in host lists did not work
880 for iplsearch lookups.
882 This has been fixed by arranging for IPv6 addresses to be expressed with
883 colons if the lookup type is iplsearch. This is not incompatible, because
884 previously such lookups could never work.
886 The situation is now rather anomolous, since one *can* have colons in
887 ordinary lsearch keys. However, making the change in all cases is
888 incompatible and would probably break a number of configurations.
890 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
893 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
894 conversion specification without a maximum field width, thereby enabling
895 a rogue spamd server to cause a buffer overflow. While nobody in their
896 right mind would setup Exim to query an untrusted spamd server, an
897 attacker that gains access to a server running spamd could potentially
898 exploit this vulnerability to run arbitrary code as the Exim user.
900 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
901 $primary_hostname instead of what libspf2 thinks the hosts name is.
903 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
904 a directory entry by the name of the lookup key. Previously, if a
905 symlink pointed to a non-existing file or a file in a directory that
906 Exim lacked permissions to read, a lookup for a key matching that
907 symlink would fail. Now it is enough that a matching directory entry
908 exists, symlink or not. (Bugzilla 503.)
910 PH/03 The body_linecount and body_zerocount variables are now exported in the
913 PH/04 Added the $dnslist_matched variable.
915 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
916 This means they are set thereafter only if the connection becomes
919 PH/06 Added the client_condition to authenticators so that some can be skipped
920 by clients under certain conditions.
922 PH/07 The error message for a badly-placed control=no_multiline_responses left
923 "_responses" off the end of the name.
925 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
927 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
928 (without spaces) instead of just copying the configuration text.
930 PH/10 Added the /noupdate option to the ratelimit ACL condition.
932 PH/11 Added $max_received_linelength.
934 PH/12 Added +ignore_defer and +include_defer to host lists.
936 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
937 way in which PCRE > 7.0 is built.
939 PH/14 Implemented queue_only_load_latch.
941 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
942 MAIL command. The effect was to mangle the value on 64-bit systems.
944 PH/16 Another patch from the Sieve maintainer.
946 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
948 PH/18 If a system quota error occurred while trying to create the file for
949 a maildir delivery, the message "Mailbox is full" was not appended to the
950 bounce if the delivery eventually timed out. Change 4.67/27 below applied
951 only to a quota excession during the actual writing of the file.
953 PH/19 It seems that peer DN values may contain newlines (and other non-printing
954 characters?) which causes problems in log lines. The DN values are now
955 passed through string_printing() before being added to log lines.
957 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
958 and InterBase are left for another time.)
960 PH/21 Added message_body_newlines option.
962 PH/22 Guard against possible overflow in moan_check_errorcopy().
964 PH/23 POSIX allows open() to be a macro; guard against that.
966 PH/24 If the recipient of an error message contained an @ in the local part
967 (suitably quoted, of course), incorrect values were put in $domain and
968 $local_part during the evaluation of errors_copy.
974 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
975 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
978 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
979 issue a MAIL command.
981 PH/02 In an ACL statement such as
983 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
985 if a client was not listed at all, or was listed with a value other than
986 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
987 the condition was not true (as it should be), so access was not denied.
988 The bug was that the ! inversion was incorrectly passed on to the second
989 item. This has been fixed.
991 PH/03 Added additional dnslists conditions == and =& which are different from
992 = and & when the dns lookup returns more than one IP address.
994 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
995 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
997 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
998 FSYNC, which compiles an option called disable_fsync that allows for
999 bypassing fsync(). The documentation is heavily laced with warnings.
1001 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1003 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1004 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1005 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1006 including adding "make clean"; (3) Added -fPIC when compiling the test
1007 dynamically loaded module, to get rid of a warning.
1009 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1010 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1011 = 0s. The bug is otherwise harmless.
1013 PH/07 There was a bug in the dovecot authenticator such that the value of
1014 $auth1 could be overwritten, and so not correctly preserved, after a
1015 successful authentication. This usually meant that the value preserved by
1016 the server_setid option was incorrect.
1018 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1020 PH/09 Installed PCRE release 7.0.
1022 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1023 run for batched SMTP input. It is now run at the start of every message
1024 in the batch. While fixing this I discovered that the process information
1025 (output by running exiwhat) was not always getting set for -bs and -bS
1026 input. This is fixed, and it now also says "batched" for BSMTP.
1028 PH/11 Added control=no_pipelining.
1030 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1031 patch, slightly modified), and move the expansion of helo_data till after
1032 the connection is made in the smtp transport (so it can use these
1035 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1037 PH/14 Added log_selector = +pid.
1039 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1041 PH/16 Add ${if forany and ${if forall.
1043 PH/17 Added dsn_from option to vary the From: line in DSNs.
1045 PH/18 Flush SMTP output before performing a callout, unless control =
1046 no_callout_flush is set.
1048 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1049 was true (the default) a successful delivery failed to delete the retry
1050 item, thus causing premature timeout of the address. The bug is now
1053 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1055 PH/21 Long custom messages for fakedefer and fakereject are now split up
1056 into multiline reponses in the same way that messages for "deny" and
1057 other ACL rejections are.
1059 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1060 with slight modification.
1062 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1063 draft, changing the syntax and factoring some duplicate code".
1065 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1066 for deliveries of the second and subsequent messages over the same SMTP
1069 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1070 ${reduce, with only minor "tidies".
1072 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1074 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1075 expansion side effects.
1077 PH/27 When a message times out after an over-quota error from an Exim-imposed
1078 quota, the bounce message says "mailbox is full". This message was not
1079 being given when it was a system quota that was exceeded. It now should
1082 MH/03 Made $recipients available in local_scan(). local_scan() already has
1083 better access to the recipient list through recipients_list[], but
1084 $recipients can be useful in postmaster-provided expansion strings.
1086 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1087 in the case of a MAIL command with additional options following the
1088 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1089 were accidentally chopped off.
1091 PH/29 SMTP synchronization checks are implemented when a command is read -
1092 there is a check that no more input is waiting when there shouldn't be
1093 any. However, for some commands, a delay in an ACL can mean that it is
1094 some time before the response is written. In this time, more input might
1095 arrive, invalidly. So now there are extra checks after an ACL has run for
1096 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1097 pipelining has not been advertised.
1099 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1101 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1102 returned in the SMTP response when smtp_return_error_details was false.
1103 This has been fixed.
1105 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1106 instead of the C I/O that was originally supplied, because problems were
1107 reported on Solaris.
1109 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1110 Exim which did not show up earlier: it was assuming that a call to
1111 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1112 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1113 was a macro that became an assignment, so it seemed to work. This has
1114 changed to a proper function call with a void return, hence the compile
1115 error. Exim's code has been fixed.
1117 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1120 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1122 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1124 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1125 because it turns out to be too big or there is a timeout) there is no
1126 "Completed" line in the log. When some messages of this type were
1127 selected by exigrep, they were listed as "not completed". Others were
1128 picked up by some special patterns. I have improved the selection
1129 criteria to be more general.
1131 PH/38 The host_find_failed option in the manualroute router can now be set
1132 to "ignore", to completely ignore a host whose IP address cannot be
1133 found. If all hosts are ignored, the behaviour is controlled by the new
1134 host_all_ignored option.
1136 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1137 homing or because of multiple MX records with /mx) generated more than
1138 one IP address, and the following item turned out to be the local host,
1139 all the secondary addresses of the first item were incorrectly removed
1140 from the list, along with the local host and any following hosts (which
1141 is what is supposed to happen).
1143 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1144 whoever called Exim into the -H file. In the case of the daemon it was
1145 behaving confusingly. When first started, it used values for whoever
1146 started the daemon, but after a SIGHUP it used the Exim user (because it
1147 calls itself on a restart). I have changed the code so that it now always
1150 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1151 message are rejected with the same error (e.g. no authentication or bad
1152 sender address), and a DATA command is nevertheless sent (as can happen
1153 with PIPELINING or a stupid MUA), the error message that was given to the
1154 RCPT commands is included in the rejection of the DATA command. This is
1155 intended to be helpful for MUAs that show only the final error to their
1158 PH/42 Another patch from the Sieve maintainer.
1160 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1161 Eximstats - Fixed some broken HTML links and added missing column headers
1163 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1164 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1166 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1168 PH/43 Yet another patch from the Sieve maintainer.
1170 PH/44 I found a way to check for a TCP/IP connection going away before sending
1171 the response to the final '.' that terminates a message, but only in the
1172 case where the client has not sent further data following the '.'
1173 (unfortunately, this is allowed). However, in many cases there won't be
1174 any further data because there won't be any more messages to send. A call
1175 to select() can be used: if it shows that the input is "ready", there is
1176 either input waiting, or the socket has been closed. An attempt to read
1177 the next input character can distinguish the two cases. Previously, Exim
1178 would have sent an OK response which the client would never have see.
1179 This could lead to message repetition. This fix should cure that, at
1180 least in a lot of common cases.
1182 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1183 advertised in response to EHLO.
1189 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1190 fixed by 4.65/MH/01 (is this a record?) are fixed:
1192 (i) An empty string was always treated as zero by the numeric comparison
1193 operators. This behaviour has been restored.
1195 (ii) It is documented that the numeric comparison operators always treat
1196 their arguments as decimal numbers. This was broken in that numbers
1197 starting with 0 were being interpreted as octal.
1199 While fixing these problems I realized that there was another issue that
1200 hadn't been noticed. Values of message_size_limit (both the global option
1201 and the transport option) were treated as octal if they started with 0.
1202 The documentation was vague. These values are now always treated as
1203 decimal, and I will make that clear in the documentation.
1209 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1210 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1213 MH/01 Don't check that the operands of numeric comparison operators are
1214 integers when their expansion is in "skipping" mode (fixes bug
1215 introduced by 4.64-PH/07).
1217 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1218 child addresses, Exim now panics and dies. Previously, because the count
1219 is held in a short int, deliveries were likely to be lost. As such a
1220 large number of recipients for a single message is ridiculous
1221 (performance will be very, very poor), I have chosen to impose a limit
1222 rather than extend the field.
1228 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1229 leftover -K file (the existence of which was triggered by #402).
1230 While we were at it, introduced process PID as part of the -K
1231 filename. This should rule out race conditions when creating
1234 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1235 processing considerably. Previous code took too long for large mails,
1236 triggering a timeout which in turn triggers #401.
1238 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1239 in the DK code in transports.c. sendfile() is not really portable,
1240 hence the _LINUX specificness.
1242 TF/01 In the add_headers option to the mail command in an Exim filter,
1243 there was a bug that Exim would claim a syntax error in any
1244 header after the first one which had an odd number of characters
1247 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1248 callout verification, Exim cached a "reject" for the entire domain. This
1249 is correct for most verifications, but it is not correct for a recipient
1250 verification with use_sender or use_postmaster set, because in that case
1251 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1252 case of MAIL FROM:<> rejection from other early rejections (e.g.
1253 rejection of HELO). When verifying a recipient using a non-null MAIL
1254 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1255 Whatever the result of the callout, the value of the domain cache is
1256 left unchanged (for any other kind of callout, getting as far as trying
1257 RCPT means that the domain itself is ok).
1259 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1262 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1263 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1264 session, the daemon ignores ECONNECT errors and logs others; it now
1265 ignores EPIPE as well.
1267 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1268 (quoted-printable decoding).
1270 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1271 later the small subsequent patch to fix an introduced bug.
1273 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1275 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1277 PH/08 An error is now given if message_size_limit is specified negative.
1279 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1280 to be given (somewhat) arbitrary names.
1282 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1285 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1286 miscellaneous code fixes
1288 PH/10 Added the log_reject_target ACL modifier to specify where to log
1291 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1292 hostname. This is wrong, because it relates to the incoming message (and
1293 probably the interface on which it is arriving) and not to the outgoing
1294 callout (which could be using a different interface). This has been
1295 changed to use the value of the helo_data option from the smtp transport
1296 instead - this is what is used when a message is actually being sent. If
1297 there is no remote transport (possible with a router that sets up host
1298 addresses), $smtp_active_hostname is used.
1300 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1301 tweaks were necessary in order to get it to work (see also 21 below):
1302 (a) The code assumed that strncpy() returns a negative number on buffer
1303 overflow, which isn't the case. Replaced with Exim's string_format()
1305 (b) There were several signed/unsigned issues. I just did the minimum
1306 hacking in of casts. There is scope for a larger refactoring.
1307 (c) The code used strcasecmp() which is not a standard C function.
1308 Replaced with Exim's strcmpic() function.
1309 (d) The code set only $1; it now sets $auth1 as well.
1310 (e) A simple test gave the error "authentication client didn't specify
1311 service in request". It would seem that Dovecot has changed its
1312 interface. Fortunately there's a specification; I followed it and
1313 changed what the client sends and it appears to be working now.
1315 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1318 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1319 address A is aliased to B and C, where B exists and C does not. Without
1320 -v the output is "A verified" because verification stops after a
1321 successful redirection if more than one address is generated. However,
1322 with -v the child addresses are also verified. Exim was outputting "A
1323 failed to verify" and then showing the successful verification for C,
1324 with its parentage. It now outputs "B failed to verify", showing B's
1325 parentage before showing the successful verification of C.
1327 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1328 look up a TXT record in a specific list after matching in a combined
1331 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1332 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1333 they consult the DNS. I had assumed they would set it the way they
1334 wanted; and indeed my experiments on Linux seem to show that in some
1335 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1336 To be on the safe side, however, I have now made the interface to
1337 host_find_byname() similar to host_find_bydns(), with an argument
1338 containing the DNS resolver options. The host_find_byname() function now
1339 sets these options at its start, just as host_find_bydns() does. The smtp
1340 transport options dns_qualify_single and dns_search_parents are passed to
1341 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1342 of host_find_byname() use the default settings of RES_DEFNAMES
1343 (qualify_single) but not RES_DNSRCH (search_parents).
1345 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1346 spool_read_header() do less string testing, by means of a preliminary
1347 switch on the second character of optional "-foo" lines. (This is
1348 overdue, caused by the large number of possibilities that now exist.
1349 Originally there were few.) While I was there, I also converted the
1350 str(n)cmp tests so they don't re-test the leading "-" and the first
1351 character, in the hope this might squeeze out yet more improvement.
1353 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1354 flag allowing group syntax was set by the header_syntax check but not
1355 turned off, possible causing trouble later; (2) The flag was not being
1356 set at all for the header_verify test, causing "group"-style headers to
1357 be rejected. I have now set it in this case, and also caused header_
1358 verify to ignore an empty address taken from a group. While doing this, I
1359 came across some other cases where the code for allowing group syntax
1360 while scanning a header line wasn't quite right (mostly, not resetting
1361 the flag correctly in the right place). These bugs could have caused
1362 trouble for malformed header lines. I hope it is now all correct.
1364 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1365 with the "reply" argument non-NULL. The code, however (which originally
1366 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1367 but it didn't always do it. This confused somebody who was copying the
1368 code for some other use. I have removed all the tests.
1370 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1371 feature that was used to support insecure browsers during the U.S. crypto
1372 embargo. It requires special client support, and Exim is probably the
1373 only MTA that supported it -- and would never use it because real RSA is
1374 always available. This code has been removed, because it had the bad
1375 effect of slowing Exim down by computing (never used) parameters for the
1376 RSA_EXPORT functionality.
1378 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1379 authenticator to fail if there's a tab character in the incoming data
1380 (there should never be unless someone is messing about, as it's supposed
1381 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1382 if the connection is using TLS or if the remote IP is the same as the
1383 local IP, and the "valid-client-cert option" if a client certificate has
1386 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1387 authenticators. This can be used for authorization after authentication
1388 succeeds. (In the case of plaintext, it servers for both authentication
1391 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1392 if any retry times were supplied.
1394 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1395 connection, where there is no client IP address to check. In this
1396 situation, the verify now always succeeds.
1398 PH/25 Applied John Jetmore's -Mset patch.
1400 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1402 PH/27 In a string expansion for a processed (not raw) header when multiple
1403 headers of the same name were present, leading whitespace was being
1404 removed from all of them, but trailing whitespace was being removed only
1405 from the last one. Now trailing whitespace is removed from each header
1406 before concatenation. Completely empty headers in a concatenation (as
1407 before) are ignored.
1409 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1410 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1412 PH/29 [Removed. This was a change that I later backed out, and forgot to
1413 correct the ChangeLog entry (that I had efficiently created) before
1414 committing the later change.]
1416 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1417 address errors (4xx response to RCPT) over the same connection as other
1418 messages routed to the same hosts. Such deliveries are always "forced",
1419 so retry times are not inspected. This resulted in far too many retries
1420 for the affected addresses. The effect occurred only when there were more
1421 hosts than the hosts_max_try setting in the smtp transport when it had
1422 the 4xx errors. Those hosts that it had tried were not added to the list
1423 of hosts for which the message was waiting, so if all were tried, there
1424 was no problem. Two fixes have been applied:
1426 (i) If there are any address or message errors in an SMTP delivery, none
1427 of the hosts (tried or untried) are now added to the list of hosts
1428 for which the message is waiting, so the message should not be a
1429 candidate for sending over the same connection that was used for a
1430 successful delivery of some other message. This seems entirely
1431 reasonable: after all the message is NOT "waiting for some host".
1432 This is so "obvious" that I'm not sure why it wasn't done
1433 previously. Hope I haven't missed anything, but it can't do any
1434 harm, as the worst effect is to miss an optimization.
1436 (ii) If, despite (i), such a delivery is accidentally attempted, the
1437 routing retry time is respected, so at least it doesn't keep
1438 hammering the server.
1440 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1441 in ${readsocket because some servers need this prod.
1443 PH/32 Added some extra debug output when updating a wait-xxx database.
1445 PH/33 The hint "could be header name not terminated by colon", which has been
1446 given for certain expansion errors for a long time, was not being given
1447 for the ${if def:h_colon_omitted{... case.
1449 PH/34 The spec says: "With one important exception, whenever a domain list is
1450 being scanned, $domain contains the subject domain." There was at least
1451 one case where this was not true.
1453 PH/35 The error "getsockname() failed: connection reset by peer" was being
1454 written to the panic log as well as the main log, but it isn't really
1455 panic-worthy as it just means the connection died rather early on. I have
1456 removed the panic log writing for the ECONNRESET error when getsockname()
1459 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1460 runs only) independently of the message's sender address. This meant
1461 that, if the 4xx error was in fact related to the sender, a different
1462 message to the same recipient with a different sender could confuse
1463 things. In particualar, this can happen when sending to a greylisting
1464 server, but other circumstances could also provoke similar problems.
1465 I have changed the default so that the retry time for these errors is now
1466 based a combination of the sender and recipient addresses. This change
1467 can be overridden by setting address_retry_include_sender=false in the
1470 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1471 remote server are returned as part of bounce messages. This was not
1472 happening for LMTP over a pipe (the lmtp transport), but now it is the
1473 same for both kinds of LMTP.
1475 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1476 in header lines that were in fact CNAMEs. This is no longer the case.
1478 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1479 and queue runs started by the daemon processed all messages. This has
1480 been fixed so that -R and -S can now usefully be given with -q<time>.
1482 PH/40 Import PCRE release 6.7 (fixes some bugs).
1484 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1486 PH/42 Give an error if -q is specified more than once.
1488 PH/43 Renamed the variables $interface_address and $interface_port as
1489 $received_ip_address and $received_port, to make it clear that these
1490 values apply to message reception, and not to the outgoing interface when
1491 a message is delivered. (The old names remain recognized, of course.)
1493 PH/44 There was no timeout on the connect() call when using a Unix domain
1494 socket in the ${readsocket expansion. There now is.
1496 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1497 be meaningful with "accept".
1499 SC/01 Eximstats V1.43
1500 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1502 SC/02 Eximstats V1.44
1503 Use a glob alias rather than an array ref in the generated
1504 parser. This improves both readability and performance.
1506 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1507 Collect SpamAssassin and rejection statistics.
1508 Don't display local sender or destination tables unless
1509 there is data to show.
1510 Added average volumes into the top table text output.
1512 SC/04 Eximstats V1.46
1513 Collect data on the number of addresses (recipients)
1514 as well as the number of messages.
1516 SC/05 Eximstats V1.47
1517 Added 'Message too big' to the list of mail rejection
1518 reasons (thanks to Marco Gaiarin).
1520 SC/06 Eximstats V1.48
1521 Mainlog lines which have GMT offsets and are too short to
1522 have a flag are now skipped.
1524 SC/07 Eximstats V1.49 (Alain Williams)
1525 Added the -emptyok flag.
1527 SC/08 Eximstats V1.50
1528 Fixes for obtaining the IP address from reject messages.
1530 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1531 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1532 whitesspace changes from 4.64-PH/27
1534 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1537 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1538 are found, allow negative numbers in numeric criteria)
1540 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1542 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1543 to match changes made in 4.64-PH/43
1545 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1547 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1548 "domain" is actually an IP address, and force a failure. This locks out
1549 those revolvers/nameservers that support "A-for-A" lookups, in
1550 contravention of the specifications.
1552 PH/48 When a host name was looked up from an IP address, and the subsequent
1553 forward lookup of the name timed out, the host name was left in
1554 $sender_host_name, contrary to the specification.
1556 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1557 restricted to single-key lookups, Exim was not diagnosing an error if
1558 * or *@ was used with a query-style lookup.
1560 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1562 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1563 long ago, but noone interested enough thought of it. Let's just say that
1564 the "1.1" means that there are some new functions that weren't there at
1565 some point in the past.
1567 PH/51 Error processing for expansion failure of helo_data from an smtp
1568 transport during callout processing was broken.
1570 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1571 tested/used via the -bh/-bhc/-bs options.
1573 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1574 bug, fixed in subsequent PCRE releases).
1576 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1577 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1579 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1585 SC/01 Use a glob alias rather than an array ref in eximstats generated
1586 parser. This improves both readability and performance.
1588 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1589 Don't display local sender or destination tables in eximstats unless
1590 there is data to show.
1591 Added average volumes into the eximstats top table text output.
1593 SC/03 Collect data on the number of addresses (recipients) as well
1594 as the number of messages in eximstats.
1596 TF/01 Correct an error in the documentation for the redirect router. Exim
1597 does (usually) call initgroups() when daemonizing.
1599 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1600 with consistent privilege compared to when running as a daemon.
1602 TF/03 Note in the spec that $authenticated_id is not set for local
1603 submissions from trusted users.
1605 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1606 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1608 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1609 by adding some example configuration directives to the default
1610 configuration file. A little bit of work is required to uncomment the
1611 directives and define how usernames and passwords are checked, but
1612 there is now a framework to start from.
1614 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1615 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1616 without this. I don't know how relevant this is to other LDAP libraries.
1618 PH/02 Add the verb name to the "unknown ACL verb" error.
1620 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1622 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1624 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1625 directory not expanded when it should be if an expanded home directory
1626 was set for the address (which is overridden by the transport).
1628 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1631 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1632 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1633 because it is too late at that time, and has no effect.
1635 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1636 security issue with \' (bugzilla #107). I could not use the
1637 PQescapeStringConn() function, because it needs a PGconn value as one of
1640 PH/08 When testing addresses using -bt, indicate those final addresses that
1641 are duplicates that would not cause an additional delivery. At least one
1642 person was confused, thinking that -bt output corresponded to deliveries.
1643 (Suppressing duplicates isn't a good idea as you lose the information
1644 about possibly different redirections that led to the duplicates.)
1646 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1647 systems where poll() doesn't work, in particular OS X.
1649 PH/10 Added more information to debugging output for retry time not reached.
1651 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1652 operations in malware.c.
1654 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1657 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1658 syslog_duplication set false, log lines that would normally be written
1659 both the the main log and to the reject log were not written to syslog at
1662 PH/14 In the default configuration, change the use of "message" in ACL warn
1663 statements to "add_header".
1665 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1666 not followed by a command (e.g. "seen endif").
1668 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1669 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1672 PH/17 Added extra conditions to the default value of delay_warning_condition
1676 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1677 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1678 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1681 The Auto-Submitted: and various List- headers are standardised, whereas I
1682 don't think Precedence: ever was.
1684 PH/18 Refactored debugging code in route_finduser() to show more information,
1685 in particular, the error code if getpwnam() issues one.
1687 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1688 This is apparently needed in addition to the PH/07 change above to avoid
1689 any possible encoding problems.
1691 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1692 but not after initializing Perl.
1694 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1695 output them only if debugging. By default they are written stderr,
1696 apparently, which is not desirable.
1698 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1701 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1704 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1706 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1707 authenticated or an ident call has been made. Suppress the default
1708 values for $authenticated_id and $authenticated_sender (but permit -oMai
1709 and -oMas) when testing with -bh.
1711 PH/24 Re-jigged the order of the tests in the default configuration so that the
1712 tests for valid domains and recipients precede the DNS black list and CSA
1713 tests, on the grounds that those ones are more expensive.
1715 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1716 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1717 This bug exists in every version of Exim that I still have, right back to
1720 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1721 However, an attempt to turn on case-sensitivity in a regex key by
1722 including (?-i) didn't work because the subject string was already
1723 lowercased, and the effects were non-intuitive. It turns out that a
1724 one-line patch can be used to allow (?-i) to work as expected.
1730 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1731 other effects) broke the use of negated acl sub-conditions.
1733 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1736 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1737 "Deny" causes Exim to reject the incoming connection with a 554 error.
1738 Unfortunately, if there is a major crisis, such as a disk failure,
1739 tcp-wrappers gives "deny", whereas what one would like would be some
1740 kind of temporary error. A kludge has been added to help with this.
1741 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1742 554 error is used if errno is still zero or contains ENOENT (which occurs
1743 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1746 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1748 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1749 errors. Otherwise a message that provokes a temporary error (when other
1750 messages do not) can cause a whole host to time out.
1752 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1753 addresses were routed directly to files or pipes from a redirect router.
1754 File deliveries just didn't batch; pipe deliveries might have suffered
1757 PH/06 A failure to get a lock for a hints database would erroneously always say
1758 "Failed to get write lock", even when it was really a read lock.
1760 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1761 of 0600. This has been changed to use the value of the lockfile_mode
1762 option (which defaults to 0600).
1764 PH/08 Applied small patch from the Sieve maintainer.
1766 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1767 folder from quota calculations, a direct delivery into this folder messed
1768 up the contents of the maildirsize file. This was because the regex was
1769 used only to exclude .Trash (or whatever) when the size of the mailbox
1770 was calculated. There was no check that a delivery was happening into an
1771 excluded directory. This bug has been fixed by ignoring all quota
1772 processing for deliveries into excluded directories.
1774 PH/10 Added the maildirfolder_create_regex option to appendfile.
1780 PH/01 The code for finding all the local interface addresses on a FreeBSD
1781 system running IPv6 was broken. This may well have applied to all BSD
1782 systems, as well as to others that have similar system calls. The broken
1783 code found IPv4 interfaces correctly, but gave incorrect values for the
1784 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1785 that it would not match correctly against @[] and not recognize the IPv6
1788 PH/02 The ipliteral router was not recognizing addresses of the form user@
1789 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1791 PH/03 Added disable_ipv6.
1793 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1794 decoded form, because it is most often used to construct To: headers
1795 lines in autoreplies, and the decoded form may well be syntactically
1796 invalid. However, $reply_address has leading white space removed, and all
1797 newlines turned into spaces so that the autoreply transport does not
1800 PH/05 If group was specified without a user on a router, and no group or user
1801 was specified on a transport, the group from the router was ignored.
1803 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1804 for visible compile-time settings that can be used to change these
1805 numbers, for those that want even more. Backwards compatibility with old
1806 spool files has been maintained. However, going back to a previous Exim
1807 release will lost any variables that are in spool files.
1809 PH/07 Two small changes when running in the test harness: increase delay when
1810 passing a TCP/IP connection to a new process, in case the original
1811 process has to generate a bounce, and remove special handling of
1812 127.0.0.2 (sic), which is no longer necessary.
1814 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1815 be the same on different OS.
1817 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1820 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1821 whether --show-vars was specified or not
1823 JJ/02 exipick: Added support for new ACL variable spool format introduced
1826 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1827 syntactically invalid From: or Reply-to: line, and a filter used this to
1828 generate an autoreply, and therefore failed to obtain an address for the
1829 autoreply, Exim could try to deliver to a non-existent relative file
1830 name, causing unrelated and misleading errors. What now happens is that
1831 it logs this as a hard delivery error, but does not attempt to create a
1834 PH/11 The exinext utility has a -C option for testing purposes, but although
1835 the given file was scanned by exinext itself; it wasn't being passed on
1836 when Exim was called.
1838 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1839 an end-of-file indication when reading a command response.
1841 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1842 compiled. In many other places in Exim, IPv6 addresses are always
1843 recognized, so I have changed this. It also means that IPv4 domain
1844 literals of the form [IPV4:n.n.n.n] are now always recognized.
1846 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1847 used if the router is not running as root, for example, when verifying at
1848 ACL time, or when using -bh. The debugging output from this situation was
1849 non-existent - all you got was a failure to exec. I have made two
1852 (a) Failures to set uid/gid, the current directory, or a process leader
1853 in a subprocess such as that created by queryprogram now generate
1854 suitable debugging ouput when -d is set.
1856 (b) The queryprogram router detects when it is not running as root,
1857 outputs suitable debugging information if -d is set, and then runs
1858 the subprocess without attempting to change uid/gid.
1860 PH/15 Minor change to Makefile for building test_host (undocumented testing
1863 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1864 additional section of a DNS packet that returns MX or SRV records.
1865 Instead, it always explicitly searches for A/AAAA records. This avoids
1866 major problems that occur when a DNS server includes only records of one
1867 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1868 fixed another bug: if SRV records were looked up and the corresponding
1869 address records were *not* found in the additional section, the port
1870 values from the SRV records were lost.
1872 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1873 using the correct key (the original address) when searching the retry
1874 rules in order to find which one to use for generating the retry hint.
1876 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1877 adding the default one. Similarly, if it contains a Reply-To: header, the
1878 errors_reply_to option, if set, is not used.
1880 PH/19 When calculating a retry time, Exim used to measure the "time since
1881 failure" by looking at the "first failed" field in the retry record. Now
1882 it does not use this if it is later than than the arrival time of the
1883 message. Instead it uses the arrival time. This makes for better
1884 behaviour in cases where some deliveries succeed, thus re-setting the
1885 "first failed" field. An example is a quota failure for a huge message
1886 when small messages continue to be delivered. Without this change, the
1887 "time since failure" will always be short, possible causing more frequent
1888 delivery attempts for the huge message than are intended.
1889 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1891 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1892 $1, $2, $3) because the numerical variables can be reset during some
1893 expansion items (e.g. "match"), thereby losing the authentication data.
1895 PH/21 Make -bV show the size of off_t variables so that the test suite can
1896 decide whether to run tests for quotas > 2G.
1898 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1899 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1900 filecount value is greater than 2G or if a quota value is greater than 2G
1901 on a system where the size of off_t is not greater than 4, a panic error
1904 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1905 never match. The debug and -bh output now contains an explicit error
1906 message indicating a malformed IPv4 address or mask.
1908 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1909 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1910 PH/23 above applies.
1912 PH/25 Do not write to syslog when running in the test harness. The only
1913 occasion when this arises is a failure to open the main or panic logs
1914 (for which there is an explicit test).
1916 PH/26 Added the /no_tell option to "control=freeze".
1918 PH/27 If a host name lookup failed very early in a connection, for example, if
1919 the IP address matched host_lookup and the reverse lookup yielded a name
1920 that did not have a forward lookup, an error message of the form "no IP
1921 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1922 could be logged. Now it outputs the IP address instead of "NULL".
1924 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1925 allows the sender and the authenticated sender to be set when
1926 submitting a message from within Exim. Since child_open_exim() is
1927 documented for local_scan(), the new function should be too.
1929 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1930 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1931 results in an empty string is now treated as unset.
1933 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1935 PH/31 Added sender_verify_fail logging option.
1937 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1938 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1939 tidied the source and removed it altogether.
1941 PH/33 When a queue run was abandoned because the load average was too high, a
1942 log line was always written; now it is written only if the queue_run log
1943 selector is set. In addition, the log line for abandonment now contains
1944 information about the queue run such as the pid. This is always present
1945 in "start" and "stop" lines but was omitted from the "abandon" line.
1947 PH/34 Omit spaces between a header name and the colon in the error message that
1948 is given when verify = headers_syntax fails (if there are lots of them,
1949 the message gets confusing).
1951 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1952 names, as there are now some PTR records that contain slashes. This check
1953 is only to protect against broken name servers that fall over on strange
1954 characters, so the fact that it applies to all lookups doesn't matter.
1956 PH/36 Now that the new test suite is complete, we can remove some of the
1957 special code in Exim that was needed for the old test suite. For example,
1958 sorting DNS records because real resolvers return them in an arbitrary
1959 order. The new test suite's fake resolver always returns records in the
1962 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1963 bounces) except when queue_only is set, to avoid logging races between
1964 the different processes.
1966 PH/38 Panic-die if .include specifies a non-absolute path.
1968 PH/39 A tweak to the "H" retry rule from its user.
1970 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1971 a label. They prevented compilation on older perls.
1973 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1974 a warning to be raised on newish perls.
1976 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1977 on queue. Changes to match documented behaviour of showing count of
1978 messages matching specified criteria.
1980 PH/40 Changed the default ident timeout from 30s to 5s.
1982 PH/41 Added support for the use of login_cap features, on those BSD systems
1983 that have them, for controlling the resources used by pipe deliveries.
1985 PH/42 The content-scanning code uses fopen() to create files in which to put
1986 message data. Previously it was not paying any attention to the mode of
1987 the files. Exim runs with umask(0) because the rest of the code creates
1988 files with open(), and sets the required mode explicitly. Thus, these
1989 files were ending up world-writeable. This was not a big issue, because,
1990 being within the spool directory, they were not world-accessible. I have
1991 created a function called modefopen, which takes an additional mode
1992 argument. It sets umask(777), creates the file, chmods it to the required
1993 mode, then resets the umask. All the relevant calls to fopen() in the
1994 content scanning code have been changed to use this function.
1996 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1997 to 24 hours. This avoids potential overflow problems when processing G
1998 and H retry rules. I suspect nobody ever tinkers with this value.
2000 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2002 PH/45 When the plaintext authenticator is running as a client, the server's
2003 challenges are checked to ensure they are valid base64 strings. By
2004 default, the authentication attempt is cancelled if an invalid string is
2005 received. Setting client_ignore_invalid_base64 true ignores these errors.
2006 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2007 they are received. Thus, the responses can be made to depend on the
2008 challenges. If an invalid string is ignored, an empty string is placed in
2011 PH/46 Messages that are created by the autoreply transport now contains a
2012 References: header, in accordance with RFCs 2822 and 3834.
2014 PH/47 Added authenticated_sender_force to the smtp transport.
2016 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2018 PH/49 Installed latest patch from the Sieve maintainer.
2020 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2021 was also set, the appendfile transport was unnecessarily scanning a
2022 directory of message files (e.g. for maildir delivery) to find the count
2023 of files (along with the size), even though it did not need this
2024 information. It now does the scan only if it needs to find either the
2025 size of the count of files.
2027 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2029 PH/52 Two bugs concerned with error handling when the smtp transport is
2032 (i) Exim was not creating retry information for temporary errors given
2033 for individual recipients after the DATA command when the smtp transport
2034 was used in LMTP mode. This meant that they could be retried too
2035 frequently, and not timed out correctly.
2037 (ii) Exim was setting the flag that allows error details to be returned
2038 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2039 recipients that were returned after the DATA command.
2041 PH/53 This is related to PH/52, but is more general: for any failing address,
2042 when detailed error information was permitted to be returned to the
2043 sender, but the error was temporary, then after the final timeout, only
2044 "retry timeout exceeded" was returned. Now it returns the full error as
2045 well as "retry timeout exceeded".
2047 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2048 do this, and (what is worse) MTAs that accept it.
2050 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2051 will now be deprecated.
2053 PH/56 New os.c-cygwin from the Cygwin maintainer.
2055 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2056 formats (previously only available in exim formats via -bpr, -bpru,
2057 and -bpra. Now also available in native and exiqgrep formats)
2059 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2060 with very large, slow to parse queues
2062 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2064 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2066 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2067 responses to authentication challenges, though it was showing the
2068 challenges; (ii) I've removed the CR characters from the debug output for
2071 PH/58 Allow for the insertion of a newline as well as a space when a string
2072 is turned into more than one encoded-word during RFC 2047 encoding. The
2073 Sieve code now uses this.
2075 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2076 data_4xx, lost_connection, tls_required.
2078 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2079 message was being sent as an SMTP response.
2081 PH/61 Add -l and -k options to exicyclog.
2083 PH/62 When verifying, if an address was redirected to one new address, so that
2084 verification continued, and the new address failed or deferred after
2085 having set something in $address_data, the value of $address_data was not
2086 passed back to the ACL. This was different to the case when no
2087 redirection occurred. The value is now passed back in both cases.
2089 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2090 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2091 use login_cap.h, so on its own it isn't the distinguishing feature. The
2092 new name refers directly to the setclassresources() function.
2094 PH/65 Added configuration files for NetBSD3.
2096 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2098 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2099 is preferred over IPv4.
2101 PH/68 The bounce_return_message and bounce_return_body options were not being
2102 honoured for bounces generated during the reception of non-SMTP messages.
2103 In particular, this applied to messages rejected by the ACL. This bug has
2104 been fixed. However, if bounce_return_message is true and bounce_return_
2105 body is false, the headers that are returned for a non-SMTP message
2106 include only those that have been read before the error was detected.
2107 (In the case of an ACL rejection, they have all been read.)
2109 PH/69 The HTML version of the specification is now built in a directory called
2110 spec_html instead of spec.html, because the latter looks like a path with
2111 a MIME-type, and this confuses some software.
2113 PH/70 Catch two compiler warnings in sieve.c.
2115 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2116 function verify_get_ident() calls ip_connect() to connect a socket, but
2117 if the "connect()" function timed out, ip_connect() used to close the
2118 socket. However, verify_get_ident() also closes the socket later, and in
2119 between Exim writes to the log, which may get opened at this point. When
2120 the socket was closed in ip_connect(), the log could get the same file
2121 descriptor number as the socket. This naturally causes chaos. The fix is
2122 not to close the socket in ip_connect(); the socket should be closed by
2123 the function that creates it. There was only one place in the code where
2124 this was missing, in the iplookup router, which I don't think anybody now
2125 uses, but I've fixed it anyway.
2127 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2128 well as to direct DNS lookups. Otherwise the handling of names in host
2129 lists is inconsistent and therefore confusing.
2135 PH/01 Two changes to the default runtime configuration:
2137 (1) Move the checks for relay_from_hosts and authenticated clients from
2138 after to before the (commented out) DNS black list checks.
2140 (2) Add control=submission to the relay_from_hosts and authenticated
2141 clients checks, on the grounds that messages accepted by these
2142 statements are most likely to be submissions.
2144 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2146 (1) Generate an error if the third argument for the ${prvs expansion is
2149 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2152 (3) Reset the variables that are obtained from the first argument of
2153 ${prvscheck and used in the second argument before leaving the code,
2154 because their memory is reclaimed, so using them afterwards may do
2157 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2158 one (it's much easier than Tom thought :-).
2160 (5) Because of (4), we can now allow for the use of $prvscheck_result
2161 inside the third argument.
2163 PH/03 For some reason, the default setting of PATH when running a command from
2164 a pipe transport was just "/usr/bin". I have changed it to
2167 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2168 anything to be listed in the output from -bV.
2170 PH/05 When a filter generated an autoreply, the entire To: header line was
2171 quoted in the delivery log line, like this:
2173 => >A.N.Other <ano@some.domain> <original@ddress> ...
2175 This has been changed so that it extracts the operative address. There
2176 may be more than one such address. If so, they are comma-separated, like
2179 => >ano@some.domain,ona@other.domain <original@ddress> ...
2181 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2182 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2183 not being looked up in the rDNS to get a host name, Exim was showing the
2184 IP address twice in Received: lines, even though the IP addresses were
2185 identical. For example:
2187 Received: from [1.2.3.4] (helo=[1.2.3.4])
2189 However, if the real host name was known, it was omitting the HELO data
2190 if it matched the actual IP address. This has been tidied up so that it
2191 doesn't show the same IP address twice.
2193 PH/07 When both +timestamp and +memory debugging was on, the value given by
2194 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2195 called by the debug printing, thereby overwriting the timestamp buffer.
2196 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2198 PH/08 When the original message was included in an autoreply transport, it
2199 always said "this is a copy of the message, including all the headers",
2200 even if body_only or headers_only was set. It now gives an appropriate
2203 PH/09 Applied a patch from the Sieve maintainer which:
2205 o fixes some comments
2206 o adds the (disabled) notify extension core
2207 o adds some debug output for the result of if/elsif tests
2208 o points to the current vacation draft in the documentation
2209 and documents the missing references header update
2213 o fixes a bug in processing the envelope test (when testing
2214 multiple envelope elements, the last element determinted the
2217 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2218 Electronic Mail") by including:
2220 Auto-submitted: auto-generated
2222 in the messages that it generates (bounce messages and others, such as
2223 warnings). In the case of bounce messages for non-SMTP mesages, there was
2224 also a typo: it was using "Auto_submitted" (underscore instead of
2225 hyphen). Since every message generated by Exim is necessarily in response
2226 to another message, thes have all been changed to:
2228 Auto-Submitted: auto-replied
2230 in accordance with these statements in the RFC:
2232 The auto-replied keyword:
2234 - SHOULD be used on messages sent in direct response to another
2235 message by an automatic process,
2237 - MUST NOT be used on manually-generated messages,
2239 - MAY be used on Delivery Status Notifications (DSNs) and Message
2240 Disposition Notifications (MDNs),
2242 - MUST NOT be used on messages generated by automatic or periodic
2243 processes, except for messages which are automatic responses to
2246 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2247 to the default Received: header definition.
2249 PH/12 Added log selector acl_warn_skipped (default on).
2251 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2252 variables because (a) they are in the wrong storage pool and (b) even if
2253 they were copied, it wouldn't work properly because of the caching.
2255 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2256 checking when decoding. Apparently there are clients that generate
2257 overlong encoded strings. Why am I not surprised?
2259 PH/15 If the first argument of "${if match_address" was not empty, but did not
2260 contain an "@" character, Exim crashed. Now it writes a panic log message
2261 and treats the condition as false.
2263 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2265 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2266 extension "envelope-auth". The code is finished and in agreement with
2267 other implementations, but there is no documentation so far and in fact,
2268 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2269 not changing the active code.
2271 Print executed "if" and "elsif" statements when debugging is used. This
2272 helps a great deal to understand what a filter does.
2274 Document more things not specified clearly in RFC3028. I had all this
2275 sorted out, when out of a sudden new issues came to my mind. Oops."
2277 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2280 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2281 canonical form (as documented). However, after a host name lookup from
2282 the IP address, check_host() was doing a simple string comparison with
2283 addresses acquired from the DNS when checking that the found name did
2284 have the original IP as one of its addresses. Since any found IPv6
2285 addresses are likely to be in abbreviated form, the comparison could
2286 fail. Luckily, there already exists a function for doing the comparison
2287 by converting both addresses to binary, so now that is used instead of
2288 the text comparison.
2290 PH/20 There was another similar case to PH/19, when a complete host name was
2291 given in a host list; looking up its IP address could give an abbreviated
2292 form, whereas the current host's name might or might not be abbreviated.
2293 The same fix has been applied.
2299 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2300 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2303 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2304 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2306 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2308 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2309 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2310 Also, for "Auto-Submitted", treat anything other than "no" as
2311 non-personal, in accordance with RFC 3834. (Previously it treated
2312 anything starting "auto-" as non-personal.)
2314 TF/01 The control=submission/name=... option had a problem with syntax
2315 errors if the name included a slash character. The /name= option
2316 now slurps the rest of the string, so it can include any characters
2317 but it must come last in the list of options (after /sender_retain
2320 PH/05 Some modifications to the interface to the fake nameserver for the new
2328 TK/01 Added the "success_on_redirect" address verification option. See
2329 NewStuff for rationale and an example.
2331 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2333 PH/02 Patch to exigrep to allow it to work on syslog lines.
2335 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2336 fread() to skip over the body file's header line, because in Cygwin the
2337 header line is locked and is inaccessible.
2339 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2340 co-exist for some time) to make it clear that it is the Exim ID that is
2341 referenced, not the Message-ID: header line.
2343 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2344 string_format() function, because snprintf() does not exist on all
2347 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2348 ${stat: expansion item.
2350 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2351 protocol synchronization error", to keep the pedants happy.
2353 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2354 well as for IRIX systems, when gcc is being used. See the host.c source
2357 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2359 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2362 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2363 transport and to the smtp transport in LMTP mode.
2365 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2367 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2368 run a filter in a subprocess. This could lead to confusion in subsequent
2369 lookups in the parent process. There should also be a search_tidyup() at
2370 the end of the subprocess.
2372 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2373 only if the host matched helo_try_verify_hosts, which caused the
2374 verification to occur when the EHLO/HELO command was issued. The ACL just
2375 tested the remembered result. Now, if a previous verification attempt has
2376 not happened, "verify = helo" does it there and then.
2378 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2380 TK/03 Fix log output including CR from clamd.
2382 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2383 exist provoked a memory error which could cause a segfault.
2385 PH/15 Installed PCRE 6.2
2387 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2389 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2390 of the problem. Specifically, suggested +O2 rather than +O1 for the
2393 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2395 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2396 record, the port was not correctly logged when the outgoing_port log
2397 selector was set (it logged the transort's default port).
2399 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2400 fallback_hosts, and "hosts" in the smtp transport.
2402 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2403 host errors such as "Connection refused".
2405 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2406 authentication with radiusclient 0.4.9:
2408 - Error returned from rc_read_config was caught wrongly
2409 - Username/password not passed on to radius server due to wrong length.
2411 The presumption is that some radiusclient API changes for 4.51/PH/17
2412 were not taken care of correctly. The code is still untested by me (my
2413 Linux distribution still has 0.3.2 of radiusclient), but it was
2414 contributed by a Radius user.
2416 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2417 expanding the "port" option of the smtp transport.
2419 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2420 while reading a MIME header. Thanks to Tom Hughes for a patch.
2422 PH/24 Include config.h inside local_scan.h so that configuration settings are
2425 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2426 that in an ACL for RCPT (for example), you can examine exactly what was
2429 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2430 commands, but it was not correctly comparing the address with the actual
2431 client host address. Thus, it would show the EHLO address in Received:
2432 header lines when this was not necessary.
2434 PH/27 Added the % operator to ${eval:}.
2436 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2437 it should be ignoring failures (because with -C, for example, it has lost
2438 privilege). It wasn't ignoring creation failures other than "already
2441 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2444 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2445 because an input line was too long, either on its own, or by virtue of
2446 too many continuations, the temporary file was not being removed, and the
2447 return code was incorrect.
2449 PH/31 Missing "BOOL" in function definition in filtertest.c.
2451 PH/32 Applied Sieve patches from the maintainer.
2453 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2455 PH/33 Added "verify = not_blind".
2457 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2458 Local/Makefile (with some defaults set). These are used in built scripts
2459 such as exicyclog, but they have never been used in the exim_install
2460 script (though there are many overriding facilities there). I have
2461 arranged that the exim_install script now takes note of these two
2464 PH/35 Installed configuration files for Dragonfly.
2466 PH/36 When a locally submitted message by a trusted user did not contain a
2467 From: header, and the sender address was obtained from -f or from an SMTP
2468 MAIL command, and the trusted user did not use -F to supply a sender
2469 name, $originator_name was incorrectly used when constructing a From:
2470 header. Furthermore, $originator_name was used for submission mode
2471 messages from external hosts without From: headers in a similar way,
2472 which is clearly wrong.
2474 PH/37 Added control=suppress_local_fixups.
2476 PH/38 When log_selector = +received_sender was set, and the addition of the
2477 sender made the log line's construction buffer exactly full, or one byte
2478 less than full, an overflow happened when the terminating "\n" was
2481 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2482 when the result of a list match is failure because a DNS lookup failed.
2484 PH/40 RM_COMMAND is now used in the building process.
2486 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2487 the "build-* directories that it finds.
2489 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2490 address in a domain literal was a prefix of an interface address.
2492 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2493 when verifying a sender address, unless rewrite_headers is false.
2495 PH/44 Wrote a long comment about why errors_to addresses are verified as
2496 recipients, not senders.
2498 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2499 the ratelimit ACL was added.
2501 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2503 PH/46 Added extra information about PostgreSQL errors to the error string.
2505 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2506 avoiding the need to install special zones in a real server. This is
2507 backwards compatible; if it can't find the fake resolver, it drops back.
2508 Thus, both old and new test suites can be run.
2510 TF/02 Added util/ratelimit.pl
2512 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2513 clock is set back in time.
2515 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2516 Candler <B.Candler@pobox.com>.
2518 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2519 OK for addresses that are the result of redirections.
2521 PH/48 A number of further additions for the benefit of the new test suite,
2522 including a fake gethostbyname() that interfaces to the fake DNS resolver
2525 TF/06 The fix for widen_domains has also been applied to qualify_single and
2526 search_parents which are the other dnslookup options that can cause
2529 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2532 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2534 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2535 service" if it encountered an MX record with an empty target hostname.
2536 The message is now "an MX or SRV record indicated no SMTP service".
2538 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2539 if the DNS of the sending site is misconfigured. This is quite a
2540 common situation. This change restores the behaviour of treating a
2541 helo verification defer as a failure.
2543 PH/51 If self=fail was set on a router, the bounce message did not include the
2544 actual error message.
2550 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2552 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2553 command itself ended in error, the underlying message about the transport
2554 filter timeout was being overwritten with the pipe command error. Now the
2555 underlying error message should be appended to the second error message.
2557 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2559 PH/02 Reduce the amount of output that "make" produces by default. Full output
2560 can still be requested.
2562 PH/03 The warning log line about a condition test deferring for a "warn" verb
2563 was being output only once per connection, rather than after each
2564 occurrence (because it was using the same function as for successful
2565 "warn" verbs). This seems wrong, so I have changed it.
2567 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2568 it should not have, which might have caused a crash in the right
2569 circumstances, but probably never did.
2571 PH/04 Installed a modified version of Tony Finch's patch to make submission
2572 mode fix the return path as well as the Sender: header line, and to
2573 add a /name= option so that you can make the user's friendly name appear
2576 TF/03 Added the control = fakedefer ACL modifier.
2578 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2579 Mark Lowes for thorough testing.
2581 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2583 TK/03 Merged latest SRS patch from Miles Wilton.
2585 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2586 with the definition in sysexits.h (which is #included earlier).
2587 Fortunately, Exim does not actually use EX_OK. The code used to try to
2588 preserve the sysexits.h value, by assumimg that macro definitions were
2589 scanned for macro replacements. I have been disabused of this notion,
2590 so now the code just undefines EX_OK before #including unistd.h.
2592 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2593 in the smtp transport. When a block could not be written in a single
2594 write() function, the timeout was being re-applied to each part-write.
2595 This seems wrong - if the receiver was accepting one byte at a time it
2596 would take for ever. The timeout is now adjusted when this happens. It
2597 doesn't have to be particularly precise.
2599 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2600 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2602 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2603 without a domain if the check to <postmaster@domain> fails.
2605 SC/01 Eximstats: added -xls and the ability to specify output files
2606 (patch written by Frank Heydlauf).
2608 SC/02 Eximstats: use FileHandles for outputing results.
2610 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2612 SC/04 Eximstats: fixed display of large numbers with -nvr option
2614 SC/05 Eximstats: fixed merging of reports with empty tables.
2616 SC/06 Eximstats: added the -include_original_destination flag
2618 SC/07 Eximstats: removed tabs and trailing whitespace.
2620 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2622 TK/06 MBOX spool code: Add real "From " MBOX separator line
2623 so the .eml file is really in mbox format (even though
2624 most programs do not really care). Patch from Alex Miller.
2626 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2627 The latter is generated from $received_to and is only set if the
2628 message has one envelope recipient. SA can use these headers,
2629 obviously out-of-the-box. Patch from Alex Miller.
2631 PH/08 The ${def test on a variable was returning false if the variable's
2632 value was "0", contrary to what the specification has always said!
2633 The result should be true unless the variable is empty.
2635 PH/09 The syntax error of a character other than { following "${if
2636 def:variable_name" (after optional whitespace) was not being diagnosed.
2637 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2638 accidental colon was present, for example, could give incorrect results.
2640 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2641 result is used (not including appendfile, where other changes are about
2644 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2645 This involved changing a lot of size variables from int to off_t. It
2646 should work with maildirs and everything.
2648 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2649 spamd dying while we are connected to it.
2651 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2654 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2655 function for BDB 4.3.
2657 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2659 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2660 the sources for PCRE have been re-arranged and more files are now
2663 PH/15 The code I had for printing potentially long long variables in PH/11
2664 above was not the best (it lost precision). The length of off_t variables
2665 is now inspected at build time, and an appropriate printing format (%ld
2666 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2667 to be "long long int" or "long int". This is needed for the internal
2668 formatting function string_vformat().
2670 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2671 the configuration file to be ":syslog", then the script "guesses" where
2672 the logs files are, rather than using the compiled in default. In our
2673 case the guess is not the same as the compiled default, so the script
2674 suddenly stopped working when I started to use syslog. The patch checks
2675 to see if log_file_path is "". If so, it attempts to read it from exim
2676 with no configuration file to get the compiled in version, before it
2677 falls back to the previous guessing code."
2679 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2680 implementing BATV in an Exim configuration. See NewStuff for the gory
2683 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2684 Makefile that are specific to HP-UX.
2686 PH/18 If the "use_postmaster" option was set for a recipient callout together
2687 with the "random" option, the postmaster address was used as the MAIL
2688 FROM address for the random test, but not for the subsequent recipient
2689 test. It is now used for both.
2691 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2692 patch removes a few documentation additions to RFC 3028, because the
2693 latest draft now contains them. It adds the new en;ascii-case comparator
2694 and a new error check for 8bit text in MIME parts. Comparator and
2695 require names are now matched exactly. I enabled the subaddress
2696 extension, but it is not well tested yet (read: it works for me)."
2698 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2699 rework some of the code of TK/09 above to avoid the hardwired use of
2700 "%lld" and "long long". Replaced the call to snprintf() with a call to
2703 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2704 records point to non-existent hosts", "retry timeout exceeded", and
2705 "retry time not reached for any host after a long failure period".
2707 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2708 experimental DomainKeys support:
2710 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2711 (2) On an error such as an illegally used "control", the wrong name for
2712 the control was given.
2714 These problems did NOT occur unless DomainKeys support was compiled.
2716 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2718 PH/24 Added ${if match_ip condition.
2720 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2721 databases so that it will be absolutely obvious if a crash occurs in the
2722 DB library. This is a regular occurrence (often caused by mis-matched
2725 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2726 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2727 picked up on a user's system that detects such things. There doesn't seem
2728 to be a gcc warning option for this - only an attribute that has to be
2729 put on the function's prototype. It seems that in Fedora Core 4 they have
2730 set this on a number of new functions. No doubt there will be more in due
2733 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2734 specify a transport. However, if an address that was verified by such a
2735 router was the subject of a callout, Exim crashed because it tried to
2736 read the rcpt_include_affixes from the non-existent transport. Now it
2737 just assumes that the setting of that option is false. This bug was
2738 introduced by 4.51/PH/31.
2740 PH/28 Changed -d+all to exclude +memory, because that information is very
2741 rarely of interest, but it makes the output a lot bigger. People tend to
2742 do -d+all out of habit.
2744 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2745 code in os-type was giving problems when libc.so lives in lib64, like on
2748 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2749 aren't the modern standard, and it seems that some systems' include files
2750 don't always have them. Exim was already checking for some of the newer
2751 ones like T_AAAA, and defining it itself. I've added checks for all the
2752 record types that Exim uses.
2754 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2755 not automatically generating a new one, as it is supposed to. This
2756 prevented TLS from working. If the file did exist, but contained invalid
2757 data, a new version was generated, as expected. It was only the case of a
2758 non-existent file that was broken.
2760 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2761 with a change in libdomainkeys > 0.64.
2763 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2764 from DNS. If the selector record carries the flag, it now has
2765 precedence over the domain-wide flag.
2767 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2769 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2770 the use of a port name that isn't defined in /etc/services) occurred, the
2771 message was deferred as in a normal delivery, and thus remained on the
2772 spool, instead of being failed because of the mua_wrapper setting. This
2773 is now fixed, and I tidied up some of the mua_wrapper messages at the
2776 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2777 the messages in a hash of arrays rather than using individual hashes.
2778 This is a bit cleaner and results in dramatic memory savings, albeit
2779 at a slight CPU cost.
2781 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2782 as requested by Marc Sherman.
2784 SC/10 Eximstats: added histograms for user specified patterns as requested
2787 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2789 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2790 fopen() in the content-scanning modules that did not already have it.
2796 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2797 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2799 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2801 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2803 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2804 to test Sieve filters that use "vacation".
2806 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2807 that changes the way the GnuTLS parameters are stored in the cache file.
2808 The new format can be generated externally. For backward compatibility,
2809 if the data in the cache doesn't make sense, Exim assumes it has read an
2810 old-format file, and it generates new data and writes a new file. This
2811 means that you can't go back to an older release without removing the
2814 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2815 work if there are any delivery delays because "one_time" forces the
2816 parent to be marked "delivered", so its unseen clone is never tried
2817 again. For this reason, Exim now forbids the simultaneous setting of
2820 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2821 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2822 entry if you want to know the details. The fix, however, neglected to
2823 consider the case when local delivery batching is involved. The test for
2824 "previously delivered" was not happening when checking to see if an
2825 address could be batched with a previous (undelivered) one; under
2826 certain circumstances this could lead to multiple deliveries to the same
2829 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2830 in its include files, and this causes problems building Exim.
2832 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2833 header_syntax) but Exim was just ignoring anything given after a slash.
2834 In particular, this caused confusion with an attempt to use "verify =
2835 reverse_host_lookup/defer_ok". An error is now given when options are
2836 supplied for verify items that do not have them. (Maybe reverse_host_
2837 lookup should have a defer_ok option, but that's a different point.)
2839 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2840 defined by RFC 821) to 2048, because there were problems with some AUTH
2841 commands, and RFC 1869 says the size should be increased for extended
2842 SMTP commands that take arguments.
2844 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2847 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2848 "unknown" error; now it says that the functionality isn't in the binary.
2850 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2851 an address' error message when a string expansion fails (syntax or
2852 whatever). Otherwise the password may appear in the log. Following change
2853 PH/42 below, there is no longer a chance of it appearing in a bounce
2856 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2858 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2859 of its addresses was ever tried. (Bugzilla bug #2.)
2861 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2862 the result incorrectly in the debug output. (It correctly added a newline
2863 to what was transported.)
2865 TF/01 Added $received_time.
2867 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2868 commented out examples of how to interface to a virus scanner and to
2869 SpamAssassin. Also added commented examples of av_scanner and
2870 spamd_address settings.
2872 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2873 and controls are allowed in which ACLs. There were a couple of minor
2874 errors. Some of the entries in the conditions table (which is a table of
2875 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2876 negation of where the condition IS allowed.
2878 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2880 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2881 header file does not have a version number, so I've had to invent a new
2882 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2883 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2884 radiusclient), but it was contributed by a Radius user.
2886 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2887 files or directories, for OpenSSL.
2889 PH/19 When an Exim process that is running as root has to create an Exim log
2890 file, it does so in a subprocess that runs as exim:exim so as to get the
2891 ownership right at creation (otherwise, other Exim processes might see
2892 the file with the wrong ownership). There was no test for failure of this
2893 fork() call, which would lead to the process getting stuck as it waited
2894 for a non-existent subprocess. Forks do occasionally fail when resources
2895 run out. I reviewed all the other calls to fork(); they all seem to check
2898 PH/20 When checking for unexpected SMTP input at connect time (before writing
2899 the banner), Exim was not dealing correctly with a non-positive return
2900 from the read() function. If the client had disconnected by this time,
2901 the result was a log entry for a synchronization error with an empty
2902 string after "input=" when read() returned zero. If read() returned -1
2903 (an event I could not check), uninitialized data bytes were printed.
2904 There were reports of junk text (parts of files, etc) appearing after
2907 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2909 PH/22 Added support for macro redefinition, and (re)definition in between
2910 driver and ACL definitions.
2912 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2913 forgetting to use the resulting value; it was using the unexpanded value.
2915 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2916 hadn't been configured. The fix is from Juergen Kreileder, who
2917 understands it better than I do:
2919 "Here's what I see happening with three configured cyrus_sasl
2920 authenticators configured (plain, login, cram-md5):
2922 On startup auth_cyrus_sasl_init() gets called for each of these.
2923 This means three calls to sasl_listmech() without a specified mech_list.
2924 => SASL tests which mechs of all available mechs actually work
2925 => three warnings about OTP not working
2926 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2928 With the patch, sasl_listmech() also gets called three times. But now
2929 SASL's mech_list option is set to the server_mech specified in the the
2930 authenticator. Or in other words, the answer from sasl_listmech()
2931 gets limited to just the mech you're testing for (which is different
2933 => the return list contains just 'plain' or 'login', 'cram-md5' or
2934 nothing depending on the value of ob->server_mech.
2936 I've just tested the patch: Authentication still works fine,
2937 unavailable mechs specified in the exim configuration are still
2938 caught, and the auth.log warnings about OTP are gone."
2940 PH/25 When debugging is enabled, the contents of the command line are added
2941 to the debugging output, even when log_selector=+arguments is not
2944 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2945 answer is "GNU", and only if the return is "GNU/something" is the answer
2948 PH/27 $acl_verify_message is now set immediately after the failure of a
2949 verification in an ACL, and so is available in subsequent modifiers. In
2950 particular, the message can be preserved by coding like this:
2952 warn !verify = sender
2953 set acl_m0 = $acl_verify_message
2955 Previously, $acl_verify_message was set only while expanding "message"
2956 and "log_message" when a very denied access.
2958 PH/28 Modified OS/os.c-Linux with
2960 -#ifndef OS_LOAD_AVERAGE
2961 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2963 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2964 nomenclature these days.)
2966 PH/29 Installed patch from the Sieve maintainer that adds the options
2967 sieve_useraddress and sieve_subaddress to the redirect router.
2969 PH/30 In these circumstances:
2970 . Two addresses routed to the same list of hosts;
2971 . First host does not offer TLS;
2972 . First host accepts first address;
2973 . First host gives temporary error to second address;
2974 . Second host offers TLS and a TLS session is established;
2975 . Second host accepts second address.
2976 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2977 and peerdn, if requested) that were in fact used only for the second
2980 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2981 attention to any local part prefix or suffix that was matched by the
2982 router that accepted the address. It now behaves in the same way as it
2983 does for delivery: the affixes are removed from the local part unless
2984 rcpt_include_affixes is set on the transport.
2986 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2987 timeout during the DATA phase of an incoming message.
2989 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2990 applied a patch sanctioned by the Sieve maintainer.
2992 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2993 the uid or gid is negative. A case of a negative gid caused this to be
2994 noticed. The fix allows for either to be negative.
2996 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2997 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3000 PH/36 The change PH/12 above was broken. Fixed it.
3002 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3003 the grounds that routing the same address twice would always produce the
3004 same answer. This might have been true once, but it is certainly no
3005 longer true now. Routing a child address may depend on the previous
3006 routing that produced that child. Some complicated redirection strategies
3007 went wrong when messages had multiple recipients, and made Exim's
3008 behaviour dependent on the order in which the addresses were given.
3010 I have moved the duplicate checking until after the routing is complete.
3011 Exim scans the addresses that are assigned to local and remote
3012 transports, and removes any duplicates. This means that more work will be
3013 done, as duplicates will always all be routed, but duplicates are
3014 presumably rare, so I don't expect this is of any significance.
3016 For deliveries to pipes, files, and autoreplies, the duplicate checking
3017 still happens during the routing process, since they are not going to be
3020 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3021 It corrects a timeout issue with spamd. This is Ian's comment: "The
3022 background is that sometimes spamd either never reads data from a
3023 connection it has accepted, or it never writes response data. The exiscan
3024 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3025 blindly assumes that writes won't block so it may never time out."
3027 PH/39 Allow G after quota size as well as K and M.
3029 PH/40 The value set for $authenticated_id in an authenticator may not contain
3030 binary zeroes or newlines because the value is written to log lines and
3031 to spool files. There was no check on this. Now the value is run through
3032 the string_printing() function so that such characters are converted to
3033 printable escape sequences.
3035 PH/41 $message_linecount is a new variable that contains the total number of
3036 lines in the message. Compare $body_linecount, which is the count for the
3039 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3040 bounce and delay warning messages, except in certain special cases, which
3043 (a) An SMTP error message from a remote host;
3044 (b) A message specified in a :fail: redirection;
3045 (c) A message specified in a "fail" command in a system filter;
3046 (d) A message specified in a FAIL return from the queryprogram router;
3047 (e) A message specified by the cannot_route_message router option.
3049 In these cases only, Exim does include the error details in bounce and
3050 warning messages. There are also a few cases where bland messages such
3051 as "unrouteable address" or "local delivery error" are given.
3053 PH/43 $value is now also set for the "else" part of a ${run expansion.
3055 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3056 being worked on, but at least Exim now implements the latest version to
3059 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3060 process to complete was treated as a delivery failure, a timeout while
3061 writing the message to the pipe was logged, but erroneously treated as a
3062 successful delivery. Such timeouts include transport filter timeouts. For
3063 consistency with the overall process timeout, these timeouts are now
3064 treated as errors, giving rise to delivery failures by default. However,
3065 there is now a new Boolean option for the pipe transport called
3066 timeout_defer, which, if set TRUE, converts the failures into defers for
3067 both kinds of timeout. A transport filter timeout is now identified in
3070 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3071 systems where "make" and "gmake" are different, calling "gmake" at top
3072 level broke things. I've arranged for the value of $(MAKE) to be passed
3073 from the Makefile to this script so that it can call the same version of
3077 A note about Exim versions 4.44 and 4.50
3078 ----------------------------------------
3080 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3081 changes of various kinds. As a consequence, a big documentation update was
3082 needed. This delayed the release for rather longer than seemed good, especially
3083 in the light of a couple of (minor) security issues. Therefore, the changes
3084 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3085 release. So 4.44 and 4.50 are in effect two different branches that both start
3088 I have left the 4.50 change log unchanged; it contains all the changes since
3089 4.43. The change log for 4.44 is below; many of its items are identical to
3090 those for 4.50. This seems to be the most sensible way to preserve the
3091 historical information.
3097 1. Minor wording change to the doc/README.SIEVE file.
3099 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3100 computation of the current number of files was incorrect.
3102 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3103 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3106 4. Give more explanation in the error message when the command for a transport
3107 filter fails to execute.
3109 5. There are several places where Exim runs a non-Exim command in a
3110 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3111 was being done only for the command run by the queryprogram router. It is
3112 now done for all such subprocesses. The other cases are: ${run, transport
3113 filters, and the commands run by the lmtp and pipe transports.
3115 6. Added CONFIGURE_GROUP build-time option.
3117 7. Some older OS have a limit of 256 on the maximum number of file
3118 descriptors. Exim was using setrlimit() to set 1000 as a large value
3119 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3120 systems. I've change it so that if it can't get 1000, it tries for 256.
3122 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3123 was an oversight, and furthermore, ever since the addition of extra
3124 controls (e.g. 4.43/32), the checks on when to allow different forms of
3125 "control" were broken. There should now be diagnostics for all cases when a
3126 control that does not make sense is encountered.
3128 9. Added the /retain_sender option to "control=submission".
3130 10. $recipients is now available in the predata ACL (oversight).
3132 11. Tidy the search cache before the fork to do a delivery from a message
3133 received from the command line. Otherwise the child will trigger a lookup
3134 failure and thereby defer the delivery if it tries to use (for example) a
3135 cached ldap connection that the parent has called unbind on.
3137 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3138 of $address_data from the recipient verification was clobbered by the
3139 sender verification.
3141 13. The value of address_data from a sender verification is now available in
3142 $sender_address_data in subsequent conditions in the ACL statement.
3144 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3146 15. Added a new option "connect=<time>" to callout options, to set a different
3149 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3150 was its contents. (It was OK if the option was not defined at all.)
3152 17. A "Completed" log line is now written for messages that are removed from
3153 the spool by the -Mrm option.
3155 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3156 information about exactly what failed.
3158 19. Added -dd to debug only the daemon process.
3160 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3161 handles timeouts, both on the server side and network timeouts. Renamed the
3162 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3164 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3165 It is now set to "smtps".
3167 22. $host_address is now set to the target address during the checking of
3168 ignore_target_hosts.
3170 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3171 being passed; this would have caused $sender_host_name to have been used if
3172 matching the list had actually called for a host name (not very likely,
3173 since this list is usually IP addresses). A host name is now passed as
3176 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3177 code that specifies a non-restarting handler (typically sigaction() in
3178 modern systems) in an attempt to fix a rare and obscure crash bug.
3180 25. Narrowed the window for a race in the daemon that could cause it to ignore
3181 SIGCHLD signals. This is not a major problem, because they are used only to
3182 wake it up if nothing else does.
3184 26. A malformed maildirsize file could cause Exim to calculate negative values
3185 for the mailbox size or file count. Odd effects could occur as a result.
3186 The maildirsize information is now recalculated if the size or filecount
3189 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3190 support for a long time. Removed HAVE_SYS_VFS_H.
3192 28. Installed the latest version of exipick from John Jetmore.
3194 29. In an address list, if the pattern was not a regular expression, an empty
3195 subject address (from a bounce message) matched only if the pattern was an
3196 empty string. Non-empty patterns were not even tested. This was the wrong
3197 because it is perfectly reasonable to use an empty address as part of a
3198 database query. An empty address is now tested by patterns that are
3199 lookups. However, all the other forms of pattern expect the subject to
3200 contain a local part and a domain, and therefore, for them, an empty
3201 address still always fails if the pattern is not itself empty.
3203 30. Exim went into a mad DNS loop when attempting to do a callout where the
3204 host was specified on an smtp transport, and looking it up yielded more
3205 than one IP address.
3207 31. Re-factored the code for checking spool and log partition space into a
3208 function that finds that data and another that does the check. The former
3209 is then used to implement four new variables: $spool_space, $log_space,
3210 $spool_inodes, and $log_inodes.
3212 32. The RFC2047 encoding function was originally intended for short strings
3213 such as real names; it was not keeping to the 75-character limit for
3214 encoded words that the RFC imposes. It now respects the limit, and
3215 generates multiple encoded words if necessary. To be on the safe side, I
3216 have increased the buffer size for the ${rfc2047: expansion operator from
3219 33. It is now permitted to omit both strings after an "if" condition; if the
3220 condition is true, the result is "true". As before, when the second string
3221 is omitted, a false condition yields an empty string. This makes it less
3222 cumbersome to write custom ACL and router conditions.
3224 34. Failure to deliver a bounce message always caused it to be frozen, even if
3225 there was an errors_to setting on the router. The errors_to setting is now
3228 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3229 canonical form (fully expanded) before being placed in
3230 $sender_host_address.
3232 36. The table in the code that translates DNS record types into text (T_A to
3233 "A" for instance) was missing entries for NS and CNAME. It is just possible
3234 that this could have caused confusion if both these types were looked up
3235 for the same domain, because the text type is used as part of Exim's
3236 per-process caching. But the chance of anyone hitting this buglet seems
3239 37. The dnsdb lookup has been extended in a number of ways.
3241 (1) There is a new type, "zns", which walks up the domain tree until it
3242 finds some nameserver records. It should be used with care.
3244 (2) There is a new type, "mxh", which is like "mx" except that it returns
3245 just the host names, not the priorities.
3247 (3) It is now possible to give a list of domains (or IP addresses) to be
3248 looked up. The behaviour when one of the lookups defers can be
3249 controlled by a keyword.
3251 (4) It is now possible to specify the separator character for use when
3252 multiple records are returned.
3254 38. The dnslists ACL condition has been extended: it is now possible to supply
3255 a list of IP addresses and/or domains to be looked up in a particular DNS
3258 39. Added log_selector=+queue_time_overall.
3260 40. When running the queue in the test harness, wait just a tad after forking a
3261 delivery process, to get repeatability of debugging output.
3263 41. Include certificate and key file names in error message when GnuTLS fails
3264 to set them up, because the GnuTLS error message doesn't include the name
3265 of the failing file when there is a problem reading it.
3267 42. Allow both -bf and -bF in the same test run.
3269 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3271 44. The "Exiscan patch" is now merged into the mainline Exim source.
3273 45. Sometimes the final signoff response after QUIT could fail to get
3274 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3275 < 0 before doing a fflush(). This bug looks as though it goes back to the
3276 introduction of TLS in release 3.20, but "sometimes" must have been rare
3277 because the tests only now provoked it.
3279 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3280 (this can affect the format of dates).
3282 47. exim_tidydb, when checking for the continued existence of a message for
3283 which it has found a message-specific retry record, was not finding
3284 messages that were in split spool directories. Consequently, it was
3285 deleting retry records that should have stayed in existence.
3287 48. Steve fixed some bugs in eximstats.
3289 49. The SPA authentication driver was not abandoning authentication and moving
3290 on to the next authenticator when an expansion was forced to fail,
3291 contradicting the general specification for all authenticators. Instead it
3292 was generating a temporary error. It now behaves as specified.
3294 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3295 (the order specifies the preference for clients). The order is now AES256,
3296 AES128, 3DES, ARCFOUR128.
3298 51. Small patch to Sieve code - explicitly set From: when generating an
3301 52. Exim crashed if a remote delivery caused a very long error message to be
3302 recorded - for instance if somebody sent an entire SpamAssassin report back
3303 as a large number of 550 error lines. This bug was coincidentally fixed by
3304 increasing the size of one of Exim's internal buffers (big_buffer) that
3305 happened as part of the Exiscan merge. However, to be on the safe side, I
3306 have made the code more robust (and fixed the comments that describe what
3309 53. Now that there can be additional text after "Completed" in log lines (if
3310 the queue_time_overall log selector is set), a one-byte patch to exigrep
3311 was needed to allow it to recognize "Completed" as not the last thing in
3314 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3315 patch that reportedly fixes this has been added. I am not expert enough to
3316 create a test for it. This is what the patch creator wrote:
3318 "I found a little strange behaviour of ldap code when working with
3319 Windows 2003 AD Domain, where users was placed in more than one
3320 Organization Units. When I tried to give exim partial DN, the exit code
3321 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3322 But simultaneously result of request was absolutely normal ldap result,
3323 so I produce this patch..."
3325 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3326 REFERENCE, so I have modified the code to exclude the patch when that macro
3329 55. Some experimental protocols are using DNS PTR records for new purposes. The
3330 keys for these records are domain names, not reversed IP addresses. The
3331 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3332 leaves it alone. Component reversal etc. now happens only for IP addresses.
3335 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3337 57. Double the size of the debug message buffer (to 2048) so that more of very
3338 long debug lines gets shown.
3340 58. The exicyclog utility now does better if the number of log files to keep
3341 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3343 59. Two changes related to the smtp_active_hostname option:
3345 (1) $smtp_active_hostname is now available as a variable.
3346 (2) The default for smtp_banner uses $smtp_active_hostname instead
3347 of $primary_hostname.
3349 60. The host_aton() function is supposed to be passed a string that is known
3350 to be a valid IP address. However, in the case of IPv6 addresses, it was
3351 not checking this. This is a hostage to fortune. Exim now panics and dies
3352 if the condition is not met. A case was found where this could be provoked
3353 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3354 components; fortuitously, this particular loophole had already been fixed
3355 by change 4.50/55 above.
3357 If there are any other similar loopholes, the new check in host_aton()
3358 itself should stop them being exploited. The report I received stated that
3359 data on the command line could provoke the exploit when Exim was running as
3360 exim, but did not say which command line option was involved. All I could
3361 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3362 running as the user.
3365 61. There was a buffer overflow vulnerability in the SPA authentication code
3366 (which came originally from the Samba project). I have added a test to the
3367 spa_base64_to_bits() function which I hope fixes it.
3370 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3371 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3373 63. The daemon start-up calls getloadavg() while still root for those OS that
3374 need the first call to be done as root, but it missed one case: when
3375 deliver_queue_load_max is set with deliver_drop_privilege. This is
3376 necessary for the benefit of the queue runner, because there is no re-exec
3377 when deliver_drop_privilege is set.
3379 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3380 This has been fixed.
3382 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3383 was in use, was not putting the data itself into the right store pool;
3384 consequently, it could be overwritten for a subsequent message in the same
3385 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3388 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3390 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3391 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3392 the function were treating the return as a boolean value, which happened to
3393 work because 0=false and not-0=true, but is not correct code.
3395 68. The host_aton() function was not handling scoped IPv6 addresses (those
3396 with, for example, "%eth0" on the end) correctly.
3398 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3399 negated items (that is, ~something) in unsigned ints. Some compilers
3400 apparently mutter when there is no cast.
3402 70. If an address verification called from an ACL failed, and did not produce a
3403 user-specific message (i.e. there was only a "system" message), nothing was
3404 put in $acl_verify_message. In this situation, it now puts the system
3407 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3408 session; change 4.31/43 added the unwanted input to the log line - except
3409 that it did not do this in the start of session case. It now does.
3411 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3412 This is wrong and can cause the other end to generate a synchronization
3413 error if it is another Exim or anything else that does the synchronization
3414 check. A QUIT command is no longer sent after a timeout.
3416 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3417 during host lookups.
3419 74. The defer_ok option of callout verification was not working if it was used
3420 when verifying addresses in header lines, that is, for this case:
3422 verify = header_sender/callout=defer_ok
3424 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3425 those file descriptors could be used for SMTP connections. If anything
3426 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3427 could be sent to the SMTP client, causing chaos. The daemon now opens
3428 stdin, stdout, and stderr to /dev/null when it puts itself into the
3431 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3432 log by default. The user can override this with suitable Perl magic.
3434 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3435 the log message when discard triggers, was not working for the DATA ACL or
3436 for the non-SMTP ACL.
3438 78. Error message wording change in sieve.c.
3440 79. If smtp_accept_max_per_host was set, the number of connections could be
3441 restricted to fewer than expected, because the daemon was trying to set up
3442 a new connection before checking whether the processes handling previous
3443 connections had finished. The check for completed processes is now done
3444 earlier. On busy systems, this bug wouldn't be noticed because something
3445 else would have woken the daemon, and it would have reaped the completed
3448 80. If a message was submitted locally by a user whose login name contained one
3449 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3450 It caused a spool format error. I have fixed the spool reading code. A
3451 related problem was that the "from" clause in the Received: line became
3452 illegal because of the space(s). It is now covered by ${quote_local_part.
3454 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3457 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3459 83. Patch from Sieve maintainer to fix unterminated string problem in
3460 "vacation" handling.
3462 84. Some minor changes to the Linux configuration files to help with other
3463 OS variants using glibc.
3465 85. One more patch for Sieve to update vacation handling to latest spec.
3468 ----------------------------------------------------
3469 See the note above about the 4.44 and 4.50 releases.
3470 ----------------------------------------------------
3476 1. Change 4.43/35 introduced a bug that caused file counts to be
3477 incorrectly computed when quota_filecount was set in an appendfile
3480 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3481 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3484 3. Give more explanation in the error message when the command for a transport
3485 filter fails to execute.
3487 4. There are several places where Exim runs a non-Exim command in a
3488 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3489 was being done only for the command run by the queryprogram router. It is
3490 now done for all such subprocesses. The other cases are: ${run, transport
3491 filters, and the commands run by the lmtp and pipe transports.
3493 5. Some older OS have a limit of 256 on the maximum number of file
3494 descriptors. Exim was using setrlimit() to set 1000 as a large value
3495 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3496 systems. I've change it so that if it can't get 1000, it tries for 256.
3498 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3499 was an oversight, and furthermore, ever since the addition of extra
3500 controls (e.g. 4.43/32), the checks on when to allow different forms of
3501 "control" were broken. There should now be diagnostics for all cases when a
3502 control that does not make sense is encountered.
3504 7. $recipients is now available in the predata ACL (oversight).
3506 8. Tidy the search cache before the fork to do a delivery from a message
3507 received from the command line. Otherwise the child will trigger a lookup
3508 failure and thereby defer the delivery if it tries to use (for example) a
3509 cached ldap connection that the parent has called unbind on.
3511 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3512 of $address_data from the recipient verification was clobbered by the
3513 sender verification.
3515 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3516 was its contents. (It was OK if the option was not defined at all.)
3518 11. A "Completed" log line is now written for messages that are removed from
3519 the spool by the -Mrm option.
3521 12. $host_address is now set to the target address during the checking of
3522 ignore_target_hosts.
3524 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3525 being passed; this would have caused $sender_host_name to have been used if
3526 matching the list had actually called for a host name (not very likely,
3527 since this list is usually IP addresses). A host name is now passed as
3530 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3531 code that specifies a non-restarting handler (typically sigaction() in
3532 modern systems) in an attempt to fix a rare and obscure crash bug.
3534 15. Narrowed the window for a race in the daemon that could cause it to ignore
3535 SIGCHLD signals. This is not a major problem, because they are used only to
3536 wake it up if nothing else does.
3538 16. A malformed maildirsize file could cause Exim to calculate negative values
3539 for the mailbox size or file count. Odd effects could occur as a result.
3540 The maildirsize information is now recalculated if the size or filecount
3543 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3544 support for a long time. Removed HAVE_SYS_VFS_H.
3546 18. Updated exipick to current release from John Jetmore.
3548 19. Allow an empty sender to be matched against a lookup in an address list.
3549 Previously the only cases considered were a regular expression, or an
3552 20. Exim went into a mad DNS lookup loop when doing a callout where the
3553 host was specified on the transport, if the DNS lookup yielded more than
3556 21. The RFC2047 encoding function was originally intended for short strings
3557 such as real names; it was not keeping to the 75-character limit for
3558 encoded words that the RFC imposes. It now respects the limit, and
3559 generates multiple encoded words if necessary. To be on the safe side, I
3560 have increased the buffer size for the ${rfc2047: expansion operator from
3563 22. Failure to deliver a bounce message always caused it to be frozen, even if
3564 there was an errors_to setting on the router. The errors_to setting is now
3567 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3568 canonical form (fully expanded) before being placed in
3569 $sender_host_address.
3571 24. Updated eximstats to version 1.33
3573 25. Include certificate and key file names in error message when GnuTLS fails
3574 to set them up, because the GnuTLS error message doesn't include the name
3575 of the failing file when there is a problem reading it.
3577 26. Expand error message when OpenSSL has problems setting up cert/key files.
3580 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3581 (this can affect the format of dates).
3583 28. exim_tidydb, when checking for the continued existence of a message for
3584 which it has found a message-specific retry record, was not finding
3585 messages that were in split spool directories. Consequently, it was
3586 deleting retry records that should have stayed in existence.
3588 29. eximstats updated to version 1.35
3589 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3590 1.35 - bugfix such that pie charts by volume are generated correctly
3592 30. The SPA authentication driver was not abandoning authentication and moving
3593 on to the next authenticator when an expansion was forced to fail,
3594 contradicting the general specification for all authenticators. Instead it
3595 was generating a temporary error. It now behaves as specified.
3597 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3598 (the order specifies the preference for clients). The order is now AES256,
3599 AES128, 3DES, ARCFOUR128.
3601 31. Small patch to Sieve code - explicitly set From: when generating an
3604 32. Exim crashed if a remote delivery caused a very long error message to be
3605 recorded - for instance if somebody sent an entire SpamAssassin report back
3606 as a large number of 550 error lines. This bug was coincidentally fixed by
3607 increasing the size of one of Exim's internal buffers (big_buffer) that
3608 happened as part of the Exiscan merge. However, to be on the safe side, I
3609 have made the code more robust (and fixed the comments that describe what
3612 33. Some experimental protocols are using DNS PTR records for new purposes. The
3613 keys for these records are domain names, not reversed IP addresses. The
3614 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3615 leaves it alone. Component reversal etc. now happens only for IP addresses.
3618 34. The host_aton() function is supposed to be passed a string that is known
3619 to be a valid IP address. However, in the case of IPv6 addresses, it was
3620 not checking this. This is a hostage to fortune. Exim now panics and dies
3621 if the condition is not met. A case was found where this could be provoked
3622 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3623 components; fortuitously, this particular loophole had already been fixed
3624 by change 4.50/55 or 4.44/33 above.
3626 If there are any other similar loopholes, the new check in host_aton()
3627 itself should stop them being exploited. The report I received stated that
3628 data on the command line could provoke the exploit when Exim was running as
3629 exim, but did not say which command line option was involved. All I could
3630 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3631 running as the user.
3634 35. There was a buffer overflow vulnerability in the SPA authentication code
3635 (which came originally from the Samba project). I have added a test to the
3636 spa_base64_to_bits() function which I hope fixes it.
3639 36. The daemon start-up calls getloadavg() while still root for those OS that
3640 need the first call to be done as root, but it missed one case: when
3641 deliver_queue_load_max is set with deliver_drop_privilege. This is
3642 necessary for the benefit of the queue runner, because there is no re-exec
3643 when deliver_drop_privilege is set.
3645 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3646 was in use, was not putting the data itself into the right store pool;
3647 consequently, it could be overwritten for a subsequent message in the same
3648 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3651 38. Sometimes the final signoff response after QUIT could fail to get
3652 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3653 < 0 before doing a fflush(). This bug looks as though it goes back to the
3654 introduction of TLS in release 3.20, but "sometimes" must have been rare
3655 because the tests only now provoked it.
3661 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3662 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3663 Now it returns an integer. A number of calls were still expecting a T/F
3664 return. Fortuitously, in all cases, the tests worked in OK situations,
3665 which is the norm. However, things would have gone wrong on any write
3666 failures on the smtp file descriptor. This function is used when sending
3667 messages over SMTP and also when doing verify callouts.
3669 2. When Exim is called to do synchronous delivery of a locally submitted
3670 message (the -odf or -odi options), it no longer closes stderr before doing
3673 3. Implemented the mua_wrapper option.
3675 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3677 5. Implemented the functions header_remove(), header_testname(),
3678 header_add_at_position(), and receive_remove_recipient(), and exported them
3681 6. If an ACL "warn" statement specified the addition of headers, Exim already
3682 inserted X-ACL-Warn: at the start if there was no header name. However, it
3683 was not making this test for the second and subsequent header lines if
3684 there were newlines in the string. This meant that an invalid header could
3685 be inserted if Exim was badly configured.
3687 7. Allow an ACL "warn" statement to add header lines at the start or after all
3688 the Received: headers, as well as at the end.
3690 8. Added the rcpt_4xx retry error code.
3692 9. Added postmaster_mailfrom=xxx to callout verification option.
3694 10. Added mailfrom=xxxx to the callout verification option, for verify=
3697 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3698 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3700 12. Inserted some casts to stop certain compilers warning when using pointer
3701 differences as field lengths or precisions in printf-type calls (mostly
3702 affecting debugging statements).
3704 13. Added optional readline() support for -be (dynamically loaded).
3706 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3707 same clock tick as a message's arrival, so that its received time was the
3708 same as the "first fail" time on the retry record, and that message
3709 remained on the queue past the ultimate address timeout, every queue runner
3710 would try a delivery (because it was past the ultimate address timeout) but
3711 after another failure, the ultimate address timeout, which should have then
3712 bounced the address, did not kick in. This was a "< instead of <=" error;
3713 in most cases the first failure would have been in the next clock tick
3714 after the received time, and all would be well.
3716 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3717 being recognized when the domain list was tested by the match_domain
3718 condition in an expansion string.
3720 16. Added the ${str2b64: operator.
3722 17. Exim was always calling setrlimit() to set a large limit for the number of
3723 processes, without checking whether the existing limit was already
3724 adequate. (It did check for the limit on file descriptors.) Furthermore,
3725 errors from getrlimit() and setrlimit() were being ignored. Now they are
3726 logged to the main and panic logs, but Exim does carry on, to try to do its
3727 job under whatever limits there are.
3729 18. Imported PCRE 5.0.
3731 19. Trivial typo in log message " temporarily refused connection" (the leading
3734 20. If the log selector return_path_on_delivery was set and an address was
3735 redirected to /dev/null, the delivery process crashed because it assumed
3736 that a return path would always be set for a "successful" delivery. In this
3737 case, the whole delivery is bypassed as an optimization, and therefore no
3740 21. Internal re-arrangement: the function for sending a challenge and reading
3741 a response while authentication was assuming a zero-terminated challenge
3742 string. It's now changed to take a pointer and a length, to allow for
3743 binary data in such strings.
3745 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3747 23. Exim was not respecting finduser_retries when seeking the login of the
3748 uid under which it was called; it was always trying 10 times. (The default
3749 setting of finduser_retries is zero.) Also, it was sleeping after the final
3750 failure, which is pointless.
3752 24. Implemented tls_on_connect_ports.
3754 25. Implemented acl_smtp_predata.
3756 26. If the domain in control=submission is set empty, Exim assumes that the
3757 authenticated id is a complete email address when it generates From: or
3758 Sender: header lines.
3760 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3761 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3762 chown and chgrp in /bin and hostname in /usr/bin.
3764 28. Exim was keeping the "process log" file open after each use, just as it
3765 does for the main log. This opens the possibility of it remaining open for
3766 long periods when the USR1 signal hits a daemon. Occasional processlog
3767 errors were reported, that could have been caused by this. Anyway, it seems
3768 much more sensible not to leave this file open at all, so that is what now
3771 29. The long-running daemon process does not normally write to the log once it
3772 has entered its main loop, and it closes the log before doing so. This is
3773 so that log files can straightforwardly be renamed and moved. However,
3774 there are a couple of unusual error situations where the daemon does write
3775 log entries, and I had neglected to close the log afterwards.
3777 30. The text of an SMTP error response that was received during a remote
3778 delivery was being truncated at 512 bytes. This is too short for some of
3779 the long messages that one sometimes sees. I've increased the limit to
3782 31. It is now possible to make retry rules that apply only when a message has a
3783 specific sender, in particular, an empty sender.
3785 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3786 it possible to be selective about when SMTP synchronization is enforced.
3788 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3790 32. Implemented hosts_connection_nolog.
3792 33. Added an ACL for QUIT.
3794 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3797 35. Added mailbox_size and mailbox_filecount to appendfile.
3799 36. Added control = no_multiline_responses to ACLs.
3801 37. There was a bug in the logic of the code that waits for the clock to tick
3802 in the case where the clock went backwards by a substantial amount such
3803 that the microsecond fraction of "now" was more than the microsecond
3804 fraction of "then" (but the whole seconds number was less).
3806 38. Added support for the libradius Radius client library this is found on
3807 FreeBSD (previously only the radiusclient library was supported).
3813 1. When certain lookups returned multiple values in the form name=value, the
3814 quoting of the values was not always being done properly. Specifically:
3815 (a) If the value started with a double quote, but contained no whitespace,
3817 (b) If the value contained whitespace other than a space character (i.e.
3818 tabs or newlines or carriage returns) it was not quoted.
3819 This fix has been applied to the mysql and pgsql lookups by writing a
3820 separate quoting function and calling it from the lookup code. The fix
3821 should probably also be applied to nisplus, ibase and oracle lookups, but
3822 since I cannot test any of those, I have not disturbed their existing code.
3824 2. A hit in the callout cache for a specific address caused a log line with no
3825 reason for rejecting RCPT. Now it says "Previous (cached) callout
3826 verification failure".
3828 3. There was an off-by-one bug in the queryprogram router. An over-long
3829 return line was truncated at 256 instead of 255 characters, thereby
3830 overflowing its buffer with the terminating zero. As well as fixing this, I
3831 have increased the buffer size to 1024 (and made a note to document this).
3833 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3834 when Exim is waiting for an SMTP response from a remote server, Exim
3835 restarts its select() call on the socket, thereby resetting its timeout.
3836 This is not a problem when such interrupts are rare. Somebody set up a cron
3837 job to run exiwhat every 2 minutes, which is less than the normal select()
3838 timeout (5 or 10 minutes). This meant that the select() timeout never
3839 kicked in because it was always reset. I have fixed this by comparing the
3840 time when an interrupt arrives with the time at the start of the first call
3841 to select(). If more time than the timeout has elapsed, the interrupt is
3842 treated as a timeout.
3844 5. Some internal re-factoring in preparation for the addition of Sieve
3845 extensions (by MH). In particular, the "personal" test is moved to a
3846 separate function, and given an option for scanning Cc: and Bcc: (which is
3847 not set for Exim filters).
3849 6. When Exim created an email address using the login of the caller as the
3850 local part (e.g. when creating a From: or Sender: header line), it was not
3851 quoting the local part when it contained special characters such as @.
3853 7. Installed new OpenBSD configuration files.
3855 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3856 try to make them clearer.
3858 9. Callout options, other than the timeout value, were being ignored when
3859 verifying sender addresses in header lines. For example, when using
3861 verify = header_sender/callout=no_cache
3863 the cache was (incorrectly) being used.
3865 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3866 only the Cygwin environment.
3868 11. When return_path_on_delivery was set as a log selector, if different remote
3869 addresses in the same message used different return paths and parallel
3870 remote delivery occurred, the wrong values would sometimes be logged.
3871 (Whenever a remote delivery process finished, the return path value from
3872 the most recently started remote delivery process was logged.)
3874 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3875 lines when AUTH and/or TLS are in use. This is the "received protocol"
3876 field. Exim used to use "asmtp" for authenticated SMTP, without any
3877 indication (in the protocol name) for TLS use. Now it follows the RFC and
3878 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3879 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3880 names appear in log lines as well as in Received: header lines.
3882 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3883 extensions, and comparison tests, and to fix some bugs.
3885 14. Changes to the "personal" filter test:
3887 (1) The test was buggy in that it was just doing the equivalent of
3888 "contains" tests on header lines. For example, if a user's address was
3889 anne@some.where, the "personal" test would incorrectly be true for
3891 To: susanne@some.where
3893 This test is now done by extracting each address from the header in turn,
3894 and checking the entire address. Other tests that are part of "personal"
3895 are now done using regular expressions (for example, to check local parts
3896 of addresses in From: header lines).
3898 (2) The list of non-personal local parts in From: addresses has been
3899 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3900 taken from the Sieve specification recommendations.
3902 (3) If the message contains any header line starting with "List-" it is
3903 treated as non-personal.
3905 (4) The test for "circular" in the Subject: header line has been removed
3906 because it now seems ill-conceived.
3908 15. Minor typos in src/EDITME comments corrected.
3910 16. Installed latest exipick from John Jetmore.
3912 17. If headers_add on a router specified a text string that was too long for
3913 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3914 of string_sprintf() is now avoided.
3916 18. $message_body_size was not set (it was always zero) when running the DATA
3917 ACL and the local_scan() function.
3919 19. For the "mail" command in an Exim filter, no default was being set for
3920 the once_repeat time, causing a random time value to be used if "once" was
3921 specified. (If the value happened to be <= 0, no repeat happened.) The
3922 default is now 0s, meaning "never repeat". The "vacation" command was OK
3923 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3924 (I found it when inspecting the code).
3926 20. There is now an overall timeout for performing a callout verification. It
3927 defaults to 4 times the callout timeout, which applies to individual SMTP
3928 commands during the callout. The overall timeout applies when there is more
3929 than one host that can be tried. The timeout is checked before trying the
3930 next host. This prevents very long delays if there are a large number of
3931 hosts and all are timing out (e.g. when the network connections are timing
3932 out). The value of the overall timeout can be changed by specifying an
3933 additional sub-option for "callout", called "maxwait". For example:
3935 verify = sender/callout=5s,maxwait=20s
3937 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3938 to the end before writing, but this should make it even safer).
3940 22. Exim was forgetting that it had advertised PIPELINING for the second and
3941 subsequent messages on an SMTP connection. It was also not resetting its
3942 memory on STARTTLS and an internal HELO.
3944 23. When Exim logs an SMTP synchronization error within a session, it now
3945 records whether PIPELINING has been advertised or not.
3947 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3948 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3949 rather than long int.
3951 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3953 26. Added the never_mail option to autoreply.
3959 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3960 crash if the getsockname() call failed; this can happen if a connection is
3961 closed very soon after it is established. The problem was simply in the
3962 order in which certain operations were done, causing Exim to try to write
3963 to the SMTP stream before it had set up the file descriptor. The bug has
3964 been fixed by making things happen in the correct order.
3970 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3971 before the connection was closed, thus losing the rejection response.
3973 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3974 some early Solaris releases, but causes trouble in current releases where
3975 socklen_t is defined.
3977 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3980 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3983 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3984 that releasing the top of it at the end releases what was used for sub-
3985 expansions (unless the block got too big). However, discard this block if
3986 the first thing is a variable or header, so that we can use its block when
3987 it is dynamic (useful for very large $message_headers, for example).
3989 6. Lookups now cache *every* query, not just the most recent. A new, separate
3990 store pool is used for this. It can be recovered when all lookup caches are
3991 flushed. Lookups now release memory at the end of their result strings.
3992 This has involved some general refactoring of the lookup sources.
3994 7. Some code has been added to the store_xxx() functions to reduce the amount
3995 of flapping under certain conditions.
3997 8. log_incoming_interface used to affect only the <= reception log lines. Now
3998 it causes the local interface and port to be added to several more SMTP log
3999 lines, for example "SMTP connection from", and rejection lines.
4001 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4003 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4005 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4006 could be overwritten at the end of the current message (or the start of a
4007 new message if it was set in a HELO ACL). The value is now preserved for
4008 the duration of the SMTP connection.
4010 12. If a transport had a headers_rewrite setting, and a matching header line
4011 contained an unqualified address, that address was qualified, even if it
4012 did not match any rewriting rules. The underlying bug was that the values
4013 of the flags that permit the existence of unqualified sender and recipient
4014 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4015 non-local messages, and by -bnq for local messages) were not being
4016 preserved with the message after it was received.
4018 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4019 "next input=" as part of the text comprising the host identity instead of
4020 the correct text. The code was using the same buffer for two different
4021 strings. However, depending on which order the printing function evaluated
4022 its arguments, the bug did not always show up. Under Linux, for example, my
4023 test suite worked just fine.
4025 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4026 doesn't work with some older versions of Perl. It has been changed to "my",
4027 which in any case is probably the better facility to use.
4029 15. A really picky compiler found some instances of statements for creating
4030 error messages that either had too many or two few arguments for the format
4033 16. The size of the buffer for calls to the DNS resolver has been increased
4034 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4035 for addresses that have a lot of PTR records. This alleviates a problem; it
4036 does not fully solve it.
4038 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4039 buffer now truncates the list and logs the incident, which is the same
4040 action as happens when Exim is looking up a host name and its aliases.
4041 Previously in this situation something unpredictable would happen;
4042 sometimes it was "internal error: store_reset failed".
4044 18. If a server dropped the connection unexpectedly when an Exim client was
4045 using GnuTLS and trying to read a response, the client delivery process
4046 crashed while trying to generate an error log message.
4048 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4049 string, for example:
4051 warn message = H1: something\nH2: something
4053 the text was added as a single header line from Exim's point of view
4054 though it ended up OK in the delivered message. However, searching for the
4055 second and subsequent header lines using $h_h2: did not work. This has been
4056 fixed. Similarly, if a system filter added multiple headers in this way,
4057 the routers could not see them.
4059 20. Expanded the error message when iplsearch is called with an invalid key to
4060 suggest using net-iplsearch in a host list.
4062 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4063 ACLs are no longer actually imposed (and a message to that effect is
4066 22. If a "gecos" field in a passwd entry contained escaped characters, in
4067 particular, if it contained a \" sequence, Exim got it wrong when building
4068 a From: or a Sender: header from that name. A second bug also caused
4069 incorrect handling when an unquoted " was present following a character
4070 that needed quoting.
4072 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4073 was not being matched caselessly.
4075 24. Arranged for all hyphens in the exim.8 source to be escaped with
4078 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4079 properly. Recipient callout cache records were still being keyed to include
4080 the sender, even when use_sender was set false. This led to far more
4081 callouts that were necessary. The sender is no longer included in the key
4082 when use_sender is false.
4084 26. Added "control = submission" modifier to ACLs.
4086 27. Added the ${base62d: operator to decode base 62 numbers.
4088 28. dnsdb lookups can now access SRV records.
4090 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4091 the configuration file.
4093 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4094 (-v) mode. This makes the output for a verbose queue run more intelligible.
4096 31. Added a use_postmaster feature to recipient callouts.
4098 32. Added the $body_zerocount variable, containing the number of binary zero
4099 bytes in the message body.
4101 33. The time of last modification of the "new" subdirectory is now used as the
4102 "mailbox time last read" when there is a quota error for a maildir
4105 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4107 35. Added +ignore_unknown as a special item in host lists.
4109 36. Code for decoding IPv6 addresses in host lists is now included, even if
4110 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4111 address was recognized as an IP address, but was then not correctly decoded
4112 into binary, causing unexpected and incorrect effects when compared with
4119 1. Very minor rewording of debugging text in manualroute to say "list of
4120 hosts" instead of "hostlist".
4122 2. If verify=header_syntax was set, and a header line with an unqualified
4123 address (no domain) and a large number of spaces between the end of the
4124 name and the colon was received, the reception process suffered a buffer
4125 overflow, and (when I tested it) crashed. This was caused by some obsolete
4126 code that should have been removed. The fix is to remove it!
4128 3. When running in the test harness, delay a bit after writing a bounce
4129 message to get a bit more predictability in the log output.
4131 4. Added a call to search_tidyup() just before forking a reception process. In
4132 theory, someone could use a lookup in the expansion of smtp_accept_max_
4133 per_host which, without the tidyup, could leave open a database connection.
4135 5. Added the variables $recipient_data and $sender_data which get set from a
4136 lookup success in an ACL "recipients" or "senders" condition, or a router
4137 "senders" option, similar to $domain_data and $local_part_data.
4139 6. Moved the writing of debug_print from before to after the "senders" test
4142 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4143 problems for message scanning, either using a data ACL, or using
4144 local_scan() because the Received: header was not generated till after they
4145 were called (in order to set the time as the time of reception completion).
4146 I have revised the way this works. The header is now generated after the
4147 body is received, but before the ACL or local_scan() are called. After they
4148 are run, the timestamp in the header is updated.
4154 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4155 before starting a queue runner without re-exec. This happened only when
4156 deliver_drop_privilege was set or when the Exim user was set to root. The
4157 effect of the bug was that timeouts during subsequent deliveries caused
4158 crashes instead of being properly handled. The handler is now left at its
4159 default (and expected) setting.
4161 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4162 message, again when deliver_drop_privilege is set or Exim is run as root.
4163 The bug described in (1) was not present in this case, but the tidying up
4164 of the other signals was missing. I have made the two cases consistent.
4166 3. The ignore_target_hosts setting on a manualroute router was being ignored
4167 for hosts that were looked up using the /MX notation.
4169 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4172 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4173 operated on the sender address. After changing the $sender_address to <>
4174 for the sender address verify, Exim was re-instated it as the original
4175 (before rewriting) address, but remembering that it had rewritten it, so it
4176 wasn't rewriting it again. This bug also had the effect of breaking the
4177 sender address verification caching when the sender address was rewritten.
4179 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4180 This has been changed so that if the ip literal address matches
4181 ignore_target_hosts, the router declines.
4183 7. Added expansion conditions match_domain, match_address, and match_local_
4184 part (NOT match_host).
4186 8. The placeholder for the Received: header didn't have a length field set.
4188 9. Added code to Exim itself and to exim_lock to test for a specific race
4189 condition that could lead to file corruption when using MBX delivery. The
4190 issue is with the lockfile that is created in /tmp. If this file is removed
4191 after a process has opened it but before that process has acquired a lock,
4192 there is the potential for a second process to recreate the file and also
4193 acquire a lock. This could lead to two Exim processes writing to the file
4194 at the same time. The added code performs the same test as UW imapd; it
4195 checks after acquiring the lock that its file descriptor still refers to
4196 the same named file.
4198 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4199 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4202 11. Added the smtp_active_hostname option. If used, this will typically be made
4203 to depend on the incoming interface address. Because $interface_address is
4204 not set up until the daemon has forked a reception process, error responses
4205 that can happen earlier (such as "too many connections") no longer contain
4208 12. If an expansion in a condition on a "warn" statement fails because a lookup
4209 defers, the "warn" statement is abandoned, and the next ACL statement is
4210 processed. Previously this caused the whole ACL to be aborted.
4212 13. Added the iplsearch lookup type.
4214 14. Added ident_timeout as a log selector.
4216 15. Added tls_certificate_verified as a log selector.
4218 16. Added a global option tls_require_ciphers (compare the smtp transport
4219 option of the same name). This controls incoming TLS connections.
4221 17. I finally figured out how to make tls_require_ciphers do a similar thing
4222 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4223 before starting the TLS session.
4225 18. Tabs are now shown as \t in -bP output.
4227 19. If the log selector return_path_on_delivery was set, Exim crashed when
4228 bouncing a message because it had too many Received: header lines.
4230 20. If two routers both had headers_remove settings, and the first one included
4231 a superfluous trailing colon, the final name in the first list and the
4232 first name in the second list were incorrectly joined into one item (with a
4233 colon in the middle).
4239 1. Added -C and -D options to the exinext utility, mainly to make it easier
4240 to include in the automated testing, but these could be helpful when
4241 multiple configurations are in use.
4243 2. The exinext utility was not formatting the output nicely when there was
4244 an alternate port involved in the retry record key, nor when there was a
4245 message id as well (for retries that were specific to a specific message
4246 and a specific host). It was also confused by IPv6 addresses, because of
4247 the additional colons they contain. I have fixed the IPv4 problem, and
4248 patched it up to do a reasonable job for IPv6.
4250 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4251 delivery, the log line now contains "pipelined" if PIPELINING was used.
4253 4. An SMTP transport process used to panic and die if the bind() call to set
4254 an explicit outgoing interface failed. This has been changed; it is now
4255 treated in the same way as a connect() failure.
4257 5. A reference to $sender_host_name in the part of a conditional expansion
4258 that was being skipped was still causing a DNS lookup. This no longer
4261 6. The def: expansion condition was not recognizing references to header lines
4262 that used bh_ and bheader_.
4264 7. Added the _cache feature to named lists.
4266 8. The code for checking quota_filecount in the appendfile transport was
4267 allowing one more file than it should have been.
4269 9. For compatibility with Sendmail, the command line option
4277 and sets the incoming protocol and host name (for trusted callers). The
4278 host name and its colon can be omitted when only the protocol is to be set.
4279 Note the Exim already has two private options, -pd and -ps, that refer to
4280 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4281 "s", but I don't think that's a major issue.
4283 10. A number of refactoring changes to the code, none of which should affect
4286 (a) The number of logging options was getting close to filling up the
4287 32-bit word that was used as a bit map. I have split them into two classes:
4288 those that are passed in the argument to log_write(), and those that are
4289 only ever tested independently outside of that function. These are now in
4290 separate 32-bit words, so there is plenty of room for expansion again.
4291 There is no change in the user interface or the logging behaviour.
4293 (b) When building, for example, log lines, the code previously used a
4294 macro that called string_cat() twice, in order to add two strings. This is
4295 not really sufficiently general. Furthermore, there was one instance where
4296 it was actually wrong because one of the argument was used twice, and in
4297 one call a function was used. (As it happened, calling the function twice
4298 did not affect the overall behaviour.) The macro has been replaced by a
4299 function that can join an arbitrary number of extra strings onto a growing
4302 (c) The code for expansion conditions now uses a table and a binary chop
4303 instead of a serial search (which was left over from when there were very
4304 few conditions). Also, it now recognizes conditions like "pam" even when
4305 the relevant support is not compiled in: a suitably worded error message is
4306 given if an attempt is made to use such a condition.
4308 11. Added ${time_interval:xxxxx}.
4310 12. A bug was causing one of the ddress fields not to be passed back correctly
4311 from remote delivery subprocesses. The field in question was not being
4312 subsequently used, so this caused to problems in practice.
4314 13. Added new log selectors queue_time and deliver_time.
4316 14. Might have fixed a bug in maildirsizefile handling that threw up
4317 "unexpected character" debug warnings, and recalculated the data
4318 unnecessarily. In any case, I expanded the warning message to give more
4321 15. Added the message "Restricted characters in address" to the statements in
4322 the default ACL that block characters like @ and % in local parts.
4324 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4325 Three changes have been made:
4327 (a) There was a serious bug; a negative response to MAIL caused the whole
4328 recipient domain to be cached as invalid, thereby blocking all messages
4329 to all local parts at the same domain, from all senders. This bug has
4330 been fixed. The domain is no longer cached after a negative response to
4331 MAIL if the sender used is not empty.
4333 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4336 (c) A new callout option, "use_sender" has been added for people who want
4337 the modified behaviour.
4343 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4346 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4347 indeed breaks things for older releases.
4349 3. Added additional logging to the case where there is a problem reading data
4350 from a filter that is running in a subprocess using a pipe, in order to
4351 try to track down a specific problem.
4353 4. Testing facility fudge: when running in the test harness and attempting
4354 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4355 getting "No route to host". Convert this to a timeout.
4357 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4360 6. Some OS don't have socklen_t but use size_t instead. This affects the
4361 fifth argument of getsockopt() amongst other things. This is now
4362 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4363 can be set for individual OS. I have set it for SunOS5, OSF1, and
4364 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4365 some earlier ones do not.
4367 7. Change 4.30/15 was not doing the test caselessly.
4369 8. The standard form for an IPv6 address literal was being rejected by address
4370 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4371 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4372 this, as well as the form without the "IPv6" on the front (but only when
4373 address literals are enabled, of course).
4375 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4377 10. Exim crashed if a message with an empty sender address specified by -f
4378 encountered a router with an errors_to setting. This could be provoked only
4379 by a command such as
4383 where an empty string was supplied; "<>" did not hit this bug.
4385 11. Installed PCRE release 4.5.
4387 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4388 remained set. It is now erased.
4390 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4391 times from message ids (which are base 36 rather than the normal 62).
4393 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4394 were being counted as actual protocol errors, and logged if the log
4395 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4396 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4397 and DATA following a set of rejected RCPTs do not count as protocol errors.
4398 In other words, Exim assumes they were pipelined, though this may not
4399 actually be the case. Of course, in all cases the client gets an
4400 appropriate error code.
4402 15. If a lookup fails in an ACL condition, a message about the failure may
4403 be available; it is used if testing the ACL cannot continue, because most
4404 such messages specify what the cause of the deferral is. However, some
4405 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4406 that caused an old message to be retained and used if a later statement
4407 caused a defer, replacing the real cause of the deferral.
4409 16. If an IP address had so many PTR records that the DNS lookup buffer
4410 was not large enough to hold them, Exim could crash while trying to process
4411 the truncated data. It now detects and logs this case.
4413 17. Further to 4.21/58, another change has been made: if (and only if) the
4414 first line of a message (the first header line) ends with CRLF, a bare LF
4415 in a subsequent header line has a space inserted after it, so as not to
4416 terminate the header.
4418 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4419 unnecessarily, used atoi() instead of strtol(), and didn't check the
4420 termination when getting file sizes from file names by regex.
4422 19. Completely re-implemented the support for maildirsize files, in the light
4423 of a number of problems with the previous contributed implementation
4424 (4.30/29). In particular:
4426 . If the quota is zero, the maildirsize file is maintained, but no quota is
4429 . If the maildir directory does not exist, it is created before any attempt
4430 to write a maildirsize file.
4432 . The quota value in the file is just a cache; if the quota is changed in
4433 the transport, the new value overrides.
4435 . A regular expression is available for excluding directories from the
4438 20. The autoreply transport checks the characters in options that define the
4439 message's headers; it allows continued headers, but it was checking with
4440 isspace() after an embedded newline instead of explicitly looking for a
4443 21. If all the "regular" hosts to which an address was routed had passed their
4444 expiry times, and had not reached their retry times, the address was
4445 bounced, even if fallback hosts were defined. Now Exim should go on to try
4448 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4449 equivalent code in the SMTP transport. Some hosts send humungous responses
4450 to HELO/EHLO, more than 1024 it seems.
4452 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4453 gives compiler warnings in some environments. I've now done it "properly",
4456 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4457 (because of problems with the built-in one) was declared to return uschar *
4458 instead of char *, causing compiler failure.
4460 25. Fixed a file descriptor leak when processing alias/forward files.
4462 26. Fixed a minor format string issue in dbfn.c.
4464 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4466 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4467 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4470 29. When a sender address is verified, it is cached, to save repeating the test
4471 when there is more than one recipient in a message. However, when the
4472 verification involves a callout, it is possible for different callout
4473 options to be set for different recipients. It is too complicated to keep
4474 track of this in the cache, so now Exim always runs a verification when a
4475 callout is required, relying on the callout cache for the optimization.
4476 The overhead is duplication of the address routing, but this should not be
4479 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4480 to be verified with callout=postmaster, and the main callout worked but the
4481 postmaster check failed, the verification correctly failed. However, if a
4482 subsequent RCPT command asked for sender verification *without* the
4483 postmaster check, incorrect caching caused this verification also to fail,
4486 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4487 it was not caching the DNS options (qualify_single, search_parents) that
4488 were used when the lookup failed. A subsequent lookup with different
4489 options therefore always gave the same answer, though there were cases
4490 where it should not have. (Example: a "domains = !$mx_any" option on a
4491 dnslookup router: the "domains" option is always processed without any
4492 widening, but the router might have qualify_single set.) Now Exim uses the
4493 cached value only when the same options are set.
4495 32. Added John Jetmore's "exipick" utility to the distribution.
4497 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4498 than a timeout (e.g. a certificate is required, and is not provided), an
4499 Exim server now closes the connection immediately. Previously it waited for
4500 the client to close - but if the client is SSL, it seems that they each
4501 wait for each other, leading to a delay before one of them times out.
4503 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4504 maintained 0.8.x compatibility because I don't think many are using it, and
4505 it is clearly obsolete.
4507 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4510 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4511 client certificate was expired. A simple patch fixes this, though I don't
4512 understand the full logic of why the verify callback is called multiple
4515 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4516 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4517 which causes problems with some clients (such as the Certicom SSL Plus
4518 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4519 disables the coutermeasure allowing Eudora to connect."
4521 38. Exim was not checking that a write() to a log file succeeded. This could
4522 lead to Bad Things if a log got too big, in particular if it hit a file
4523 size limit. Exim now panics and dies if it cannot write to a log file, just
4524 as it does if it cannot open a log file.
4526 39. Modified OS/Makefile-Linux so that it now contains
4528 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4530 The two -D definitions ensure that Exim is compiled with large file
4531 support, which makes it possible to handle log files that are bigger than
4534 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4535 instance) a domain was checked against a named list that involved a lookup,
4536 causing $domain_data to be set, then another domain was checked against the
4537 same list, then the first domain was re-checked, the value of $domain_data
4538 after the final check could be wrong. In particular, if the second check
4539 failed, it could be set empty. This bug probably also applied to
4542 41. The strip_trailing_dot option was not being applied to the address given
4543 with the -f command-line option.
4545 42. The code for reading a message's header from the spool was incrementing
4546 $received_count, but never initializing it. This meant that the value was
4547 incorrect (doubled) while delivering a message in the same process in which
4548 it was received. In the most common configuration of Exim, this never
4549 happens - a fresh exec is done - but it can happen when
4550 deliver_drop_privilege is set.
4552 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4553 it now includes up to 150 characters of the unexpected data in the log
4556 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4557 and building data strings. The size of both of these buffers was 10 000
4558 bytes - far larger than anybody would *ever* want, thought I. Needless to
4559 say, somebody hit the limit. I have increased the maximum line length to
4560 20 000 and the maximum data length of concatenated lines to 100 000. I have
4561 also fixed two bugs, because there was no checking on these buffers. Tsk,
4562 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4563 buffer is too small.
4565 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4566 lsearch lookups. Now it does.
4568 46. When parsing a route_list item in a manualroute router, a fixed-length
4569 buffer was used for the list of hosts. I made this 1024 bytes long,
4570 thinking that nobody would ever have a list of hosts that long. Wrong.
4571 Somebody had a whole pile of complicated expansion conditions, and the
4572 string was silently truncated, leading to an expansion error. It turns out
4573 that it is easier to change to an unlimited length (owing to other changes
4574 that have happened since this code was originally written) than to build
4575 structure for giving a limitation error. The length of the item that
4576 expands into the list of hosts is now unlimited.
4578 47. The lsearch lookup could not handle data where the length of text line was
4579 more than 4095 characters. Such lines were truncated, leading to shortened
4580 data being returned. It should now handle lines of any length.
4582 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4583 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4586 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4587 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4588 no longer happens in comment lines. A list of replacements is now placed
4589 at the head of all of the source files, except those whose only change is
4590 to replace PERL_COMMAND in the very first #! line.
4592 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4593 messages on the queue, with a bottom-up merge sort, using code contributed
4594 by Michael Haardt. This should make operations like -bp somewhat faster on
4595 large queues. It won't affect queue runners, except when queue_run_in_order
4598 51. Installed eximstats 1.31 in the distribution.
4600 52. Added support for SRV lookups to the dnslookup router.
4602 53. If an ACL referred to $message_body or $message_body_end, the value was not
4603 reset for any messages that followed in the same SMTP session.
4605 54. The store-handling optimization for building very long strings was not
4606 differentiating between the different store pools. I don't think this
4607 actually made any difference in practice, but I've tidied it.
4609 55. While running the routers to verify a sender address, $sender_address
4610 was still set to the sender address. This is wrong, because when routing to
4611 send a bounce to the sender, it would be empty. Therefore, I have changed
4612 it so that, while verifying a sender address, $sender_address is set to <>.
4613 (There is no change to what happens when verifying a recipient address.)
4615 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4616 target A or AAAA records (if not already returned) without resetting the
4617 qualify_single or search_parents options of the DNS resolver. These are
4618 inappropriate in this case because the targets of MX and SRV records must
4619 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4620 target that, when qualified, matched something in the local domain. These
4621 two options are now turned off when doing these lookups.
4623 57. It seems that at least some releases of Reiserfs (which does not have the
4624 concept of a fixed number of inodes) returns zero and not -1 for the
4625 number of available inodes. This interacted badly with check_spool_inodes,
4626 which assumed that -1 was the "no such thing" setting. What I have done is
4627 to check that the total number of inodes is greater than zero before doing
4628 the test of how many are available.
4630 58. When a "warn" ACL statement has a log_message modifier, the message is
4631 remembered, and not repeated. This is to avoid a lot of repetition when a
4632 message has many recipients that cause the same warning to be written.
4633 Howewer, Exim was preserving the list of already written lines for an
4634 entire SMTP session, which doesn't seem right. The memory is now reset if a
4635 new message is started.
4637 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4638 debugging output unless log_rewrite was also set.
4640 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4641 of (int)(handle) when we know that handle contains (void *)(-1).
4643 61. The Exim daemon panic-logs an error return when it closes the incoming
4644 connection. However "connection reset by peer" seems to be common, and
4645 isn't really an error worthy of noting specially, so that particular error
4648 62. When Exim is trying to find all the local interfaces, it used to panic and
4649 die if the ioctl to get the interface flags failed. However, it seems that
4650 on at least one OS (Solaris 9) it is possible to have an interface that is
4651 included in the list of interfaces, but for which you get a failure error
4652 for this call. This happens when the interface is not "plumbed" into a
4653 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4654 failure of the "get flags" call assumes that the interface is down.
4656 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4657 makes life easier for people who are doing arithmetic on fields extracted
4658 from dates, where you often get leading zeros that should not be
4659 interpreted as octal.
4661 64. Added qualify_domain to the redirect router, to override the global
4664 65. If a pathologically long header line contained very many addresses (the
4665 report of this problem mentioned 10 000) and each of them was rewritten,
4666 Exim could use up a very large amount of memory. (It kept on making new
4667 copies of the header line as it rewrote, and never released the old ones.)
4668 At the expense of a bit more processing, the header rewriting function has
4669 been changed so that it no longer eats memory in this way.
4671 66. The generation of the Received: header has been moved from the time that a
4672 message starts to be received, to the time that it finishes. The timestamp
4673 in the Received: header should now be very close to that of the <= log
4674 line. There are two side-effects of this change:
4676 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4677 logged header lines no longer include the local Received: line, because
4678 it has not yet been created. The same applies to a copy of the message
4679 that is returned to a non-SMTP sender when a message is rejected.
4681 (b) When a filter file is tested using -bf, no additional Received: header
4682 is added to the test message. After some thought, I decided that this
4685 This change does not affect the value of $received_for. It is still set
4686 after address rewriting, but before local_scan() is called.
4688 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4690 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4691 gave an unhelpful panic error message, and a defer error. I have managed to
4692 change this behaviour so that it now rejects any supplied certificate,
4693 which seems right, as the list of acceptable certificates is empty.
4695 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4696 gave an unhelpful defer error. I have not managed to make this reject any
4697 supplied certificates, but the error message it gives is "no certificate
4698 supplied", which is not helpful.
4700 70. exigrep's output now also includes lines that are not associated with any
4701 message, but which match the given pattern. Implemented by a patch from
4702 Martin Sluka, which also tidied up the Perl a bit.
4704 71. Recipient callout verification, like sender verification, was using <> in
4705 the MAIL FROM command. This isn't really the right thing, since the actual
4706 sender may affect whether the remote host accepts the recipient or not. I
4707 have changed it to use the actual sender in the callout; this means that
4708 the cache record is now keyed on a recipient/sender pair, not just the
4709 recipient address. There doesn't seem to be a real danger of callout loops,
4710 since a callout by the remote host to check the sender would use <>.
4711 [SEE ABOVE: changed after hitting problems.]
4713 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4714 temporary errors. However, in the case of such a code being given after
4715 the end of a data transmission (i.e. after ".") Exim was failing to write
4716 a retry record for the message. (Yes, there was some broken host that was
4717 actually sending 8xx at this point.)
4719 73. An unknown lookup type in a host list could cause Exim to panic-die when
4720 the list was checked. (An example that provoked this was putting <; in the
4721 middle of a list instead of at the start.) If this happened during a DATA
4722 ACL check, a -D file could be left lying around. This kind of configuration
4723 error no longer causes Exim to die; instead it causes a defer errror. The
4724 incident is still logged to the main and panic logs.
4726 74. Buglet left over from Exim 3 conversion. The message "too many messages
4727 in one connection" was written to the rejectlog but not the mainlog, except
4728 when address rewriting (yes!) was being logged.
4730 75. Added write_rejectlog option.
4732 76. When a system filter was run not as root (that is, when system_filter_user
4733 was set), the values of the $n variables were not being returned to the
4734 main process; thus, they were not subsequently available in the $sn
4737 77. Added +return_path_on_delivery log selector.
4739 78. A connection timeout was being treated differently from recipients deferred
4740 when testing hosts_max_try with a message that was older than the host's
4741 retry timeout. (The host should not be counted, thus allowing all hosts to
4742 be tried at least once before bouncing.) This may have been the cause of an
4743 occasionally reported bug whereby a message would remain on the queue
4744 longer than the retry timeout, but would be bounced if a delivery was
4745 forced. I say "may" because I never totally pinned down the problem;
4746 setting up timeout/retry tests is difficult. See also the next item.
4748 79. The ultimate address timeout was not being applied to errors that involved
4749 a combination of host plus message (for example, a timeout on a MAIL
4750 command). When an address resolved to a number of possible hosts, and they
4751 were not all tried for each delivery (e.g. because of hosts_max_try), a
4752 message could remain on the queue longer than the retry timeout.
4754 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4757 81. Fixed an obscure SMTP outgoing bug which required at least the following
4758 conditions: (a) there was another message waiting for the same server;
4759 (b) the server returned 5xx to all RCPT commands in the first message so
4760 that the message was not completed; (c) the server dropped the connection
4761 or gave a negative response to the RSET that Exim sends to abort the
4762 transaction. The observed case was a dropped connection after DATA that had
4763 been sent in pipelining mode. That is, the server had advertised PIPELINING
4764 but was not implementing it correctly. The effect of the bug was incorrect
4765 behaviour, such as trying another host, and this could lead to a crash.
4771 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4772 and daemon.c were passed as pointers to ints; they should have been
4773 pointers to socklen_t variables (which are typically unsigned ints).
4775 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4778 3. Fixed a really odd bug that affected only the testing scheme; patching a
4779 certain fixed string in the binary changed the value of another string that
4780 happened to be identical to the end of the original first string.
4782 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4783 name", it returns that address as the IP address. On some operating
4784 systems (e.g. Solaris), it also passes back the IP address string as the
4785 "host name". However, on others (e.g. Linux), it passes back an empty
4786 string. Exim wasn't checking for this, and was changing the host name to an
4787 empty string, assuming it had been canonicized.
4789 5. Although rare, it is permitted to have more than one PTR record for a given
4790 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4791 all the names associated with an address, because they do in Solaris.
4792 However, it seems that they do not in Linux for data that comes from the
4793 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4794 I found this out when I moved to a new Linux workstation and tried to run
4795 the Exim test suite.
4797 To get round this problem I have changed the code so that it now does its
4798 own call to the DNS to look up PTR records when searching for a host name.
4799 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4800 addresses that are only in /etc/hosts are still found.
4802 This behaviour is, however, controlled by an option called host_lookup_
4803 order, which defaults to "bydns:byaddr". If people want to use the other
4804 order, or indeed, just use one or the other means of lookup, they can
4805 specify it in this variable.
4807 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4808 some operating systems, this comes back from gethostbyaddr() as an empty
4809 string, and this is what Exim used to test for. However, it seems that in
4810 other systems, "." is yielded. Exim now tests for this case too.
4812 7. The values of check_spool_space and check_log_space are now held internally
4813 as a number of kilobytes instead of an absolute number of bytes. If a
4814 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4815 kilobyte. This means that much larger values can be stored.
4817 8. Exim monitor: an attempt to get the action menu when not actually pointing
4818 at a message produces an empty menu entitled "No message selected". This
4819 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4820 no entries in it ("Shell widget menu has zero width and/or height"). So I
4821 have added a single, blank menu entry in this case.
4823 9. Added ${quote_local_part.
4825 10. MIME decoding is now applied to the contents of Subject: header lines when
4828 11. Now that a reference to $sender_host_address automatically causes a reverse
4829 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4830 host lookup before query-style lookups in lists that might use this
4831 variable. This has therefore been abolished, and the "net-" prefix is no
4832 longer necessary for query-style lookups.
4834 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4835 have been a typo for LFLAGS, so it has been changed.
4837 13. The install script calls Exim with "-C /dev/null" in order to find the
4838 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4839 to be output. Howeve, since Exim outputs its version number before the
4840 error, it didn't break the script. It just looked ugly. I fixed this by
4841 always allowing "-C /dev/null" if the caller is root.
4843 14. Ignore overlarge ACL variable number when reading spool file - insurance
4844 against a later release with more variables having written the file.
4846 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4847 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4848 this, as well as the form without the "IPv6" on the front.
4850 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4851 OS/Makefile-Darwin file.
4853 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4854 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4856 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4857 variable. However, some OS use ints for inodes. I've added cast to long int
4858 to get rid of the compiler warning.
4860 19. I had forgotten to lock out "/../" in configuration file names when
4861 ALT_CONFIG_PREFIX was set.
4863 20. Routers used for verification do not need to specify transports. However,
4864 if such a router generated a host list, and callout was configured, Exim
4865 crashed, because it could not find a port number from the (non-existent)
4866 transport. It now assumes port 25 in this circumstance.
4868 21. Added the -t option to exigrep.
4870 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4871 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4872 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4873 them is set, LOOKUP_LSEARCH is forced.
4875 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4877 24. Added sender and host information to the "rejected by local_scan()" log
4878 line; previously there was no indication of these.
4880 25. Added .include_if_exists.
4882 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4883 Linux. It turns out that not all file systems support this. Apparently some
4884 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4885 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4886 not supported on the file descriptor, is now ignored when Exim is trying to
4887 sync a directory. This applies only to Linux.
4889 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4891 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4892 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4895 29. Added support for maildirsize files from supplied patch (modified a bit).
4897 30. The use of :fail: followed by an empty string could lead Exim to respond to
4898 sender verification failures with (e.g.):
4900 550 Verification failed for <xxx>
4901 550 Sender verify failed
4903 where the first response line was missing the '-' that indicates it is not
4904 the final line of the response.
4906 31. The loop for finding the name of the user that called Exim had a hardwired
4907 limit of 10; it now uses the value of finduser_retries, which is used for
4908 all other user lookups.
4910 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4913 33. Exim was neglecting to zero errno before one call of strtol() when
4914 expanding a string and expecting an integer value. On some systems this
4915 resulted in spurious "integer overflow" errors. Also, it was casting the
4916 result into an int without checking.
4918 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4919 did not work. The code looks as if it has *never* worked, though it appears
4920 to have been documented since at least releast 1.62. I have made it work.
4922 35. The "timeout_DNS" error in retry rules, also documented since at least
4923 1.62, also never worked. As it isn't clear exactly what this means, and
4924 clearly it isn't a major issue, I have abolished the feature by treating it
4925 as "timeout", and writing a warning to the main and panic logs.
4927 36. The display of retry rules for -brt wasn't always showing the error code
4930 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4931 timeout_connect_A, timeout_connect_MX.
4933 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4934 to the empty sender.
4936 39. The daemon was not analysing the content of -oX till after it had closed
4937 stderr and disconnected from the controlling terminal. This meant that any
4938 syntax errors were only noted on the panic log, and the return code from
4939 the command was 0. By re-arranging the code a little, I've made the
4940 decoding happen first, so such errors now appear on stderr, and the return
4941 code is 1. However, the actual setting up of the sockets still happens in
4942 the disconnected process, so errors there are still only recorded on the
4945 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4946 connections (as happens on some IP stacks) was logged at start up time as
4947 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4948 it from "IPv6 and IPv4", which means that two separate sockets are being
4951 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4952 says whether AF_INET or AF_INET6 was passed as an argument.
4954 42. Exiwhat output was messed up when time zones were included in log
4957 43. Exiwhat now gives more information about the daemon's listening ports,
4958 and whether -tls-on-connect was used.
4960 44. The "port" option of the smtp transport is now expanded.
4962 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4963 silently ignored. Now an error message is written to the main and panic
4966 46. There's a new ACL modifier called "logwrite" which writes to a log file
4967 as soon as it is encountered.
4969 47. Added $local_user_uid and $local_user_gid at routing time.
4971 48. Exim crashed when trying to verify a sender address that was being
4974 49. Exim was recognizing only a space character after ".include". It now also
4975 recognizes a tab character.
4977 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4978 extracting the relevant information from the specification. The man page no
4979 longer contains scrambled data for the -d option, and I've added a section
4980 at the front about calling Exim under different names.
4982 51. Added "extra_headers" argument to the "mail" command in filter files.
4984 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4987 53. Installed eximstats 1.29.
4989 54. Added transport_filter_timeout as a generic transport option.
4991 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4992 Cc: header lines. This was required by RFC 822, but it not required by RFC
4995 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4996 incoming messages that did not have them. Now it does so only if the
4997 message originates locally, that is, if there is no associated remote host
4998 address. When Resent- header lines are present, this applies to the Resent-
4999 lines rather than the non-Resent- lines.
5001 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5002 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5004 58. Messages for configuration errors now include the name of the main
5005 configuration file - useful now that there may be more than one file in a
5006 list (.included file names were always shown).
5008 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5009 for those rare installations that do not start the daemon as root or run it
5010 setuid root. I've cut out the call to initgroups() if the daemon is not
5013 60. The Exim user and group can now be bound into the binary as text strings
5014 that are looked up at the start of Exim's processing.
5016 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5018 62. Added $mailstore_basename variable.
5020 63. Installed patch to sieve.c from Michael Haardt.
5022 64. When Exim failed to open the panic log after failing to open the main log,
5023 the original message it was trying to log was written to stderr and debug
5024 output, but if they were not available (the usual case in production), it
5025 was lost. Now it is written to syslog before the two lines that record the
5026 failures to open the logs.
5028 65. Users' Exim filters run in subprocesses under the user's uid. It is
5029 possible for a "deliver" command or an alias in a "personal" command to
5030 provoke an address rewrite. If logging of address rewriting is configured,
5031 this fails because the process is not running as root or exim. There may be
5032 a better way of dealing with this, but for the moment (because 4.30 needs
5033 to be released), I have disabled address rewrite logging when running a
5034 filter in a non-root, non-exim process.
5040 1. The buildconfig auxiliary program wasn't quoting the value set for
5041 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5042 not defined. This bug was masked in 4.22 by the effect that was fixed in
5045 2. Some messages that were rejected after a message id was allocated were
5046 shown as "incomplete" by exigrep. It no longer does this for messages that
5047 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5049 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5050 have allow_domain_literals set, the ID did not get logged in the <= line.
5051 Domain literals are now always recognized in Message-ID: header lines.
5053 4. The first argument for a ${extract expansion item is the key name or field
5054 number. Leading and trailing spaces in this item were not being ignored,
5055 causing some misleading effects.
5057 5. When deliver_drop_privilege was set, single queue runner processes started
5058 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5059 same command in the process it spins off) were not dropping privilege.
5061 6. When the daemon running as "exim" started a queue runner, it always
5062 re-executed Exim in the spun-off process. This is a waste of effort when
5063 deliver_drop_privilege is set. The new process now just calls the
5064 queue-runner function directly.
5070 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5073 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5074 ignored. Though the use of -oP was forcing the writing of a pid file, it
5075 was always written to the default place.
5077 3. If the message "no IP address found for host xxxx" is generated during
5078 incoming verification, it is now followed by identification of the incoming
5079 connection (so you can more easily find what provoked it).
5081 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5083 5. Added some features to "harden" Exim a bit more against certain attacks:
5085 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5086 be put in Local/Makefile. This is like the never_users runtime option,
5087 but it cannot be overridden. The default setting is "root".
5089 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5090 prefix string with which any file named in a -C command line option
5093 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5094 is retained for -C and -D only if the caller of Exim is root. Without
5095 it, the exim user may also use -C and -D and retain privilege.
5097 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5098 command line option is disabled.
5100 6. Macro names set by the -D option must start with an upper case letter, just
5101 like macro names defined in the configuration file.
5103 7. Added "dereference=" facility to LDAP.
5105 8. Two instances of the typo "uknown" in the source files are fixed.
5107 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5108 the Configure-Makefile script screwed up while processing it.
5110 10. Incorporated PCRE 4.4.
5112 11. The SMTP synchronization check was not operating right at the start of an
5113 SMTP session. For example, it could not catch a HELO sent before the client
5114 waited for the greeting. There is now a check for outstanding input at the
5115 point when the greeting is written. Because of the duplex, asynchronous
5116 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5117 way, but not yet received, when the check is performed.
5119 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5120 on TCP/IP sockets, because this apparently causes some broken clients to
5123 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5124 unchanged) from the Cygwin maintainer.
5126 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5127 was supported instead of testing for mbx. Effectively a typo.
5129 15. The spa authenticator server code was not checking that the input it
5130 received was valid base64.
5132 16. The debug output line for the "set" modifier in ACLs was not showing the
5133 name of the variable that was being set.
5135 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5137 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5138 lookup on its own. Something else was needed to trigger the lookup. For
5139 example, a match in host_lookup or the need for a host name in a host list.
5140 Now, if $sender_host_name is referenced and the host name has not yet been
5141 looked up, a lookup is performed. If the lookup fails, the variable remains
5142 empty, and $host_lookup_failed is set to "1".
5144 19. Added "eqi" as a case-independent comparison operator.
5146 20. The saslauthd authentication condition could segfault if neither service
5147 nor realm was specified.
5149 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5150 error message that was logged was misleading, and incoming SMTP
5151 connections were dropped. The message is now more accurate, and temporary
5152 errors are given to SMTP connections.
5154 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5155 (except RSET) with a 421 error, until QUIT is received. However, it was
5156 failing to send a response to QUIT.
5158 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5159 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5160 ACL is now run after the helo_try_verify_hosts code.
5162 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5163 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5164 case-independent, so other case variants are also recognized). Apparently
5165 some systems use these upper case variants.
5167 25. If more than two messages were waiting for the same host, and a transport
5168 filter was specified for the transport, Exim sent two messages over the
5169 same TCP/IP connection, and then failed with "socket operation on non-
5170 socket" when it tried to send the third.
5172 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5174 27. The extern definition of crypt16() in expand.c was not being excluded when
5175 the OS had its own crypt16() function.
5177 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5178 as a preferred synonym for return_size_limit, both as an option and as an
5181 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5183 30. Changed the default configuration ACL to relax the local part checking rule
5184 for addresses that are not in any local domains. For these addresses,
5185 slashes and pipe symbols are allowed within local parts, but the sequence
5186 /../ is explicitly forbidden.
5188 31. SPA server authentication was not clearing the challenge buffer before
5191 32. log_message in a "warn" ACL statement was writing to the reject log as
5192 well as to the main log, which contradicts the documentation and doesn't
5193 seem right (because no rejection is happening). So I have stopped it.
5195 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5196 However, I am unable to do any testing of this.
5198 34. Fixed an infelicity in the appendfile transport. When checking directories
5199 for a mailbox, to see if any needed to be created, it was accidentally
5200 using path names with one or more superfluous leading slashes; tracing
5201 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5203 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5204 contents are added to the log line that is written for every discarded
5205 recipient. (Previously a log_message setting was ignored.)
5207 36. The ${quote: operator now quotes the string if it is empty.
5209 37. The install script runs exim in order to find its version number. If for
5210 some reason other than non-existence or emptiness, which it checks, it
5211 could not run './exim', it was installing it with an empty version number,
5212 i.e. as "exim-". This error state is now caught, and the installation is
5215 38. An argument was missing from the function that creates an error message
5216 when Exim fails to connect to the socket for saslauthd authentication.
5217 This could cause Exim to crash, or give a corrupted message.
5219 39. Added isip, isip4, and isip6 to ${if conditions.
5221 40. The ACL variables $acl_xx are now saved with the message, and can be
5222 accessed later in routers, transports, and filters.
5224 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5225 strings in the file are not string-expanded.
5227 42. If a MAIL command specified a SIZE value that was too large to fit into an
5228 int variable, the check against message_size_limit failed. Such values are
5229 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5230 day this will have to be increased, but I don't think I want to be around
5231 when emails are that large.
5238 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5239 iconv() is not standard in FreeBSD.
5241 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5242 IPv6 enabled. The observed symptom was a segmentation fault on return from
5243 the function os_common_find_running_interfaces() in src/os.c.
5245 3. In the check_special_case() function in daemon.c I had used "errno" as an
5246 argument name, which causes warnings on some systems. This was basically a
5247 typo, since it was named "eno" in the comments!
5249 4. The code that waits for the clock to tick (at a resolution of some fraction
5250 of a second) so as to ensure message-id uniqueness was always waiting for
5251 at least one whole tick, when it could have waited for less. [This is
5252 almost certainly not relevant at current processor speeds, where it is
5253 unlikely to ever wait at all. But we try to future-proof.]
5255 5. The function that sleeps for a time interval that includes fractions of a
5256 second contained a race. It did not block SIGALRM between setting the
5257 timer, and suspending (a couple of lines later). If the interval was short
5258 and the sigsuspend() was delayed until after it had expired, the suspension
5259 never ended. On busy systems this could lead to processes getting stuck for
5262 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5263 process, before it forks any delivery processes. The open lookup caching
5264 mechanism meant that the open file or database connection was passed into
5265 the delivery process. The problem was that delivery processes always tidy
5266 up cached lookup data. This could cause a problem for the next delivery
5267 process started by the queue runner, because the external queue runner
5268 process does not know about the closure. So the next delivery process
5269 still has data in the lookup cache. In the case of a file lookup, there was
5270 no problem because closing a file descriptor in a subprocess doesn't affect
5271 the parent. However, if the lookup was caching a connection to a database,
5272 the connection was closed, and the second delivery process was likely to
5273 see errors such as "PGSQL: query failed: server closed the connection
5274 unexpectedly". The problem has been fixed by closing all cached lookups
5275 in a queue runner before running a delivery process.
5277 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5278 seem to have the "const" qualifier which it has on other OS. I've
5281 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5282 *of the same type* (client or server) with the same public name that an
5283 error should be diagnosed.
5285 9. When Exim looked up a host name for an IP address, but failed to find the
5286 original IP address when looking up the host name (a safety check), it
5287 output the message "<ip address> does not match any IP for NULL", which was
5288 confusing, to say the least. The bug was that the host name should have
5289 appeared instead of "NULL".
5291 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5292 user that is built into the binary, and the -C or -D options is used, root
5293 privilege is dropped before the configuration file is read. In addition,
5294 logging is switched to stderr instead of the normal log files. If the
5295 configuration then re-defines the Exim user, the unprivileged environment
5296 is probably not what is expected, so Exim logs a panic warning message (but
5299 However, if deliver_drop_privilege is set, the unprivileged state may well
5300 be exactly what is intended, so the warning has been cut out in that case,
5301 and Exim is allowed to try to write to its normal log files.
5307 1. smtp_return_error_details was not giving details for temporary sender
5308 or receiver verification errors.
5310 2. Diagnose a configuration error if two authenticators have the same public
5313 3. Exim used not to create the message log file for a message until the first
5314 delivery attempt. This could be confusing when incoming messages were held
5315 for policy or load reasons. The message log file is now created at the time
5316 the message is received, and an initial "Received" line is written to it.
5318 4. The automatically generated man page for command line options had a minor
5319 bug that caused no ill effects; however, a more serious problem was that
5320 the procedure for building the man page automatically didn't always
5321 operate. Consequently, release 4.20 contains an out-of-date version. This
5322 shouldn't happen again.
5324 5. When building Exim with embedded Perl support, the script that builds the
5325 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5326 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5328 6. The freeze_tell option was not being used for messages that were frozen on
5329 arrival, either by an ACL or by local_scan().
5331 7. Added the smtp_incomplete_transaction log selector.
5333 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5334 was accepting AUTH without a new EHLO.
5336 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5337 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5340 10. Logging of TCP/IP connections (when configured) now happens in the main
5341 daemon process instead of the child process, so that the TCP/IP connection
5342 count is more accurate (but it can never be perfect).
5344 11. The use of "drop" in a nested ACL was not being handled correctly in the
5345 outer ACL. Now, if condition failure induced by the nested "drop" causes
5346 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5347 or "require"), the connection is dropped.
5349 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5350 that yield "discard" can now be used with an "accept" or a "discard" verb,
5351 but an error is generated for any others (because I can't see a useful way
5352 to define what should happen).
5354 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5355 are now processed in the same way as lines in the Exim configuration file.
5356 In particular, continuation lines are supported.
5358 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5360 15. Added -ti meaning -t -i.
5362 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5363 domains, and warn by logging if others are found.
5365 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5366 returned from GIFCONF when Exim is trying to find the list of interfaces on
5367 a host. The code in os.c has been modified to copy each ifreq to an aligned
5368 structure in all cases.
5370 Also, in some cases, the returned ifreq's were being copied to a 'struct
5371 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5372 means the last couple of bytes of an IPv6 address could be chopped if the
5373 ifreq contained only a normal sockaddr (14 bytes storage).
5375 18. Named domain lists were not supported in the hosts_treat_as_local option.
5376 An entry such as +xxxx was not recognized, and was treated as a literal
5379 19. Ensure that header lines added by a DATA ACL are included in the reject log
5380 if the ACL subsequently rejects the message.
5382 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5383 MD5 (which is deprecated).
5385 21. When testing a filter file using -bf, Exim was writing a message when it
5386 took the sender from a "From " line in the message, but it was not doing so
5387 when it took $return_path from a Return-Path: header line. It now does.
5389 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5390 with a valid header line field name (a series of printing characters
5391 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5393 23. Changed "disc" in the source to "disk" to conform to the documentation and
5394 the book and for uniformity.
5396 24. Ignore Sendmail's -Ooption=value command line item.
5398 25. When execve() failed while trying to run a command in a pipe transport,
5399 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5400 could be confused with a return value of 69 from the command itself. This
5401 has been changed to 127, the value the shell returns if it is asked to run
5402 a non-existent command. The wording for the related log line suggests a
5403 non-existent command as the problem.
5405 26. If received_header_text expands to an empty string, do not add a Received:
5406 header line to the message. (Well, it adds a token one on the spool, but
5407 marks it "old" so that it doesn't get used or transmitted.)
5409 27. Installed eximstats 1.28 (addition of -nt option).
5411 28. There was no check for failure on the call to getsockname() in the daemon
5412 code. This can fail if there is a shortage of resources on the system, with
5413 ENOMEM, for example. A temporary error is now given on failure.
5415 29. Contrary to the C standard, it seems that in some environments, the
5416 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5417 program. Exim now does this explicitly; it affects the formatting of
5418 timestamps using strftime().
5420 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5421 complaints. I've now initialized all the variables, to avoid this.
5423 32. Header lines added by a system filter were not being "seen" during
5424 transport-time rewrites.
5426 33. The info_callback() function passed to OpenSSL is set up with type void
5427 (*)(SSL *, int, int), as described somewhere. However, when calling the
5428 function (actually a macro) that sets it up, the type void(*)() is
5429 expected. I've put in a cast to prevent warnings from picky compilers.
5431 34. If a DNS black list lookup found a CNAME record, but there were no A
5432 records associated with the domain it pointed at, Exim crashed.
5434 35. If a DNS black list lookup returned more than one A record, Exim ignored
5435 all but the first. It now scans all returned addresses if a particular IP
5436 value is being sought. In this situation, the contents of the
5437 $dnslist_value variable are a list of all the addresses, separated by a
5440 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5441 to accept a host name and all its aliases if the forward lookup for any of
5442 them yielded the IP address of the incoming connection. Now it accepts only
5443 those names whose forward lookup yields the correct IP address. Any other
5444 names are discarded. This closes a loophole whereby a rogue DNS
5445 administrator could create reverse DNS records to break through a
5446 wildcarded host restriction in an ACL.
5448 37. If a user filter or a system filter that ran in a subprocess used any of
5449 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5450 the wrong values were passed to the pipe command ($thisaddress had the
5451 value of $0, $0 had the value of $1, etc). This bug was introduced by
5452 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5454 38. Improved the line breaking for long SMTP error messages from ACLs.
5455 Previously, if there was no break point between 40 and 75 characters, Exim
5456 left the rest of the message alone. Two changes have been made: (a) I've
5457 reduced the minimum length to 35 characters; (b) if it can't find a break
5458 point between 35 and 75 characters, it looks ahead and uses the first one
5459 that it finds. This may give the occasional overlong line, but at least the
5460 remaining text gets split now.
5462 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5463 file descriptors might be low, and that setting 1000 would always raise it.
5464 It turns out that in some environments, the limit is already over 1000 and
5465 that lowering it causes trouble. So now Exim takes care not to decrease it.
5467 40. When delivering a message, the value of $return_path is set to $sender_
5468 address at the start of routing (routers may change the value). By an
5469 oversight, this default was not being set up when an address was tested by
5470 -bt or -bv, which affected the outcome if any router or filter referred to
5473 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5474 log only when "message" or "log_message" are set. However, if one of the
5475 conditions was an address verification, or a call to a nested ACL, the
5476 messages generated by the underlying test were being passed through. This
5477 no longer happens. The underlying message is available in $acl_verify_
5478 message for both "message" and "log_message" expansions, so it can be
5479 passed through if needed.
5481 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5482 new expansion $bh_ to give the encoded byte string without charset
5483 translation. Translation happens only if iconv() is available; HAVE_ICONV
5484 indicates this at build time. HEADERS_CHARSET gives the charset to
5485 translate to; headers_charset can change it in the configuration, and
5486 "headers charset" can change it in an individual filter file.
5488 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5489 that creates RFC 2047 encoded "words" labels them as that charset instead
5490 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5491 expansion operator; (ii) when Exim creates a From: line for a local
5492 message; (iii) when a header line is rewritten to include a "phrase" part.
5494 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5495 buggy, causing it to skip the first lines of messages whose message ID
5496 ended in 'D'. This would not have bitten before Exim release 4.14, because
5497 message IDs were unlikely to end in 'D' before then. The effect was to have
5498 incorrect size information for certain domains.
5500 45. #include "config.h" was missing at the start of the crypt16.c module. This
5501 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5504 46. If there was a timeout during a "random" callout check, Exim treated it as
5505 a failure of the random address, and carried on sending RSET and the real
5506 address. If the delay was just some slowness somewhere, the response to the
5507 original RCPT would be taken as a response to RSET and so on, causing
5508 mayhem of various kinds.
5510 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5511 when I implemented it. It didn't allow for the fact that some option values
5512 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5513 the right test for positive values.
5515 48. Domain names in DNS records are case-independent. Exim always looks them up
5516 in lower case. Some resolvers return domain names in exactly the case they
5517 appear in the zone file, that is, they may contain uppercase letters. Not
5518 all resolvers do this - some return always lower case. Exim was treating a
5519 change of case by a resolver as a change of domain, similar to a widening
5520 of a domain abbreviation. This triggered its re-routing code and so it was
5521 trying to route what was effectively the same domain again. This normally
5522 caused routing to fail (because the router wouldn't handle the domain
5523 twice). Now Exim checks for this case specially, and just changes the
5524 casing of the domain that it ultimately uses when it transmits the message
5527 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5530 50. If a filter generated a file delivery with a non-absolute name (possible if
5531 no home directory exists for the router), the forbid_file option was not
5534 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5535 the existing equality matching.
5537 52. Exim was using ints instead of ino_t variables in some places where it was
5538 dealing with inode numbers.
5540 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5541 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5542 called TMPDIR, and if it finds it is different, it changes its value.
5544 54. The smtp_printf() function is now made available to local_scan() so
5545 additional output lines can be written before returning. There is also an
5546 smtp_fflush() function to enable the detection of a dropped connection.
5547 The variables smtp_input and smtp_batched_input are exported to
5550 55. Changed the default runtime configuration: the message "Unknown user"
5551 has been removed from the ACL, and instead placed on the localuser router,
5552 using the cannot_route_message feature. This means that any verification
5553 failures that generate their own messages won't get overridden. Similarly,
5554 the "Unrouteable address" message that was in the ACL for unverifiable
5555 relay addresses has also been removed.
5557 56. Added hosts_avoid_esmtp to the smtp transport.
5559 57. The exicyclog script was not checking for the esoteric option
5560 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5561 will work only if exicyclog is run under the appropriate euid.
5563 58. Following a discussion on the list, the rules by which Exim recognises line
5564 endings on incoming messages have been changed. The -dropcr and drop_cr
5565 options are now no-ops, retained only for backwards compatibility. The
5566 following line terminators are recognized: LF CRLF CR. However, special
5567 processing applies to CR:
5569 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5570 nor a local message in the state where . is a terminator.
5572 (ii) If a bare CR is encountered in a header line, an extra space is added
5573 after the line terminator so as not to end the header. The reasoning
5574 behind this is that bare CRs in header lines are most likely either
5575 to be mistakes, or people trying to play silly games.
5577 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5578 was being incorrectly given as 18 bytes larger than it should have been.
5579 This is a VOB (very old bug).
5581 60. This may never have affected anything current, but just in case it has:
5582 When the local host is found other than at the start of a list of hosts,
5583 the local host, those with the same MX, and any that follow, are discarded.
5584 When the list in question was part of a longer list of hosts, the following
5585 hosts (not currently being processed) were also being discarded. This no
5586 longer happens. I'm not sure if this situation could ever has previously
5589 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5592 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5593 header. This is something that is recommended in a new Internet Draft, and
5594 is something that is documented as being done by Sendmail. There are two
5595 possible values. For messages generated by the autoreply transport, Exim
5598 Auto-Submitted: auto-replied
5600 whereas for all other generated messages (e.g. bounces) it adds
5602 Auto-Submitted: auto-generated
5604 63. The "personal" condition in filters now includes a test for the
5605 Auto-Submitted: header. If it contains the string "auto-" the message it
5606 not considered personal.
5608 64. Added rcpt_include_affixes as a generic transport option.
5610 65. Added queue_only_override (default true).
5612 66. Added the syslog_duplication option.
5614 67. If what should have been the first header line of a message consisted of
5615 a space followed by a colon, Exim was mis-interpreting it as a header line.
5616 It isn't of course - it is syntactically invalid and should therefore be
5617 treated as the start of the message body. The misbehaviour could have
5618 caused a number of strange effects, including loss of data in subsequent
5619 header lines, and spool format errors.
5621 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5622 client host had authenticated. This control can now be exercised by an ACL
5623 for more flexibility.
5625 69. By default, callouts do not happen when testing with -bh. There is now a
5626 variant, -bhc, which does actually run the callout code, including
5627 consulting and updating the callout cache.
5629 70. Added support for saslauthd authentication, courtesy of Alexander
5632 71. If statvfs() failed on the spool or log directories while checking their
5633 size for availability, Exim confusingly gave the error "space shortage".
5634 Furthermore, in debugging mode it crashed with a floating point exception.
5635 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5636 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5637 serious problem, Exim now writes to the main and panic logs when this
5638 happens, with details of the failure. It then refuses to accept the
5639 incoming message, giving the message "spool directory problem" or "log
5640 directory problem" with a 421 code for SMTP messages.
5642 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5643 0, 1, and 2 exist, because some OS complain for execs without them (see
5644 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5645 descriptors. However, the code omitted to check that the open succeeded,
5646 causing mysterious errors if for some reason the permissions on /dev/null
5647 got screwed. Now Exim writes a message to the main and panic logs, and
5648 bombs out if it can't open /dev/null.
5650 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5651 interact so that it is all more flexible. It is supposed to remain
5652 backwards compatible. Also added extra_local_interfaces.
5654 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5655 to bomb out with an assertion failure - to the client this appears as a
5656 connection drop. This problem occurs in the part of the code that was taken
5657 from the Samba project. Fortunately, the assertion is in a very simple
5658 function, so I have fixed this by reproducing the function inline in the
5659 one place where it is called, and arranging for authentication to fail
5660 instead of killing the process with assert().
5662 75. The SPA client code was not working when the server requested OEM rather
5663 than Unicode encoding.
5665 76. Added code to make require_files with a specific uid setting more usable in
5666 the case where statting the file as root fails - usually a non-root-mounted
5667 NFS file system. When this happens and the failure is EACCES, Exim now
5668 forks a subprocess and does the per-uid checking as the relevant uid.
5670 77. Added process_log_path.
5672 78. If log_file_path was not explicitly set, a setting of check_log_space or
5673 check_log_inodes was ignored.
5675 79. If a space check for the spool or log partitions fails, the incident is now
5676 logged. Of course, in the latter case the data may get lost...
5678 80. Added the %p formatting code to string_format() so that it can be used to
5679 print addresses in debug_print(). Adjusted all the address printing in the
5680 debugging in store.c to use %p rather than %d.
5682 81. There was a concern that a line of code in smtp_in.c could overflow a
5683 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5684 initially expressed, the concern was not well-founded, because trailing
5685 spaces are removed early. However, if the trailing spaces were followed by
5686 a NULL, they did not get removed, so the overflow was possible. Two fixes
5689 (a) I re-wrote the offending code in a cleaner fashion.
5690 (b) If an incoming SMTP command contains a NULL character, it is rejected
5693 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5694 runs initgroups(), so that if the Exim user is in any additional groups,
5695 they will be used during message reception.
5701 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob