1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
9 PP/02 Make -n do something, by making it not do something.
10 When combined with -bP, the name of an option is not output.
12 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
15 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
16 $sender_host_name and config options to manage this, and basic check
19 PP/05 DSCP support for outbound connections and control modifier for inbound.
21 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
22 (Only plugin which currently uses this is kerberos4, which nobody should
23 be using, but we should make it available and other future plugins might
24 conceivably use it, even though it would break NAT; stuff *should* be
25 using channel bindings instead).
27 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
28 name; added for Sendmail compatibility; requires admin caller.
29 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
30 just ignore it); requires trusted caller.
31 Also parse but ignore: -Ac -Am -X<logfile>
34 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
36 TL/02 Add +smtp_confirmation as a default logging option.
38 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
39 Patch by Magnus Holmgren from 2007-02-20.
41 TL/04 Bugzilla 1281 - Spec typo.
42 Bugzilla 1283 - Spec typo.
43 Bugzilla 1290 - Spec grammar fixes.
45 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
47 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
49 JH/02 Support "G" suffix to numbers in ${if comparisons.
51 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
53 NM/01 Bugzilla 1197 - Spec typo
54 Bugzilla 1196 - Spec examples corrections
56 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
58 PP/09 Add gnutls_enable_pkcs11 option.
60 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
61 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
63 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
64 "acl {{name}{arg}...}", and optional args on acl condition
67 JH/05 Permit multiple router/transport headers_add/remove lines.
69 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
71 JH/07 Avoid using a waiting database for a single-message-only transport.
72 Performance patch from Paul Fisher. Bugzilla 1262.
74 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
77 JH/09 Add $headers_added variable, with content from use of ACL modifier
78 add_header (but not yet added to the message). Bugzilla 199.
81 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
82 Pulled from Bugzilla 817 by Wolfgang Breyha.
89 PP/01 Handle short writes when writing local log-files.
90 In practice, only affects FreeBSD (8 onwards).
91 Bugzilla 1053, with thanks to Dmitry Isaikin.
93 NM/01 Bugzilla 949 - Documentation tweak
95 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
98 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
100 PP/02 Implemented gsasl authenticator.
102 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
104 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
105 `pkg-config foo` for cflags/libs.
107 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
108 with rest of GSASL and with heimdal_gssapi.
110 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
111 `pkg-config foo` for cflags/libs for the TLS implementation.
113 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
114 properties get this fed in as external SSF. A number of robustness
115 and debugging improvements to the cyrus_sasl authenticator.
117 PP/08 cyrus_sasl server now expands the server_realm option.
119 PP/09 Bugzilla 1214 - Log authentication information in reject log.
120 Patch by Jeremy Harris.
122 PP/10 Added dbmjz lookup type.
124 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
126 PP/12 MAIL args handles TAB as well as SP, for better interop with
127 non-compliant senders.
128 Analysis and variant patch by Todd Lyons.
130 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
131 Bug report from Lars Müller <lars@samba.org> (via SUSE),
132 Patch from Dirk Mueller <dmueller@suse.com>
134 PP/13 tls_peerdn now print-escaped for spool files.
135 Observed some $tls_peerdn in wild which contained \n, which resulted
136 in spool file corruption.
138 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
139 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
140 or write after TLS renegotiation, which otherwise led to messages
143 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
144 as a tracking header (ie: a signed header comes before the signature).
145 Patch from Wolfgang Breyha.
147 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
148 comma-sep list; embedded commas doubled.
150 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
152 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
154 Report and patch from Dmitry Banschikov.
156 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
157 Removed SSL_clear() after SSL_new() which led to protocol negotiation
158 failures. We appear to now support TLS1.1+ with Exim.
160 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
161 lets Exim select keys and certificates based upon TLS SNI from client.
162 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
163 before an outbound SMTP session. New log_selector, +tls_sni.
165 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
166 NULL dereference. Report and patch from Alun Jones.
168 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
169 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
170 Not seeing resolver debug output on NetBSD, but suspect this is a
171 resolver implementation change.
173 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
174 Left warnings. Added "eximon gdb" invocation mode.
176 PP/21 Defaulting "accept_8bitmime" to true, not false.
178 PP/22 Added -bw for inetd wait mode support.
180 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
181 locate the relevant includes and libraries. Made this the default.
183 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
184 Bugzilla 1246, report and most of solution from Tomasz Kusy.
186 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
187 This may cause build issues on older platforms.
189 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
190 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
191 gnutls_require_mac & gnutls_require_protocols (no longer supported).
192 Added SNI support via GnuTLS too.
193 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
195 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
197 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
198 Applied second patch from Janne, implementing suggestion to default
199 multiple-strings-in-record handling to match SPF spec.
201 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
203 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
204 read-only, out of scope).
205 Patch from Wolfgang Breyha, report from Stuart Northfield.
207 PP/29 Fix three issues highlighted by clang analyser static analysis.
208 Only crash-plausible issue would require the Cambridge-specific
209 iplookup router and a misconfiguration.
210 Report from Marcin Mirosław.
212 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
214 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
215 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
216 As part of this, removing so much warning spew let me fix some minor
217 real issues in debug logging.
219 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
220 assignment on my part. Fixed.
222 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
223 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
224 Janne Snabb (who went above and beyond: thank you).
226 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
227 string otherwise requires a connection and a bunch more work and it's
228 relatively easy to get wrong. Should also expose TLS library linkage
231 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
232 64-bit ${eval} (JH/03).
234 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
235 GNU libc to support some of the 64-bit stuff, should not lead to
236 conflicts. Defined before os.h is pulled in, so if a given platform
237 needs to override this, it can.
239 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
240 protection layer was required, which is not implemented.
241 Bugzilla 1254, patch from Wolfgang Breyha.
243 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
244 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
245 tls_dhparam take prime identifiers. Also unbreak combination of
246 OpenSSL+DH_params+TLSSNI.
248 PP/39 Disable SSLv2 by default in OpenSSL support.
254 PP/01 Solaris build fix for Oracle's LDAP libraries.
255 Bugzilla 1109, patch from Stephen Usher.
257 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
259 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
262 TF/02 Fix a couple more cases where we did not log the error message
263 when unlink() failed. See also change 4.74-TF/03.
265 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
266 lock up or crash if it happened to be inside a call to libc when it
267 got a SIGUSR1 from exiwhat.
269 The SIGUSR1 handler appends the current process status to the process
270 log which is later printed by exiwhat. It used to use the general
271 purpose logging code to do this, but several functions it calls are
272 not safe for signals.
274 The new output code in the SIGUSR1 handler is specific to the process
275 log, and simple enough that it's easy to inspect for signal safety.
276 Removing some special cases also simplifies the general logging code.
277 Removing the spurious timestamps from the process log simplifies
280 TF/04 Improved ratelimit ACL condition.
282 The /noupdate option has been deprecated in favour of /readonly which
283 has clearer semantics. The /leaky, /strict, and /readonly update modes
284 are mutually exclusive. The update mode is no longer included in the
285 database key; it just determines when the database is updated. (This
286 means that when you upgrde Exim will forget old rate measurements.)
288 Exim now checks that the per_* options are used with an update mode that
289 makes sense for the current ACL. For example, when Exim is processing a
290 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
291 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
292 must specify per_mail/readonly. If you omit the update mode it defaults to
293 /leaky where that makes sense (as before) or /readonly where required.
295 The /noupdate option is now undocumented but still supported for
296 backwards compatibility. It is equivalent to /readonly except that in
297 ACLs where /readonly is required you may specify /leaky/noupdate or
298 /strict/noupdate which are treated the same as /readonly.
300 A useful new feature is the /count= option. This is a generalization
301 of the per_byte option, so that you can measure the throughput of other
302 aggregate values. For example, the per_byte option is now equivalent
303 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
305 The per_rcpt option has been generalized using the /count= mechanism
306 (though it's more complicated than the per_byte equivalence). When it is
307 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
308 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
309 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
310 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
311 Note that using per_rcpt with a non-readonly update mode in more than
312 one ACL will cause the recipients to be double-counted. (The per_mail
313 and per_byte options don't have this problem.)
315 The handling of very low rates has changed slightly. If the computed rate
316 is less than the event's count (usually one) then this event is the first
317 after a long gap. In this case the rate is set to the same as this event's
318 count, so that the first message of a spam run is counted properly.
320 The major new feature is a mechanism for counting the rate of unique
321 events. The new per_addr option counts the number of different
322 recipients that someone has sent messages to in the last time period. It
323 behaves like per_rcpt if all the recipient addresses are different, but
324 duplicate recipient addresses do not increase the measured rate. Like
325 the /count= option this is a general mechanism, so the per_addr option
326 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
327 example, measure the rate that a client uses different sender addresses
328 with the options per_mail/unique=$sender_address. There are further
329 details in the main documentation.
331 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
333 TF/06 Removed a few PCRE remnants.
335 TF/07 Automatically extract Exim's version number from tags in the git
336 repository when doing development or release builds.
338 PP/02 Raise smtp_cmd_buffer_size to 16kB.
339 Bugzilla 879. Patch from Paul Fisher.
341 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
342 Heavily based on revision 40f9a89a from Simon Arlott's tree.
345 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
347 PP/05 Variable $av_failed, true if the AV scanner deferred.
348 Bugzilla 1078. Patch from John Horne.
350 PP/06 Stop make process more reliably on build failure.
351 Bugzilla 1087. Patch from Heiko Schlittermann.
353 PP/07 Make maildir_use_size_file an _expandable_ boolean.
354 Bugzilla 1089. Patch from Heiko Schlittermann.
356 PP/08 Handle ${run} returning more data than OS pipe buffer size.
357 Bugzilla 1131. Patch from Holger Weiß.
359 PP/09 Handle IPv6 addresses with SPF.
360 Bugzilla 860. Patch from Wolfgang Breyha.
362 PP/10 GnuTLS: support TLS 1.2 & 1.1.
364 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
367 PP/11 match_* no longer expand right-hand-side by default.
368 New compile-time build option, EXPAND_LISTMATCH_RHS.
369 New expansion conditions, "inlist", "inlisti".
371 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
373 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
375 PP/14 fix log_write() format string regression from TF/03.
376 Bugzilla 1152. Patch from Dmitry Isaikin.
382 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
384 PP/02 Harmonised TLS library version reporting; only show if debugging.
385 Layout now matches that introduced for other libraries in 4.74 PP/03.
387 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
389 PP/04 New "dns_use_edns0" global option.
391 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
394 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
395 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
397 TK/01 Updated PolarSSL code to 0.14.2.
398 Bugzilla 1097. Patch from Andreas Metzler.
400 PP/07 Catch divide-by-zero in ${eval:...}.
403 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
406 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
407 format-string attack -- SECURITY: remote arbitrary code execution.
409 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
410 time unintentionally subject to list matching rules, letting the header
411 cause arbitrary Exim lookups (of items which can occur in lists, *not*
412 arbitrary string expansion). This allowed for information disclosure.
414 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
415 INT_MIN/-1 -- value coerced to INT_MAX.
421 NM/01 Workround for PCRE version dependency in version reporting
424 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
425 This fixes portability to compilers other than gcc, notably
426 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
428 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
429 makefiles for portability to HP-UX and POSIX correctness.
431 PP/01 Permit LOOKUP_foo enabling on the make command-line.
432 Also via indented variable definition in the Makefile.
433 (Debugging by Oliver Heesakkers).
435 PP/02 Restore caching of spamd results with expanded spamd_address.
436 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
438 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
439 Improves build reliability. Fix from: Frank Elsner
441 NM/02 Fix wide character breakage in the rfc2047 coding
442 Fixes bug 1064. Patch from Andrey N. Oktyabrski
444 NM/03 Allow underscore in dnslist lookups
445 Fixes bug 1026. Patch from Graeme Fowler
447 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
448 Code patches from Adam Ciarcinski of NetBSD.
450 NM/04 Fixed exiqgrep to cope with mailq missing size issue
453 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
454 is logged, to avoid truncation. Patch from John Horne.
456 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
457 Patch from Jakob Hirsch.
459 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
460 SQL string expansion failure details.
461 Patch from Andrey Oktyabrski.
463 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
464 Patch from Simon Arlott.
466 PP/09 New lookups functionality failed to compile on old gcc which rejects
467 extern declarations in function scope.
468 Patch from Oliver Fleischmann
470 PP/10 Use sig_atomic_t for flags set from signal handlers.
471 Check getgroups() return and improve debugging.
472 Fixed developed for diagnosis in bug 927 (which turned out to be
475 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
476 Patch from Mark Zealey.
478 PP/12 Bugzilla 1056: Improved spamd server selection.
479 Patch from Mark Zealey.
481 PP/13 Bugzilla 1086: Deal with maildir quota file races.
482 Based on patch from Heiko Schlittermann.
484 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
485 Patch from Uwe Doering, sign-off by Michael Haardt.
487 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
488 variable declaration deep within a block. Bug and patch from
491 PP/15 lookups-Makefile IRIX compatibilty coercion.
493 PP/16 Make DISABLE_DKIM build knob functional.
495 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
496 Patch by Simon Arlott
498 TF/03 Fix valgrind.h portability to C89 compilers that do not support
499 variable argument macros. Our copy now differs from upstream.
505 TF/01 Failure to get a lock on a hints database can have serious
506 consequences so log it to the panic log.
508 TF/02 Log LMTP confirmation messages in the same way as SMTP,
509 controlled using the smtp_confirmation log selector.
511 TF/03 Include the error message when we fail to unlink a spool file.
513 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
514 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
515 for maintaining out-of-tree patches for some time.
517 PP/01 Bugzilla 139: Documentation and portability issues.
518 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
519 Handle per-OS dynamic-module compilation flags.
521 PP/02 Let /dev/null have normal permissions.
522 The 4.73 fixes were a little too stringent and complained about the
523 permissions on /dev/null. Exempt it from some checks.
524 Reported by Andreas M. Kirchwitz.
526 PP/03 Report version information for many libraries, including
527 Exim version information for dynamically loaded libraries. Created
528 version.h, now support a version extension string for distributors
529 who patch heavily. Dynamic module ABI change.
531 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
532 privilege escalation vulnerability whereby the Exim run-time user
533 can cause root to append content of the attacker's choosing to
536 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
539 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
540 If dropping privileges for untrusted macros, we disabled normal logging
541 on the basis that it would fail; for the Exim run-time user, this is not
542 the case, and it resulted in successful deliveries going unlogged.
543 Fixed. Reported by Andreas Metzler.
549 PP/01 Date: & Message-Id: revert to normally being appended to a message,
550 only prepend for the Resent-* case. Fixes regression introduced in
551 Exim 4.70 by NM/22 for Bugzilla 607.
553 PP/02 Include check_rfc2047_length in configure.default because we're seeing
554 increasing numbers of administrators be bitten by this.
556 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
558 PP/03 Bugzilla 994: added openssl_options main configuration option.
560 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
562 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
564 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
566 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
567 without a peer certificate, leading to a segfault because of an
568 assumption that peers always have certificates. Be a little more
569 paranoid. Problem reported by Martin Tscholak.
571 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
572 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
573 NB: ClamAV planning to remove STREAM in "middle of 2010".
574 CL also introduces -bmalware, various -d+acl logging additions and
575 more caution in buffer sizes.
577 PP/09 Implemented reverse_ip expansion operator.
579 PP/10 Bugzilla 937: provide a "debug" ACL control.
581 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
583 PP/12 Bugzilla 973: Implement --version.
585 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
587 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
589 PP/15 Bugzilla 816: support multiple condition rules on Routers.
591 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
592 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
593 ignore trailing whitespace.
595 JJ/02 prevent non-panic DKIM error from being sent to paniclog
597 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
600 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
601 Notification from Dr Andrew Aitchison.
603 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
604 ExtendedDetectionInfo response format.
605 Notification from John Horne.
607 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
610 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
611 XSL and documented dependency on system catalogs, with examples of how
614 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
617 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
618 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
619 configuration file which is writeable by the Exim user or group.
621 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
622 of configuration files to cover files specified with the -C option if
623 they are going to be used with root privileges, not just the default
626 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
627 option (effectively making it always true).
629 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
630 files to be used while preserving root privileges.
632 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
633 that rogue child processes cannot use them.
635 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
636 run-time user, instead of root.
638 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
639 Exim run-time user without dropping privileges.
641 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
642 result string, instead of calling string_vformat() twice with the same
645 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
646 for other users. Others should always drop root privileges if they use
647 -C on the command line, even for a whitelisted configure file.
649 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
651 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
657 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
658 $data_path, and $header_path variables; fixed documentation bugs and
661 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
662 exipick to access non-standard spools, including the "frozen" queue
665 NM/01 Bugzilla 965: Support mysql stored procedures.
666 Patch from Alain Williams
668 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
670 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
671 Patch from Andreas Metzler
673 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
674 Patch from Kirill Miazine
676 NM/05 Bugzilla 671: Added umask to procmail example.
678 JJ/03 installed exipick 20100323.0, fixing doc bug
680 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
681 directory. Notification and patch from Dan Rosenberg.
683 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
685 TK/02 Improve log output when DKIM signing operation fails.
687 MH/01 Treat the transport option dkim_domain as a colon separated
688 list, not as a single string, and sign the message with each element,
689 omitting multiple occurences of the same signer.
691 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
692 Bugzilla 985, 986. Patch by Simon Arlott
694 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
695 Patch by Simon Arlott
697 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
698 MBX locking. Notification from Dan Rosenberg.
704 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
706 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
708 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
710 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
712 NM/04 Bugzilla 915: Flag broken perl installation during build.
718 TK/01 Added patch by Johannes Berg that expands the main option
719 "spamd_address" if it starts with a dollar sign.
721 TK/02 Write list of recipients to X-Envelope-Sender header when building
722 the mbox-format spool file for content scanning (suggested by Jakob
725 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
726 (http://www.dcc-servers.net/) support via dccifd. Activated by
727 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
729 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
730 by Mark Daniel Reidel <mr@df.eu>.
732 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
733 When building exim an external PCRE library is now needed -
734 PCRE is a system library on the majority of modern systems.
735 See entry on PCRE_LIBS in EDITME file.
737 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
738 conversation. Added nologin parameter to request.
739 Patch contributed by Kirill Miazine.
741 TF/01 Do not log submission mode rewrites if they do not change the address.
743 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
745 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
746 log files in place. Contributed by Roberto Lima.
748 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
750 TF/03 Bugzilla 615: When checking the local_parts router precondition
751 after a local_part_suffix or local_part_prefix option, Exim now
752 does not use the address's named list lookup cache, since this
753 contains cached lookups for the whole local part.
755 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
756 Robert Millan. Documentation is in experimental-spec.txt.
758 TF/04 Bugzilla 668: Fix parallel build (make -j).
760 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
762 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
763 Patch provided by Jan Srzednicki.
765 TF/05 Leading white space used to be stripped from $spam_report which
766 wrecked the formatting. Now it is preserved.
768 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
769 that they are available at delivery time.
771 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
773 TF/08 TLS error reporting now respects the incoming_interface and
774 incoming_port log selectors.
776 TF/09 Produce a more useful error message if an SMTP transport's hosts
777 setting expands to an empty string.
779 NM/06 Bugzilla 744: EXPN did not work under TLS.
780 Patch provided by Phil Pennock.
782 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
783 Patch provided by Richard Godbee.
785 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
786 acl_smtp_notquit, added index entry.
788 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
789 Patch provided by Eugene Bujak.
791 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
792 accept(). Patch provided by Maxim Dounin.
794 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
795 Patch provided by Phil Pennock.
797 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
799 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
800 Patch provided by Brad "anomie" Jorsch.
802 NM/14 Bugzilla 622: Added timeout setting to transport filter.
803 Patch provided by Dean Brooks.
805 TK/05 Add native DKIM support (does not depend on external libraries).
807 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
808 Patch provided by Graeme Fowler.
810 NM/16 Bugzilla 851: Documentation example syntax fix.
812 NM/17 Changed NOTICE file to remove references to embedded PCRE.
814 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
817 NM/19 Bugzilla 745: TLS version reporting.
818 Patch provided by Phil Pennock.
820 NM/20 Bugzilla 167: bool: condition support.
821 Patch provided by Phil Pennock.
823 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
824 clients. Patch provided by Phil Pennock.
826 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
827 Patch provided by Brad "anomie" Jorsch.
829 NM/23 Bugzilla 687: Fix misparses in eximstats.
830 Patch provided by Heiko Schlittermann.
832 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
833 Patch provided by Heiko Schlittermann.
835 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
836 plus update to original patch.
838 NM/26 Bugzilla 799: Documentation correction for ratelimit.
840 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
841 Patch provided by David Brownlee.
843 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
845 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
847 NM/30 Bugzilla 888: TLS documentation bugfixes.
849 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
851 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
852 Unlike the original bugzilla I have changed all shell scripts in src tree.
854 NM/33 Bugzilla 898: Transport filter timeout fix.
855 Patch by Todd Rinaldo.
857 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
858 Patch by Serge Demonchaux.
860 NM/35 Bugzilla 39: Base64 decode bug fixes.
861 Patch by Jakob Hirsch.
863 NM/36 Bugzilla 909: Correct connect() call in dcc code.
865 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
867 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
869 NM/39 Bugzilla 911: Fixed MakeLinks build script.
875 TK/01 Add preliminary DKIM support. Currently requires a forked version of
876 ALT-N's libdkim that I have put here:
877 http://duncanthrax.net/exim-experimental/
879 Note to Michael Haardt: I had to rename some vars in sieve.c. They
880 were called 'true' and it seems that C99 defines that as a reserved
881 keyword to be used with 'bool' variable types. That means you could
882 not include C99-style headers which use bools without triggering
883 build errors in sieve.c.
885 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
886 as mailq or other aliases. Changed the --help handling significantly
887 to do whats expected. exim_usage() emits usage/help information.
889 SC/01 Added the -bylocaldomain option to eximstats.
891 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
893 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
895 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
901 PH/01 Another patch from the Sieve maintainer.
903 PH/02 When an IPv6 address is converted to a string for single-key lookup
904 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
905 dots are used instead of colons so that keys in lsearch files need not
906 contain colons. This was done some time before quoting was made available
907 in lsearch files. However, iplsearch files do require colons in IPv6 keys
908 (notated using the quote facility) so as to distinguish them from IPv4
909 keys. This meant that lookups for IP addresses in host lists did not work
910 for iplsearch lookups.
912 This has been fixed by arranging for IPv6 addresses to be expressed with
913 colons if the lookup type is iplsearch. This is not incompatible, because
914 previously such lookups could never work.
916 The situation is now rather anomolous, since one *can* have colons in
917 ordinary lsearch keys. However, making the change in all cases is
918 incompatible and would probably break a number of configurations.
920 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
923 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
924 conversion specification without a maximum field width, thereby enabling
925 a rogue spamd server to cause a buffer overflow. While nobody in their
926 right mind would setup Exim to query an untrusted spamd server, an
927 attacker that gains access to a server running spamd could potentially
928 exploit this vulnerability to run arbitrary code as the Exim user.
930 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
931 $primary_hostname instead of what libspf2 thinks the hosts name is.
933 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
934 a directory entry by the name of the lookup key. Previously, if a
935 symlink pointed to a non-existing file or a file in a directory that
936 Exim lacked permissions to read, a lookup for a key matching that
937 symlink would fail. Now it is enough that a matching directory entry
938 exists, symlink or not. (Bugzilla 503.)
940 PH/03 The body_linecount and body_zerocount variables are now exported in the
943 PH/04 Added the $dnslist_matched variable.
945 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
946 This means they are set thereafter only if the connection becomes
949 PH/06 Added the client_condition to authenticators so that some can be skipped
950 by clients under certain conditions.
952 PH/07 The error message for a badly-placed control=no_multiline_responses left
953 "_responses" off the end of the name.
955 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
957 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
958 (without spaces) instead of just copying the configuration text.
960 PH/10 Added the /noupdate option to the ratelimit ACL condition.
962 PH/11 Added $max_received_linelength.
964 PH/12 Added +ignore_defer and +include_defer to host lists.
966 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
967 way in which PCRE > 7.0 is built.
969 PH/14 Implemented queue_only_load_latch.
971 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
972 MAIL command. The effect was to mangle the value on 64-bit systems.
974 PH/16 Another patch from the Sieve maintainer.
976 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
978 PH/18 If a system quota error occurred while trying to create the file for
979 a maildir delivery, the message "Mailbox is full" was not appended to the
980 bounce if the delivery eventually timed out. Change 4.67/27 below applied
981 only to a quota excession during the actual writing of the file.
983 PH/19 It seems that peer DN values may contain newlines (and other non-printing
984 characters?) which causes problems in log lines. The DN values are now
985 passed through string_printing() before being added to log lines.
987 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
988 and InterBase are left for another time.)
990 PH/21 Added message_body_newlines option.
992 PH/22 Guard against possible overflow in moan_check_errorcopy().
994 PH/23 POSIX allows open() to be a macro; guard against that.
996 PH/24 If the recipient of an error message contained an @ in the local part
997 (suitably quoted, of course), incorrect values were put in $domain and
998 $local_part during the evaluation of errors_copy.
1004 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1005 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1008 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1009 issue a MAIL command.
1011 PH/02 In an ACL statement such as
1013 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1015 if a client was not listed at all, or was listed with a value other than
1016 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1017 the condition was not true (as it should be), so access was not denied.
1018 The bug was that the ! inversion was incorrectly passed on to the second
1019 item. This has been fixed.
1021 PH/03 Added additional dnslists conditions == and =& which are different from
1022 = and & when the dns lookup returns more than one IP address.
1024 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1025 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1027 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1028 FSYNC, which compiles an option called disable_fsync that allows for
1029 bypassing fsync(). The documentation is heavily laced with warnings.
1031 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1033 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1034 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1035 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1036 including adding "make clean"; (3) Added -fPIC when compiling the test
1037 dynamically loaded module, to get rid of a warning.
1039 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1040 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1041 = 0s. The bug is otherwise harmless.
1043 PH/07 There was a bug in the dovecot authenticator such that the value of
1044 $auth1 could be overwritten, and so not correctly preserved, after a
1045 successful authentication. This usually meant that the value preserved by
1046 the server_setid option was incorrect.
1048 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1050 PH/09 Installed PCRE release 7.0.
1052 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1053 run for batched SMTP input. It is now run at the start of every message
1054 in the batch. While fixing this I discovered that the process information
1055 (output by running exiwhat) was not always getting set for -bs and -bS
1056 input. This is fixed, and it now also says "batched" for BSMTP.
1058 PH/11 Added control=no_pipelining.
1060 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1061 patch, slightly modified), and move the expansion of helo_data till after
1062 the connection is made in the smtp transport (so it can use these
1065 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1067 PH/14 Added log_selector = +pid.
1069 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1071 PH/16 Add ${if forany and ${if forall.
1073 PH/17 Added dsn_from option to vary the From: line in DSNs.
1075 PH/18 Flush SMTP output before performing a callout, unless control =
1076 no_callout_flush is set.
1078 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1079 was true (the default) a successful delivery failed to delete the retry
1080 item, thus causing premature timeout of the address. The bug is now
1083 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1085 PH/21 Long custom messages for fakedefer and fakereject are now split up
1086 into multiline reponses in the same way that messages for "deny" and
1087 other ACL rejections are.
1089 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1090 with slight modification.
1092 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1093 draft, changing the syntax and factoring some duplicate code".
1095 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1096 for deliveries of the second and subsequent messages over the same SMTP
1099 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1100 ${reduce, with only minor "tidies".
1102 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1104 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1105 expansion side effects.
1107 PH/27 When a message times out after an over-quota error from an Exim-imposed
1108 quota, the bounce message says "mailbox is full". This message was not
1109 being given when it was a system quota that was exceeded. It now should
1112 MH/03 Made $recipients available in local_scan(). local_scan() already has
1113 better access to the recipient list through recipients_list[], but
1114 $recipients can be useful in postmaster-provided expansion strings.
1116 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1117 in the case of a MAIL command with additional options following the
1118 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1119 were accidentally chopped off.
1121 PH/29 SMTP synchronization checks are implemented when a command is read -
1122 there is a check that no more input is waiting when there shouldn't be
1123 any. However, for some commands, a delay in an ACL can mean that it is
1124 some time before the response is written. In this time, more input might
1125 arrive, invalidly. So now there are extra checks after an ACL has run for
1126 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1127 pipelining has not been advertised.
1129 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1131 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1132 returned in the SMTP response when smtp_return_error_details was false.
1133 This has been fixed.
1135 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1136 instead of the C I/O that was originally supplied, because problems were
1137 reported on Solaris.
1139 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1140 Exim which did not show up earlier: it was assuming that a call to
1141 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1142 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1143 was a macro that became an assignment, so it seemed to work. This has
1144 changed to a proper function call with a void return, hence the compile
1145 error. Exim's code has been fixed.
1147 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1150 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1152 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1154 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1155 because it turns out to be too big or there is a timeout) there is no
1156 "Completed" line in the log. When some messages of this type were
1157 selected by exigrep, they were listed as "not completed". Others were
1158 picked up by some special patterns. I have improved the selection
1159 criteria to be more general.
1161 PH/38 The host_find_failed option in the manualroute router can now be set
1162 to "ignore", to completely ignore a host whose IP address cannot be
1163 found. If all hosts are ignored, the behaviour is controlled by the new
1164 host_all_ignored option.
1166 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1167 homing or because of multiple MX records with /mx) generated more than
1168 one IP address, and the following item turned out to be the local host,
1169 all the secondary addresses of the first item were incorrectly removed
1170 from the list, along with the local host and any following hosts (which
1171 is what is supposed to happen).
1173 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1174 whoever called Exim into the -H file. In the case of the daemon it was
1175 behaving confusingly. When first started, it used values for whoever
1176 started the daemon, but after a SIGHUP it used the Exim user (because it
1177 calls itself on a restart). I have changed the code so that it now always
1180 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1181 message are rejected with the same error (e.g. no authentication or bad
1182 sender address), and a DATA command is nevertheless sent (as can happen
1183 with PIPELINING or a stupid MUA), the error message that was given to the
1184 RCPT commands is included in the rejection of the DATA command. This is
1185 intended to be helpful for MUAs that show only the final error to their
1188 PH/42 Another patch from the Sieve maintainer.
1190 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1191 Eximstats - Fixed some broken HTML links and added missing column headers
1193 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1194 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1196 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1198 PH/43 Yet another patch from the Sieve maintainer.
1200 PH/44 I found a way to check for a TCP/IP connection going away before sending
1201 the response to the final '.' that terminates a message, but only in the
1202 case where the client has not sent further data following the '.'
1203 (unfortunately, this is allowed). However, in many cases there won't be
1204 any further data because there won't be any more messages to send. A call
1205 to select() can be used: if it shows that the input is "ready", there is
1206 either input waiting, or the socket has been closed. An attempt to read
1207 the next input character can distinguish the two cases. Previously, Exim
1208 would have sent an OK response which the client would never have see.
1209 This could lead to message repetition. This fix should cure that, at
1210 least in a lot of common cases.
1212 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1213 advertised in response to EHLO.
1219 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1220 fixed by 4.65/MH/01 (is this a record?) are fixed:
1222 (i) An empty string was always treated as zero by the numeric comparison
1223 operators. This behaviour has been restored.
1225 (ii) It is documented that the numeric comparison operators always treat
1226 their arguments as decimal numbers. This was broken in that numbers
1227 starting with 0 were being interpreted as octal.
1229 While fixing these problems I realized that there was another issue that
1230 hadn't been noticed. Values of message_size_limit (both the global option
1231 and the transport option) were treated as octal if they started with 0.
1232 The documentation was vague. These values are now always treated as
1233 decimal, and I will make that clear in the documentation.
1239 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1240 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1243 MH/01 Don't check that the operands of numeric comparison operators are
1244 integers when their expansion is in "skipping" mode (fixes bug
1245 introduced by 4.64-PH/07).
1247 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1248 child addresses, Exim now panics and dies. Previously, because the count
1249 is held in a short int, deliveries were likely to be lost. As such a
1250 large number of recipients for a single message is ridiculous
1251 (performance will be very, very poor), I have chosen to impose a limit
1252 rather than extend the field.
1258 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1259 leftover -K file (the existence of which was triggered by #402).
1260 While we were at it, introduced process PID as part of the -K
1261 filename. This should rule out race conditions when creating
1264 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1265 processing considerably. Previous code took too long for large mails,
1266 triggering a timeout which in turn triggers #401.
1268 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1269 in the DK code in transports.c. sendfile() is not really portable,
1270 hence the _LINUX specificness.
1272 TF/01 In the add_headers option to the mail command in an Exim filter,
1273 there was a bug that Exim would claim a syntax error in any
1274 header after the first one which had an odd number of characters
1277 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1278 callout verification, Exim cached a "reject" for the entire domain. This
1279 is correct for most verifications, but it is not correct for a recipient
1280 verification with use_sender or use_postmaster set, because in that case
1281 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1282 case of MAIL FROM:<> rejection from other early rejections (e.g.
1283 rejection of HELO). When verifying a recipient using a non-null MAIL
1284 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1285 Whatever the result of the callout, the value of the domain cache is
1286 left unchanged (for any other kind of callout, getting as far as trying
1287 RCPT means that the domain itself is ok).
1289 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1292 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1293 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1294 session, the daemon ignores ECONNECT errors and logs others; it now
1295 ignores EPIPE as well.
1297 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1298 (quoted-printable decoding).
1300 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1301 later the small subsequent patch to fix an introduced bug.
1303 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1305 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1307 PH/08 An error is now given if message_size_limit is specified negative.
1309 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1310 to be given (somewhat) arbitrary names.
1312 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1315 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1316 miscellaneous code fixes
1318 PH/10 Added the log_reject_target ACL modifier to specify where to log
1321 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1322 hostname. This is wrong, because it relates to the incoming message (and
1323 probably the interface on which it is arriving) and not to the outgoing
1324 callout (which could be using a different interface). This has been
1325 changed to use the value of the helo_data option from the smtp transport
1326 instead - this is what is used when a message is actually being sent. If
1327 there is no remote transport (possible with a router that sets up host
1328 addresses), $smtp_active_hostname is used.
1330 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1331 tweaks were necessary in order to get it to work (see also 21 below):
1332 (a) The code assumed that strncpy() returns a negative number on buffer
1333 overflow, which isn't the case. Replaced with Exim's string_format()
1335 (b) There were several signed/unsigned issues. I just did the minimum
1336 hacking in of casts. There is scope for a larger refactoring.
1337 (c) The code used strcasecmp() which is not a standard C function.
1338 Replaced with Exim's strcmpic() function.
1339 (d) The code set only $1; it now sets $auth1 as well.
1340 (e) A simple test gave the error "authentication client didn't specify
1341 service in request". It would seem that Dovecot has changed its
1342 interface. Fortunately there's a specification; I followed it and
1343 changed what the client sends and it appears to be working now.
1345 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1348 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1349 address A is aliased to B and C, where B exists and C does not. Without
1350 -v the output is "A verified" because verification stops after a
1351 successful redirection if more than one address is generated. However,
1352 with -v the child addresses are also verified. Exim was outputting "A
1353 failed to verify" and then showing the successful verification for C,
1354 with its parentage. It now outputs "B failed to verify", showing B's
1355 parentage before showing the successful verification of C.
1357 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1358 look up a TXT record in a specific list after matching in a combined
1361 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1362 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1363 they consult the DNS. I had assumed they would set it the way they
1364 wanted; and indeed my experiments on Linux seem to show that in some
1365 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1366 To be on the safe side, however, I have now made the interface to
1367 host_find_byname() similar to host_find_bydns(), with an argument
1368 containing the DNS resolver options. The host_find_byname() function now
1369 sets these options at its start, just as host_find_bydns() does. The smtp
1370 transport options dns_qualify_single and dns_search_parents are passed to
1371 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1372 of host_find_byname() use the default settings of RES_DEFNAMES
1373 (qualify_single) but not RES_DNSRCH (search_parents).
1375 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1376 spool_read_header() do less string testing, by means of a preliminary
1377 switch on the second character of optional "-foo" lines. (This is
1378 overdue, caused by the large number of possibilities that now exist.
1379 Originally there were few.) While I was there, I also converted the
1380 str(n)cmp tests so they don't re-test the leading "-" and the first
1381 character, in the hope this might squeeze out yet more improvement.
1383 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1384 flag allowing group syntax was set by the header_syntax check but not
1385 turned off, possible causing trouble later; (2) The flag was not being
1386 set at all for the header_verify test, causing "group"-style headers to
1387 be rejected. I have now set it in this case, and also caused header_
1388 verify to ignore an empty address taken from a group. While doing this, I
1389 came across some other cases where the code for allowing group syntax
1390 while scanning a header line wasn't quite right (mostly, not resetting
1391 the flag correctly in the right place). These bugs could have caused
1392 trouble for malformed header lines. I hope it is now all correct.
1394 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1395 with the "reply" argument non-NULL. The code, however (which originally
1396 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1397 but it didn't always do it. This confused somebody who was copying the
1398 code for some other use. I have removed all the tests.
1400 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1401 feature that was used to support insecure browsers during the U.S. crypto
1402 embargo. It requires special client support, and Exim is probably the
1403 only MTA that supported it -- and would never use it because real RSA is
1404 always available. This code has been removed, because it had the bad
1405 effect of slowing Exim down by computing (never used) parameters for the
1406 RSA_EXPORT functionality.
1408 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1409 authenticator to fail if there's a tab character in the incoming data
1410 (there should never be unless someone is messing about, as it's supposed
1411 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1412 if the connection is using TLS or if the remote IP is the same as the
1413 local IP, and the "valid-client-cert option" if a client certificate has
1416 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1417 authenticators. This can be used for authorization after authentication
1418 succeeds. (In the case of plaintext, it servers for both authentication
1421 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1422 if any retry times were supplied.
1424 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1425 connection, where there is no client IP address to check. In this
1426 situation, the verify now always succeeds.
1428 PH/25 Applied John Jetmore's -Mset patch.
1430 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1432 PH/27 In a string expansion for a processed (not raw) header when multiple
1433 headers of the same name were present, leading whitespace was being
1434 removed from all of them, but trailing whitespace was being removed only
1435 from the last one. Now trailing whitespace is removed from each header
1436 before concatenation. Completely empty headers in a concatenation (as
1437 before) are ignored.
1439 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1440 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1442 PH/29 [Removed. This was a change that I later backed out, and forgot to
1443 correct the ChangeLog entry (that I had efficiently created) before
1444 committing the later change.]
1446 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1447 address errors (4xx response to RCPT) over the same connection as other
1448 messages routed to the same hosts. Such deliveries are always "forced",
1449 so retry times are not inspected. This resulted in far too many retries
1450 for the affected addresses. The effect occurred only when there were more
1451 hosts than the hosts_max_try setting in the smtp transport when it had
1452 the 4xx errors. Those hosts that it had tried were not added to the list
1453 of hosts for which the message was waiting, so if all were tried, there
1454 was no problem. Two fixes have been applied:
1456 (i) If there are any address or message errors in an SMTP delivery, none
1457 of the hosts (tried or untried) are now added to the list of hosts
1458 for which the message is waiting, so the message should not be a
1459 candidate for sending over the same connection that was used for a
1460 successful delivery of some other message. This seems entirely
1461 reasonable: after all the message is NOT "waiting for some host".
1462 This is so "obvious" that I'm not sure why it wasn't done
1463 previously. Hope I haven't missed anything, but it can't do any
1464 harm, as the worst effect is to miss an optimization.
1466 (ii) If, despite (i), such a delivery is accidentally attempted, the
1467 routing retry time is respected, so at least it doesn't keep
1468 hammering the server.
1470 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1471 in ${readsocket because some servers need this prod.
1473 PH/32 Added some extra debug output when updating a wait-xxx database.
1475 PH/33 The hint "could be header name not terminated by colon", which has been
1476 given for certain expansion errors for a long time, was not being given
1477 for the ${if def:h_colon_omitted{... case.
1479 PH/34 The spec says: "With one important exception, whenever a domain list is
1480 being scanned, $domain contains the subject domain." There was at least
1481 one case where this was not true.
1483 PH/35 The error "getsockname() failed: connection reset by peer" was being
1484 written to the panic log as well as the main log, but it isn't really
1485 panic-worthy as it just means the connection died rather early on. I have
1486 removed the panic log writing for the ECONNRESET error when getsockname()
1489 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1490 runs only) independently of the message's sender address. This meant
1491 that, if the 4xx error was in fact related to the sender, a different
1492 message to the same recipient with a different sender could confuse
1493 things. In particualar, this can happen when sending to a greylisting
1494 server, but other circumstances could also provoke similar problems.
1495 I have changed the default so that the retry time for these errors is now
1496 based a combination of the sender and recipient addresses. This change
1497 can be overridden by setting address_retry_include_sender=false in the
1500 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1501 remote server are returned as part of bounce messages. This was not
1502 happening for LMTP over a pipe (the lmtp transport), but now it is the
1503 same for both kinds of LMTP.
1505 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1506 in header lines that were in fact CNAMEs. This is no longer the case.
1508 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1509 and queue runs started by the daemon processed all messages. This has
1510 been fixed so that -R and -S can now usefully be given with -q<time>.
1512 PH/40 Import PCRE release 6.7 (fixes some bugs).
1514 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1516 PH/42 Give an error if -q is specified more than once.
1518 PH/43 Renamed the variables $interface_address and $interface_port as
1519 $received_ip_address and $received_port, to make it clear that these
1520 values apply to message reception, and not to the outgoing interface when
1521 a message is delivered. (The old names remain recognized, of course.)
1523 PH/44 There was no timeout on the connect() call when using a Unix domain
1524 socket in the ${readsocket expansion. There now is.
1526 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1527 be meaningful with "accept".
1529 SC/01 Eximstats V1.43
1530 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1532 SC/02 Eximstats V1.44
1533 Use a glob alias rather than an array ref in the generated
1534 parser. This improves both readability and performance.
1536 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1537 Collect SpamAssassin and rejection statistics.
1538 Don't display local sender or destination tables unless
1539 there is data to show.
1540 Added average volumes into the top table text output.
1542 SC/04 Eximstats V1.46
1543 Collect data on the number of addresses (recipients)
1544 as well as the number of messages.
1546 SC/05 Eximstats V1.47
1547 Added 'Message too big' to the list of mail rejection
1548 reasons (thanks to Marco Gaiarin).
1550 SC/06 Eximstats V1.48
1551 Mainlog lines which have GMT offsets and are too short to
1552 have a flag are now skipped.
1554 SC/07 Eximstats V1.49 (Alain Williams)
1555 Added the -emptyok flag.
1557 SC/08 Eximstats V1.50
1558 Fixes for obtaining the IP address from reject messages.
1560 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1561 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1562 whitesspace changes from 4.64-PH/27
1564 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1567 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1568 are found, allow negative numbers in numeric criteria)
1570 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1572 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1573 to match changes made in 4.64-PH/43
1575 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1577 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1578 "domain" is actually an IP address, and force a failure. This locks out
1579 those revolvers/nameservers that support "A-for-A" lookups, in
1580 contravention of the specifications.
1582 PH/48 When a host name was looked up from an IP address, and the subsequent
1583 forward lookup of the name timed out, the host name was left in
1584 $sender_host_name, contrary to the specification.
1586 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1587 restricted to single-key lookups, Exim was not diagnosing an error if
1588 * or *@ was used with a query-style lookup.
1590 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1592 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1593 long ago, but noone interested enough thought of it. Let's just say that
1594 the "1.1" means that there are some new functions that weren't there at
1595 some point in the past.
1597 PH/51 Error processing for expansion failure of helo_data from an smtp
1598 transport during callout processing was broken.
1600 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1601 tested/used via the -bh/-bhc/-bs options.
1603 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1604 bug, fixed in subsequent PCRE releases).
1606 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1607 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1609 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1615 SC/01 Use a glob alias rather than an array ref in eximstats generated
1616 parser. This improves both readability and performance.
1618 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1619 Don't display local sender or destination tables in eximstats unless
1620 there is data to show.
1621 Added average volumes into the eximstats top table text output.
1623 SC/03 Collect data on the number of addresses (recipients) as well
1624 as the number of messages in eximstats.
1626 TF/01 Correct an error in the documentation for the redirect router. Exim
1627 does (usually) call initgroups() when daemonizing.
1629 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1630 with consistent privilege compared to when running as a daemon.
1632 TF/03 Note in the spec that $authenticated_id is not set for local
1633 submissions from trusted users.
1635 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1636 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1638 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1639 by adding some example configuration directives to the default
1640 configuration file. A little bit of work is required to uncomment the
1641 directives and define how usernames and passwords are checked, but
1642 there is now a framework to start from.
1644 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1645 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1646 without this. I don't know how relevant this is to other LDAP libraries.
1648 PH/02 Add the verb name to the "unknown ACL verb" error.
1650 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1652 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1654 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1655 directory not expanded when it should be if an expanded home directory
1656 was set for the address (which is overridden by the transport).
1658 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1661 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1662 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1663 because it is too late at that time, and has no effect.
1665 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1666 security issue with \' (bugzilla #107). I could not use the
1667 PQescapeStringConn() function, because it needs a PGconn value as one of
1670 PH/08 When testing addresses using -bt, indicate those final addresses that
1671 are duplicates that would not cause an additional delivery. At least one
1672 person was confused, thinking that -bt output corresponded to deliveries.
1673 (Suppressing duplicates isn't a good idea as you lose the information
1674 about possibly different redirections that led to the duplicates.)
1676 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1677 systems where poll() doesn't work, in particular OS X.
1679 PH/10 Added more information to debugging output for retry time not reached.
1681 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1682 operations in malware.c.
1684 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1687 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1688 syslog_duplication set false, log lines that would normally be written
1689 both the the main log and to the reject log were not written to syslog at
1692 PH/14 In the default configuration, change the use of "message" in ACL warn
1693 statements to "add_header".
1695 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1696 not followed by a command (e.g. "seen endif").
1698 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1699 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1702 PH/17 Added extra conditions to the default value of delay_warning_condition
1706 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1707 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1708 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1711 The Auto-Submitted: and various List- headers are standardised, whereas I
1712 don't think Precedence: ever was.
1714 PH/18 Refactored debugging code in route_finduser() to show more information,
1715 in particular, the error code if getpwnam() issues one.
1717 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1718 This is apparently needed in addition to the PH/07 change above to avoid
1719 any possible encoding problems.
1721 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1722 but not after initializing Perl.
1724 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1725 output them only if debugging. By default they are written stderr,
1726 apparently, which is not desirable.
1728 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1731 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1734 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1736 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1737 authenticated or an ident call has been made. Suppress the default
1738 values for $authenticated_id and $authenticated_sender (but permit -oMai
1739 and -oMas) when testing with -bh.
1741 PH/24 Re-jigged the order of the tests in the default configuration so that the
1742 tests for valid domains and recipients precede the DNS black list and CSA
1743 tests, on the grounds that those ones are more expensive.
1745 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1746 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1747 This bug exists in every version of Exim that I still have, right back to
1750 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1751 However, an attempt to turn on case-sensitivity in a regex key by
1752 including (?-i) didn't work because the subject string was already
1753 lowercased, and the effects were non-intuitive. It turns out that a
1754 one-line patch can be used to allow (?-i) to work as expected.
1760 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1761 other effects) broke the use of negated acl sub-conditions.
1763 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1766 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1767 "Deny" causes Exim to reject the incoming connection with a 554 error.
1768 Unfortunately, if there is a major crisis, such as a disk failure,
1769 tcp-wrappers gives "deny", whereas what one would like would be some
1770 kind of temporary error. A kludge has been added to help with this.
1771 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1772 554 error is used if errno is still zero or contains ENOENT (which occurs
1773 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1776 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1778 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1779 errors. Otherwise a message that provokes a temporary error (when other
1780 messages do not) can cause a whole host to time out.
1782 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1783 addresses were routed directly to files or pipes from a redirect router.
1784 File deliveries just didn't batch; pipe deliveries might have suffered
1787 PH/06 A failure to get a lock for a hints database would erroneously always say
1788 "Failed to get write lock", even when it was really a read lock.
1790 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1791 of 0600. This has been changed to use the value of the lockfile_mode
1792 option (which defaults to 0600).
1794 PH/08 Applied small patch from the Sieve maintainer.
1796 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1797 folder from quota calculations, a direct delivery into this folder messed
1798 up the contents of the maildirsize file. This was because the regex was
1799 used only to exclude .Trash (or whatever) when the size of the mailbox
1800 was calculated. There was no check that a delivery was happening into an
1801 excluded directory. This bug has been fixed by ignoring all quota
1802 processing for deliveries into excluded directories.
1804 PH/10 Added the maildirfolder_create_regex option to appendfile.
1810 PH/01 The code for finding all the local interface addresses on a FreeBSD
1811 system running IPv6 was broken. This may well have applied to all BSD
1812 systems, as well as to others that have similar system calls. The broken
1813 code found IPv4 interfaces correctly, but gave incorrect values for the
1814 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1815 that it would not match correctly against @[] and not recognize the IPv6
1818 PH/02 The ipliteral router was not recognizing addresses of the form user@
1819 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1821 PH/03 Added disable_ipv6.
1823 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1824 decoded form, because it is most often used to construct To: headers
1825 lines in autoreplies, and the decoded form may well be syntactically
1826 invalid. However, $reply_address has leading white space removed, and all
1827 newlines turned into spaces so that the autoreply transport does not
1830 PH/05 If group was specified without a user on a router, and no group or user
1831 was specified on a transport, the group from the router was ignored.
1833 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1834 for visible compile-time settings that can be used to change these
1835 numbers, for those that want even more. Backwards compatibility with old
1836 spool files has been maintained. However, going back to a previous Exim
1837 release will lost any variables that are in spool files.
1839 PH/07 Two small changes when running in the test harness: increase delay when
1840 passing a TCP/IP connection to a new process, in case the original
1841 process has to generate a bounce, and remove special handling of
1842 127.0.0.2 (sic), which is no longer necessary.
1844 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1845 be the same on different OS.
1847 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1850 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1851 whether --show-vars was specified or not
1853 JJ/02 exipick: Added support for new ACL variable spool format introduced
1856 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1857 syntactically invalid From: or Reply-to: line, and a filter used this to
1858 generate an autoreply, and therefore failed to obtain an address for the
1859 autoreply, Exim could try to deliver to a non-existent relative file
1860 name, causing unrelated and misleading errors. What now happens is that
1861 it logs this as a hard delivery error, but does not attempt to create a
1864 PH/11 The exinext utility has a -C option for testing purposes, but although
1865 the given file was scanned by exinext itself; it wasn't being passed on
1866 when Exim was called.
1868 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1869 an end-of-file indication when reading a command response.
1871 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1872 compiled. In many other places in Exim, IPv6 addresses are always
1873 recognized, so I have changed this. It also means that IPv4 domain
1874 literals of the form [IPV4:n.n.n.n] are now always recognized.
1876 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1877 used if the router is not running as root, for example, when verifying at
1878 ACL time, or when using -bh. The debugging output from this situation was
1879 non-existent - all you got was a failure to exec. I have made two
1882 (a) Failures to set uid/gid, the current directory, or a process leader
1883 in a subprocess such as that created by queryprogram now generate
1884 suitable debugging ouput when -d is set.
1886 (b) The queryprogram router detects when it is not running as root,
1887 outputs suitable debugging information if -d is set, and then runs
1888 the subprocess without attempting to change uid/gid.
1890 PH/15 Minor change to Makefile for building test_host (undocumented testing
1893 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1894 additional section of a DNS packet that returns MX or SRV records.
1895 Instead, it always explicitly searches for A/AAAA records. This avoids
1896 major problems that occur when a DNS server includes only records of one
1897 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1898 fixed another bug: if SRV records were looked up and the corresponding
1899 address records were *not* found in the additional section, the port
1900 values from the SRV records were lost.
1902 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1903 using the correct key (the original address) when searching the retry
1904 rules in order to find which one to use for generating the retry hint.
1906 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1907 adding the default one. Similarly, if it contains a Reply-To: header, the
1908 errors_reply_to option, if set, is not used.
1910 PH/19 When calculating a retry time, Exim used to measure the "time since
1911 failure" by looking at the "first failed" field in the retry record. Now
1912 it does not use this if it is later than than the arrival time of the
1913 message. Instead it uses the arrival time. This makes for better
1914 behaviour in cases where some deliveries succeed, thus re-setting the
1915 "first failed" field. An example is a quota failure for a huge message
1916 when small messages continue to be delivered. Without this change, the
1917 "time since failure" will always be short, possible causing more frequent
1918 delivery attempts for the huge message than are intended.
1919 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1921 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1922 $1, $2, $3) because the numerical variables can be reset during some
1923 expansion items (e.g. "match"), thereby losing the authentication data.
1925 PH/21 Make -bV show the size of off_t variables so that the test suite can
1926 decide whether to run tests for quotas > 2G.
1928 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1929 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1930 filecount value is greater than 2G or if a quota value is greater than 2G
1931 on a system where the size of off_t is not greater than 4, a panic error
1934 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1935 never match. The debug and -bh output now contains an explicit error
1936 message indicating a malformed IPv4 address or mask.
1938 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1939 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1940 PH/23 above applies.
1942 PH/25 Do not write to syslog when running in the test harness. The only
1943 occasion when this arises is a failure to open the main or panic logs
1944 (for which there is an explicit test).
1946 PH/26 Added the /no_tell option to "control=freeze".
1948 PH/27 If a host name lookup failed very early in a connection, for example, if
1949 the IP address matched host_lookup and the reverse lookup yielded a name
1950 that did not have a forward lookup, an error message of the form "no IP
1951 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1952 could be logged. Now it outputs the IP address instead of "NULL".
1954 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1955 allows the sender and the authenticated sender to be set when
1956 submitting a message from within Exim. Since child_open_exim() is
1957 documented for local_scan(), the new function should be too.
1959 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1960 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1961 results in an empty string is now treated as unset.
1963 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1965 PH/31 Added sender_verify_fail logging option.
1967 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1968 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1969 tidied the source and removed it altogether.
1971 PH/33 When a queue run was abandoned because the load average was too high, a
1972 log line was always written; now it is written only if the queue_run log
1973 selector is set. In addition, the log line for abandonment now contains
1974 information about the queue run such as the pid. This is always present
1975 in "start" and "stop" lines but was omitted from the "abandon" line.
1977 PH/34 Omit spaces between a header name and the colon in the error message that
1978 is given when verify = headers_syntax fails (if there are lots of them,
1979 the message gets confusing).
1981 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1982 names, as there are now some PTR records that contain slashes. This check
1983 is only to protect against broken name servers that fall over on strange
1984 characters, so the fact that it applies to all lookups doesn't matter.
1986 PH/36 Now that the new test suite is complete, we can remove some of the
1987 special code in Exim that was needed for the old test suite. For example,
1988 sorting DNS records because real resolvers return them in an arbitrary
1989 order. The new test suite's fake resolver always returns records in the
1992 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1993 bounces) except when queue_only is set, to avoid logging races between
1994 the different processes.
1996 PH/38 Panic-die if .include specifies a non-absolute path.
1998 PH/39 A tweak to the "H" retry rule from its user.
2000 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2001 a label. They prevented compilation on older perls.
2003 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2004 a warning to be raised on newish perls.
2006 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2007 on queue. Changes to match documented behaviour of showing count of
2008 messages matching specified criteria.
2010 PH/40 Changed the default ident timeout from 30s to 5s.
2012 PH/41 Added support for the use of login_cap features, on those BSD systems
2013 that have them, for controlling the resources used by pipe deliveries.
2015 PH/42 The content-scanning code uses fopen() to create files in which to put
2016 message data. Previously it was not paying any attention to the mode of
2017 the files. Exim runs with umask(0) because the rest of the code creates
2018 files with open(), and sets the required mode explicitly. Thus, these
2019 files were ending up world-writeable. This was not a big issue, because,
2020 being within the spool directory, they were not world-accessible. I have
2021 created a function called modefopen, which takes an additional mode
2022 argument. It sets umask(777), creates the file, chmods it to the required
2023 mode, then resets the umask. All the relevant calls to fopen() in the
2024 content scanning code have been changed to use this function.
2026 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2027 to 24 hours. This avoids potential overflow problems when processing G
2028 and H retry rules. I suspect nobody ever tinkers with this value.
2030 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2032 PH/45 When the plaintext authenticator is running as a client, the server's
2033 challenges are checked to ensure they are valid base64 strings. By
2034 default, the authentication attempt is cancelled if an invalid string is
2035 received. Setting client_ignore_invalid_base64 true ignores these errors.
2036 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2037 they are received. Thus, the responses can be made to depend on the
2038 challenges. If an invalid string is ignored, an empty string is placed in
2041 PH/46 Messages that are created by the autoreply transport now contains a
2042 References: header, in accordance with RFCs 2822 and 3834.
2044 PH/47 Added authenticated_sender_force to the smtp transport.
2046 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2048 PH/49 Installed latest patch from the Sieve maintainer.
2050 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2051 was also set, the appendfile transport was unnecessarily scanning a
2052 directory of message files (e.g. for maildir delivery) to find the count
2053 of files (along with the size), even though it did not need this
2054 information. It now does the scan only if it needs to find either the
2055 size of the count of files.
2057 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2059 PH/52 Two bugs concerned with error handling when the smtp transport is
2062 (i) Exim was not creating retry information for temporary errors given
2063 for individual recipients after the DATA command when the smtp transport
2064 was used in LMTP mode. This meant that they could be retried too
2065 frequently, and not timed out correctly.
2067 (ii) Exim was setting the flag that allows error details to be returned
2068 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2069 recipients that were returned after the DATA command.
2071 PH/53 This is related to PH/52, but is more general: for any failing address,
2072 when detailed error information was permitted to be returned to the
2073 sender, but the error was temporary, then after the final timeout, only
2074 "retry timeout exceeded" was returned. Now it returns the full error as
2075 well as "retry timeout exceeded".
2077 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2078 do this, and (what is worse) MTAs that accept it.
2080 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2081 will now be deprecated.
2083 PH/56 New os.c-cygwin from the Cygwin maintainer.
2085 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2086 formats (previously only available in exim formats via -bpr, -bpru,
2087 and -bpra. Now also available in native and exiqgrep formats)
2089 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2090 with very large, slow to parse queues
2092 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2094 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2096 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2097 responses to authentication challenges, though it was showing the
2098 challenges; (ii) I've removed the CR characters from the debug output for
2101 PH/58 Allow for the insertion of a newline as well as a space when a string
2102 is turned into more than one encoded-word during RFC 2047 encoding. The
2103 Sieve code now uses this.
2105 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2106 data_4xx, lost_connection, tls_required.
2108 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2109 message was being sent as an SMTP response.
2111 PH/61 Add -l and -k options to exicyclog.
2113 PH/62 When verifying, if an address was redirected to one new address, so that
2114 verification continued, and the new address failed or deferred after
2115 having set something in $address_data, the value of $address_data was not
2116 passed back to the ACL. This was different to the case when no
2117 redirection occurred. The value is now passed back in both cases.
2119 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2120 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2121 use login_cap.h, so on its own it isn't the distinguishing feature. The
2122 new name refers directly to the setclassresources() function.
2124 PH/65 Added configuration files for NetBSD3.
2126 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2128 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2129 is preferred over IPv4.
2131 PH/68 The bounce_return_message and bounce_return_body options were not being
2132 honoured for bounces generated during the reception of non-SMTP messages.
2133 In particular, this applied to messages rejected by the ACL. This bug has
2134 been fixed. However, if bounce_return_message is true and bounce_return_
2135 body is false, the headers that are returned for a non-SMTP message
2136 include only those that have been read before the error was detected.
2137 (In the case of an ACL rejection, they have all been read.)
2139 PH/69 The HTML version of the specification is now built in a directory called
2140 spec_html instead of spec.html, because the latter looks like a path with
2141 a MIME-type, and this confuses some software.
2143 PH/70 Catch two compiler warnings in sieve.c.
2145 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2146 function verify_get_ident() calls ip_connect() to connect a socket, but
2147 if the "connect()" function timed out, ip_connect() used to close the
2148 socket. However, verify_get_ident() also closes the socket later, and in
2149 between Exim writes to the log, which may get opened at this point. When
2150 the socket was closed in ip_connect(), the log could get the same file
2151 descriptor number as the socket. This naturally causes chaos. The fix is
2152 not to close the socket in ip_connect(); the socket should be closed by
2153 the function that creates it. There was only one place in the code where
2154 this was missing, in the iplookup router, which I don't think anybody now
2155 uses, but I've fixed it anyway.
2157 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2158 well as to direct DNS lookups. Otherwise the handling of names in host
2159 lists is inconsistent and therefore confusing.
2165 PH/01 Two changes to the default runtime configuration:
2167 (1) Move the checks for relay_from_hosts and authenticated clients from
2168 after to before the (commented out) DNS black list checks.
2170 (2) Add control=submission to the relay_from_hosts and authenticated
2171 clients checks, on the grounds that messages accepted by these
2172 statements are most likely to be submissions.
2174 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2176 (1) Generate an error if the third argument for the ${prvs expansion is
2179 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2182 (3) Reset the variables that are obtained from the first argument of
2183 ${prvscheck and used in the second argument before leaving the code,
2184 because their memory is reclaimed, so using them afterwards may do
2187 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2188 one (it's much easier than Tom thought :-).
2190 (5) Because of (4), we can now allow for the use of $prvscheck_result
2191 inside the third argument.
2193 PH/03 For some reason, the default setting of PATH when running a command from
2194 a pipe transport was just "/usr/bin". I have changed it to
2197 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2198 anything to be listed in the output from -bV.
2200 PH/05 When a filter generated an autoreply, the entire To: header line was
2201 quoted in the delivery log line, like this:
2203 => >A.N.Other <ano@some.domain> <original@ddress> ...
2205 This has been changed so that it extracts the operative address. There
2206 may be more than one such address. If so, they are comma-separated, like
2209 => >ano@some.domain,ona@other.domain <original@ddress> ...
2211 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2212 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2213 not being looked up in the rDNS to get a host name, Exim was showing the
2214 IP address twice in Received: lines, even though the IP addresses were
2215 identical. For example:
2217 Received: from [1.2.3.4] (helo=[1.2.3.4])
2219 However, if the real host name was known, it was omitting the HELO data
2220 if it matched the actual IP address. This has been tidied up so that it
2221 doesn't show the same IP address twice.
2223 PH/07 When both +timestamp and +memory debugging was on, the value given by
2224 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2225 called by the debug printing, thereby overwriting the timestamp buffer.
2226 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2228 PH/08 When the original message was included in an autoreply transport, it
2229 always said "this is a copy of the message, including all the headers",
2230 even if body_only or headers_only was set. It now gives an appropriate
2233 PH/09 Applied a patch from the Sieve maintainer which:
2235 o fixes some comments
2236 o adds the (disabled) notify extension core
2237 o adds some debug output for the result of if/elsif tests
2238 o points to the current vacation draft in the documentation
2239 and documents the missing references header update
2243 o fixes a bug in processing the envelope test (when testing
2244 multiple envelope elements, the last element determinted the
2247 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2248 Electronic Mail") by including:
2250 Auto-submitted: auto-generated
2252 in the messages that it generates (bounce messages and others, such as
2253 warnings). In the case of bounce messages for non-SMTP mesages, there was
2254 also a typo: it was using "Auto_submitted" (underscore instead of
2255 hyphen). Since every message generated by Exim is necessarily in response
2256 to another message, thes have all been changed to:
2258 Auto-Submitted: auto-replied
2260 in accordance with these statements in the RFC:
2262 The auto-replied keyword:
2264 - SHOULD be used on messages sent in direct response to another
2265 message by an automatic process,
2267 - MUST NOT be used on manually-generated messages,
2269 - MAY be used on Delivery Status Notifications (DSNs) and Message
2270 Disposition Notifications (MDNs),
2272 - MUST NOT be used on messages generated by automatic or periodic
2273 processes, except for messages which are automatic responses to
2276 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2277 to the default Received: header definition.
2279 PH/12 Added log selector acl_warn_skipped (default on).
2281 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2282 variables because (a) they are in the wrong storage pool and (b) even if
2283 they were copied, it wouldn't work properly because of the caching.
2285 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2286 checking when decoding. Apparently there are clients that generate
2287 overlong encoded strings. Why am I not surprised?
2289 PH/15 If the first argument of "${if match_address" was not empty, but did not
2290 contain an "@" character, Exim crashed. Now it writes a panic log message
2291 and treats the condition as false.
2293 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2295 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2296 extension "envelope-auth". The code is finished and in agreement with
2297 other implementations, but there is no documentation so far and in fact,
2298 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2299 not changing the active code.
2301 Print executed "if" and "elsif" statements when debugging is used. This
2302 helps a great deal to understand what a filter does.
2304 Document more things not specified clearly in RFC3028. I had all this
2305 sorted out, when out of a sudden new issues came to my mind. Oops."
2307 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2310 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2311 canonical form (as documented). However, after a host name lookup from
2312 the IP address, check_host() was doing a simple string comparison with
2313 addresses acquired from the DNS when checking that the found name did
2314 have the original IP as one of its addresses. Since any found IPv6
2315 addresses are likely to be in abbreviated form, the comparison could
2316 fail. Luckily, there already exists a function for doing the comparison
2317 by converting both addresses to binary, so now that is used instead of
2318 the text comparison.
2320 PH/20 There was another similar case to PH/19, when a complete host name was
2321 given in a host list; looking up its IP address could give an abbreviated
2322 form, whereas the current host's name might or might not be abbreviated.
2323 The same fix has been applied.
2329 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2330 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2333 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2334 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2336 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2338 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2339 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2340 Also, for "Auto-Submitted", treat anything other than "no" as
2341 non-personal, in accordance with RFC 3834. (Previously it treated
2342 anything starting "auto-" as non-personal.)
2344 TF/01 The control=submission/name=... option had a problem with syntax
2345 errors if the name included a slash character. The /name= option
2346 now slurps the rest of the string, so it can include any characters
2347 but it must come last in the list of options (after /sender_retain
2350 PH/05 Some modifications to the interface to the fake nameserver for the new
2358 TK/01 Added the "success_on_redirect" address verification option. See
2359 NewStuff for rationale and an example.
2361 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2363 PH/02 Patch to exigrep to allow it to work on syslog lines.
2365 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2366 fread() to skip over the body file's header line, because in Cygwin the
2367 header line is locked and is inaccessible.
2369 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2370 co-exist for some time) to make it clear that it is the Exim ID that is
2371 referenced, not the Message-ID: header line.
2373 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2374 string_format() function, because snprintf() does not exist on all
2377 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2378 ${stat: expansion item.
2380 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2381 protocol synchronization error", to keep the pedants happy.
2383 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2384 well as for IRIX systems, when gcc is being used. See the host.c source
2387 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2389 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2392 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2393 transport and to the smtp transport in LMTP mode.
2395 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2397 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2398 run a filter in a subprocess. This could lead to confusion in subsequent
2399 lookups in the parent process. There should also be a search_tidyup() at
2400 the end of the subprocess.
2402 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2403 only if the host matched helo_try_verify_hosts, which caused the
2404 verification to occur when the EHLO/HELO command was issued. The ACL just
2405 tested the remembered result. Now, if a previous verification attempt has
2406 not happened, "verify = helo" does it there and then.
2408 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2410 TK/03 Fix log output including CR from clamd.
2412 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2413 exist provoked a memory error which could cause a segfault.
2415 PH/15 Installed PCRE 6.2
2417 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2419 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2420 of the problem. Specifically, suggested +O2 rather than +O1 for the
2423 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2425 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2426 record, the port was not correctly logged when the outgoing_port log
2427 selector was set (it logged the transort's default port).
2429 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2430 fallback_hosts, and "hosts" in the smtp transport.
2432 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2433 host errors such as "Connection refused".
2435 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2436 authentication with radiusclient 0.4.9:
2438 - Error returned from rc_read_config was caught wrongly
2439 - Username/password not passed on to radius server due to wrong length.
2441 The presumption is that some radiusclient API changes for 4.51/PH/17
2442 were not taken care of correctly. The code is still untested by me (my
2443 Linux distribution still has 0.3.2 of radiusclient), but it was
2444 contributed by a Radius user.
2446 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2447 expanding the "port" option of the smtp transport.
2449 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2450 while reading a MIME header. Thanks to Tom Hughes for a patch.
2452 PH/24 Include config.h inside local_scan.h so that configuration settings are
2455 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2456 that in an ACL for RCPT (for example), you can examine exactly what was
2459 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2460 commands, but it was not correctly comparing the address with the actual
2461 client host address. Thus, it would show the EHLO address in Received:
2462 header lines when this was not necessary.
2464 PH/27 Added the % operator to ${eval:}.
2466 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2467 it should be ignoring failures (because with -C, for example, it has lost
2468 privilege). It wasn't ignoring creation failures other than "already
2471 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2474 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2475 because an input line was too long, either on its own, or by virtue of
2476 too many continuations, the temporary file was not being removed, and the
2477 return code was incorrect.
2479 PH/31 Missing "BOOL" in function definition in filtertest.c.
2481 PH/32 Applied Sieve patches from the maintainer.
2483 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2485 PH/33 Added "verify = not_blind".
2487 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2488 Local/Makefile (with some defaults set). These are used in built scripts
2489 such as exicyclog, but they have never been used in the exim_install
2490 script (though there are many overriding facilities there). I have
2491 arranged that the exim_install script now takes note of these two
2494 PH/35 Installed configuration files for Dragonfly.
2496 PH/36 When a locally submitted message by a trusted user did not contain a
2497 From: header, and the sender address was obtained from -f or from an SMTP
2498 MAIL command, and the trusted user did not use -F to supply a sender
2499 name, $originator_name was incorrectly used when constructing a From:
2500 header. Furthermore, $originator_name was used for submission mode
2501 messages from external hosts without From: headers in a similar way,
2502 which is clearly wrong.
2504 PH/37 Added control=suppress_local_fixups.
2506 PH/38 When log_selector = +received_sender was set, and the addition of the
2507 sender made the log line's construction buffer exactly full, or one byte
2508 less than full, an overflow happened when the terminating "\n" was
2511 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2512 when the result of a list match is failure because a DNS lookup failed.
2514 PH/40 RM_COMMAND is now used in the building process.
2516 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2517 the "build-* directories that it finds.
2519 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2520 address in a domain literal was a prefix of an interface address.
2522 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2523 when verifying a sender address, unless rewrite_headers is false.
2525 PH/44 Wrote a long comment about why errors_to addresses are verified as
2526 recipients, not senders.
2528 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2529 the ratelimit ACL was added.
2531 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2533 PH/46 Added extra information about PostgreSQL errors to the error string.
2535 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2536 avoiding the need to install special zones in a real server. This is
2537 backwards compatible; if it can't find the fake resolver, it drops back.
2538 Thus, both old and new test suites can be run.
2540 TF/02 Added util/ratelimit.pl
2542 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2543 clock is set back in time.
2545 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2546 Candler <B.Candler@pobox.com>.
2548 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2549 OK for addresses that are the result of redirections.
2551 PH/48 A number of further additions for the benefit of the new test suite,
2552 including a fake gethostbyname() that interfaces to the fake DNS resolver
2555 TF/06 The fix for widen_domains has also been applied to qualify_single and
2556 search_parents which are the other dnslookup options that can cause
2559 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2562 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2564 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2565 service" if it encountered an MX record with an empty target hostname.
2566 The message is now "an MX or SRV record indicated no SMTP service".
2568 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2569 if the DNS of the sending site is misconfigured. This is quite a
2570 common situation. This change restores the behaviour of treating a
2571 helo verification defer as a failure.
2573 PH/51 If self=fail was set on a router, the bounce message did not include the
2574 actual error message.
2580 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2582 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2583 command itself ended in error, the underlying message about the transport
2584 filter timeout was being overwritten with the pipe command error. Now the
2585 underlying error message should be appended to the second error message.
2587 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2589 PH/02 Reduce the amount of output that "make" produces by default. Full output
2590 can still be requested.
2592 PH/03 The warning log line about a condition test deferring for a "warn" verb
2593 was being output only once per connection, rather than after each
2594 occurrence (because it was using the same function as for successful
2595 "warn" verbs). This seems wrong, so I have changed it.
2597 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2598 it should not have, which might have caused a crash in the right
2599 circumstances, but probably never did.
2601 PH/04 Installed a modified version of Tony Finch's patch to make submission
2602 mode fix the return path as well as the Sender: header line, and to
2603 add a /name= option so that you can make the user's friendly name appear
2606 TF/03 Added the control = fakedefer ACL modifier.
2608 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2609 Mark Lowes for thorough testing.
2611 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2613 TK/03 Merged latest SRS patch from Miles Wilton.
2615 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2616 with the definition in sysexits.h (which is #included earlier).
2617 Fortunately, Exim does not actually use EX_OK. The code used to try to
2618 preserve the sysexits.h value, by assumimg that macro definitions were
2619 scanned for macro replacements. I have been disabused of this notion,
2620 so now the code just undefines EX_OK before #including unistd.h.
2622 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2623 in the smtp transport. When a block could not be written in a single
2624 write() function, the timeout was being re-applied to each part-write.
2625 This seems wrong - if the receiver was accepting one byte at a time it
2626 would take for ever. The timeout is now adjusted when this happens. It
2627 doesn't have to be particularly precise.
2629 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2630 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2632 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2633 without a domain if the check to <postmaster@domain> fails.
2635 SC/01 Eximstats: added -xls and the ability to specify output files
2636 (patch written by Frank Heydlauf).
2638 SC/02 Eximstats: use FileHandles for outputing results.
2640 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2642 SC/04 Eximstats: fixed display of large numbers with -nvr option
2644 SC/05 Eximstats: fixed merging of reports with empty tables.
2646 SC/06 Eximstats: added the -include_original_destination flag
2648 SC/07 Eximstats: removed tabs and trailing whitespace.
2650 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2652 TK/06 MBOX spool code: Add real "From " MBOX separator line
2653 so the .eml file is really in mbox format (even though
2654 most programs do not really care). Patch from Alex Miller.
2656 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2657 The latter is generated from $received_to and is only set if the
2658 message has one envelope recipient. SA can use these headers,
2659 obviously out-of-the-box. Patch from Alex Miller.
2661 PH/08 The ${def test on a variable was returning false if the variable's
2662 value was "0", contrary to what the specification has always said!
2663 The result should be true unless the variable is empty.
2665 PH/09 The syntax error of a character other than { following "${if
2666 def:variable_name" (after optional whitespace) was not being diagnosed.
2667 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2668 accidental colon was present, for example, could give incorrect results.
2670 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2671 result is used (not including appendfile, where other changes are about
2674 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2675 This involved changing a lot of size variables from int to off_t. It
2676 should work with maildirs and everything.
2678 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2679 spamd dying while we are connected to it.
2681 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2684 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2685 function for BDB 4.3.
2687 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2689 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2690 the sources for PCRE have been re-arranged and more files are now
2693 PH/15 The code I had for printing potentially long long variables in PH/11
2694 above was not the best (it lost precision). The length of off_t variables
2695 is now inspected at build time, and an appropriate printing format (%ld
2696 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2697 to be "long long int" or "long int". This is needed for the internal
2698 formatting function string_vformat().
2700 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2701 the configuration file to be ":syslog", then the script "guesses" where
2702 the logs files are, rather than using the compiled in default. In our
2703 case the guess is not the same as the compiled default, so the script
2704 suddenly stopped working when I started to use syslog. The patch checks
2705 to see if log_file_path is "". If so, it attempts to read it from exim
2706 with no configuration file to get the compiled in version, before it
2707 falls back to the previous guessing code."
2709 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2710 implementing BATV in an Exim configuration. See NewStuff for the gory
2713 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2714 Makefile that are specific to HP-UX.
2716 PH/18 If the "use_postmaster" option was set for a recipient callout together
2717 with the "random" option, the postmaster address was used as the MAIL
2718 FROM address for the random test, but not for the subsequent recipient
2719 test. It is now used for both.
2721 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2722 patch removes a few documentation additions to RFC 3028, because the
2723 latest draft now contains them. It adds the new en;ascii-case comparator
2724 and a new error check for 8bit text in MIME parts. Comparator and
2725 require names are now matched exactly. I enabled the subaddress
2726 extension, but it is not well tested yet (read: it works for me)."
2728 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2729 rework some of the code of TK/09 above to avoid the hardwired use of
2730 "%lld" and "long long". Replaced the call to snprintf() with a call to
2733 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2734 records point to non-existent hosts", "retry timeout exceeded", and
2735 "retry time not reached for any host after a long failure period".
2737 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2738 experimental DomainKeys support:
2740 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2741 (2) On an error such as an illegally used "control", the wrong name for
2742 the control was given.
2744 These problems did NOT occur unless DomainKeys support was compiled.
2746 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2748 PH/24 Added ${if match_ip condition.
2750 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2751 databases so that it will be absolutely obvious if a crash occurs in the
2752 DB library. This is a regular occurrence (often caused by mis-matched
2755 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2756 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2757 picked up on a user's system that detects such things. There doesn't seem
2758 to be a gcc warning option for this - only an attribute that has to be
2759 put on the function's prototype. It seems that in Fedora Core 4 they have
2760 set this on a number of new functions. No doubt there will be more in due
2763 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2764 specify a transport. However, if an address that was verified by such a
2765 router was the subject of a callout, Exim crashed because it tried to
2766 read the rcpt_include_affixes from the non-existent transport. Now it
2767 just assumes that the setting of that option is false. This bug was
2768 introduced by 4.51/PH/31.
2770 PH/28 Changed -d+all to exclude +memory, because that information is very
2771 rarely of interest, but it makes the output a lot bigger. People tend to
2772 do -d+all out of habit.
2774 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2775 code in os-type was giving problems when libc.so lives in lib64, like on
2778 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2779 aren't the modern standard, and it seems that some systems' include files
2780 don't always have them. Exim was already checking for some of the newer
2781 ones like T_AAAA, and defining it itself. I've added checks for all the
2782 record types that Exim uses.
2784 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2785 not automatically generating a new one, as it is supposed to. This
2786 prevented TLS from working. If the file did exist, but contained invalid
2787 data, a new version was generated, as expected. It was only the case of a
2788 non-existent file that was broken.
2790 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2791 with a change in libdomainkeys > 0.64.
2793 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2794 from DNS. If the selector record carries the flag, it now has
2795 precedence over the domain-wide flag.
2797 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2799 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2800 the use of a port name that isn't defined in /etc/services) occurred, the
2801 message was deferred as in a normal delivery, and thus remained on the
2802 spool, instead of being failed because of the mua_wrapper setting. This
2803 is now fixed, and I tidied up some of the mua_wrapper messages at the
2806 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2807 the messages in a hash of arrays rather than using individual hashes.
2808 This is a bit cleaner and results in dramatic memory savings, albeit
2809 at a slight CPU cost.
2811 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2812 as requested by Marc Sherman.
2814 SC/10 Eximstats: added histograms for user specified patterns as requested
2817 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2819 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2820 fopen() in the content-scanning modules that did not already have it.
2826 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2827 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2829 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2831 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2833 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2834 to test Sieve filters that use "vacation".
2836 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2837 that changes the way the GnuTLS parameters are stored in the cache file.
2838 The new format can be generated externally. For backward compatibility,
2839 if the data in the cache doesn't make sense, Exim assumes it has read an
2840 old-format file, and it generates new data and writes a new file. This
2841 means that you can't go back to an older release without removing the
2844 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2845 work if there are any delivery delays because "one_time" forces the
2846 parent to be marked "delivered", so its unseen clone is never tried
2847 again. For this reason, Exim now forbids the simultaneous setting of
2850 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2851 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2852 entry if you want to know the details. The fix, however, neglected to
2853 consider the case when local delivery batching is involved. The test for
2854 "previously delivered" was not happening when checking to see if an
2855 address could be batched with a previous (undelivered) one; under
2856 certain circumstances this could lead to multiple deliveries to the same
2859 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2860 in its include files, and this causes problems building Exim.
2862 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2863 header_syntax) but Exim was just ignoring anything given after a slash.
2864 In particular, this caused confusion with an attempt to use "verify =
2865 reverse_host_lookup/defer_ok". An error is now given when options are
2866 supplied for verify items that do not have them. (Maybe reverse_host_
2867 lookup should have a defer_ok option, but that's a different point.)
2869 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2870 defined by RFC 821) to 2048, because there were problems with some AUTH
2871 commands, and RFC 1869 says the size should be increased for extended
2872 SMTP commands that take arguments.
2874 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2877 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2878 "unknown" error; now it says that the functionality isn't in the binary.
2880 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2881 an address' error message when a string expansion fails (syntax or
2882 whatever). Otherwise the password may appear in the log. Following change
2883 PH/42 below, there is no longer a chance of it appearing in a bounce
2886 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2888 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2889 of its addresses was ever tried. (Bugzilla bug #2.)
2891 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2892 the result incorrectly in the debug output. (It correctly added a newline
2893 to what was transported.)
2895 TF/01 Added $received_time.
2897 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2898 commented out examples of how to interface to a virus scanner and to
2899 SpamAssassin. Also added commented examples of av_scanner and
2900 spamd_address settings.
2902 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2903 and controls are allowed in which ACLs. There were a couple of minor
2904 errors. Some of the entries in the conditions table (which is a table of
2905 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2906 negation of where the condition IS allowed.
2908 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2910 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2911 header file does not have a version number, so I've had to invent a new
2912 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2913 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2914 radiusclient), but it was contributed by a Radius user.
2916 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2917 files or directories, for OpenSSL.
2919 PH/19 When an Exim process that is running as root has to create an Exim log
2920 file, it does so in a subprocess that runs as exim:exim so as to get the
2921 ownership right at creation (otherwise, other Exim processes might see
2922 the file with the wrong ownership). There was no test for failure of this
2923 fork() call, which would lead to the process getting stuck as it waited
2924 for a non-existent subprocess. Forks do occasionally fail when resources
2925 run out. I reviewed all the other calls to fork(); they all seem to check
2928 PH/20 When checking for unexpected SMTP input at connect time (before writing
2929 the banner), Exim was not dealing correctly with a non-positive return
2930 from the read() function. If the client had disconnected by this time,
2931 the result was a log entry for a synchronization error with an empty
2932 string after "input=" when read() returned zero. If read() returned -1
2933 (an event I could not check), uninitialized data bytes were printed.
2934 There were reports of junk text (parts of files, etc) appearing after
2937 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2939 PH/22 Added support for macro redefinition, and (re)definition in between
2940 driver and ACL definitions.
2942 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2943 forgetting to use the resulting value; it was using the unexpanded value.
2945 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2946 hadn't been configured. The fix is from Juergen Kreileder, who
2947 understands it better than I do:
2949 "Here's what I see happening with three configured cyrus_sasl
2950 authenticators configured (plain, login, cram-md5):
2952 On startup auth_cyrus_sasl_init() gets called for each of these.
2953 This means three calls to sasl_listmech() without a specified mech_list.
2954 => SASL tests which mechs of all available mechs actually work
2955 => three warnings about OTP not working
2956 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2958 With the patch, sasl_listmech() also gets called three times. But now
2959 SASL's mech_list option is set to the server_mech specified in the the
2960 authenticator. Or in other words, the answer from sasl_listmech()
2961 gets limited to just the mech you're testing for (which is different
2963 => the return list contains just 'plain' or 'login', 'cram-md5' or
2964 nothing depending on the value of ob->server_mech.
2966 I've just tested the patch: Authentication still works fine,
2967 unavailable mechs specified in the exim configuration are still
2968 caught, and the auth.log warnings about OTP are gone."
2970 PH/25 When debugging is enabled, the contents of the command line are added
2971 to the debugging output, even when log_selector=+arguments is not
2974 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2975 answer is "GNU", and only if the return is "GNU/something" is the answer
2978 PH/27 $acl_verify_message is now set immediately after the failure of a
2979 verification in an ACL, and so is available in subsequent modifiers. In
2980 particular, the message can be preserved by coding like this:
2982 warn !verify = sender
2983 set acl_m0 = $acl_verify_message
2985 Previously, $acl_verify_message was set only while expanding "message"
2986 and "log_message" when a very denied access.
2988 PH/28 Modified OS/os.c-Linux with
2990 -#ifndef OS_LOAD_AVERAGE
2991 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2993 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2994 nomenclature these days.)
2996 PH/29 Installed patch from the Sieve maintainer that adds the options
2997 sieve_useraddress and sieve_subaddress to the redirect router.
2999 PH/30 In these circumstances:
3000 . Two addresses routed to the same list of hosts;
3001 . First host does not offer TLS;
3002 . First host accepts first address;
3003 . First host gives temporary error to second address;
3004 . Second host offers TLS and a TLS session is established;
3005 . Second host accepts second address.
3006 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3007 and peerdn, if requested) that were in fact used only for the second
3010 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3011 attention to any local part prefix or suffix that was matched by the
3012 router that accepted the address. It now behaves in the same way as it
3013 does for delivery: the affixes are removed from the local part unless
3014 rcpt_include_affixes is set on the transport.
3016 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3017 timeout during the DATA phase of an incoming message.
3019 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3020 applied a patch sanctioned by the Sieve maintainer.
3022 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3023 the uid or gid is negative. A case of a negative gid caused this to be
3024 noticed. The fix allows for either to be negative.
3026 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3027 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3030 PH/36 The change PH/12 above was broken. Fixed it.
3032 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3033 the grounds that routing the same address twice would always produce the
3034 same answer. This might have been true once, but it is certainly no
3035 longer true now. Routing a child address may depend on the previous
3036 routing that produced that child. Some complicated redirection strategies
3037 went wrong when messages had multiple recipients, and made Exim's
3038 behaviour dependent on the order in which the addresses were given.
3040 I have moved the duplicate checking until after the routing is complete.
3041 Exim scans the addresses that are assigned to local and remote
3042 transports, and removes any duplicates. This means that more work will be
3043 done, as duplicates will always all be routed, but duplicates are
3044 presumably rare, so I don't expect this is of any significance.
3046 For deliveries to pipes, files, and autoreplies, the duplicate checking
3047 still happens during the routing process, since they are not going to be
3050 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3051 It corrects a timeout issue with spamd. This is Ian's comment: "The
3052 background is that sometimes spamd either never reads data from a
3053 connection it has accepted, or it never writes response data. The exiscan
3054 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3055 blindly assumes that writes won't block so it may never time out."
3057 PH/39 Allow G after quota size as well as K and M.
3059 PH/40 The value set for $authenticated_id in an authenticator may not contain
3060 binary zeroes or newlines because the value is written to log lines and
3061 to spool files. There was no check on this. Now the value is run through
3062 the string_printing() function so that such characters are converted to
3063 printable escape sequences.
3065 PH/41 $message_linecount is a new variable that contains the total number of
3066 lines in the message. Compare $body_linecount, which is the count for the
3069 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3070 bounce and delay warning messages, except in certain special cases, which
3073 (a) An SMTP error message from a remote host;
3074 (b) A message specified in a :fail: redirection;
3075 (c) A message specified in a "fail" command in a system filter;
3076 (d) A message specified in a FAIL return from the queryprogram router;
3077 (e) A message specified by the cannot_route_message router option.
3079 In these cases only, Exim does include the error details in bounce and
3080 warning messages. There are also a few cases where bland messages such
3081 as "unrouteable address" or "local delivery error" are given.
3083 PH/43 $value is now also set for the "else" part of a ${run expansion.
3085 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3086 being worked on, but at least Exim now implements the latest version to
3089 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3090 process to complete was treated as a delivery failure, a timeout while
3091 writing the message to the pipe was logged, but erroneously treated as a
3092 successful delivery. Such timeouts include transport filter timeouts. For
3093 consistency with the overall process timeout, these timeouts are now
3094 treated as errors, giving rise to delivery failures by default. However,
3095 there is now a new Boolean option for the pipe transport called
3096 timeout_defer, which, if set TRUE, converts the failures into defers for
3097 both kinds of timeout. A transport filter timeout is now identified in
3100 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3101 systems where "make" and "gmake" are different, calling "gmake" at top
3102 level broke things. I've arranged for the value of $(MAKE) to be passed
3103 from the Makefile to this script so that it can call the same version of
3107 A note about Exim versions 4.44 and 4.50
3108 ----------------------------------------
3110 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3111 changes of various kinds. As a consequence, a big documentation update was
3112 needed. This delayed the release for rather longer than seemed good, especially
3113 in the light of a couple of (minor) security issues. Therefore, the changes
3114 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3115 release. So 4.44 and 4.50 are in effect two different branches that both start
3118 I have left the 4.50 change log unchanged; it contains all the changes since
3119 4.43. The change log for 4.44 is below; many of its items are identical to
3120 those for 4.50. This seems to be the most sensible way to preserve the
3121 historical information.
3127 1. Minor wording change to the doc/README.SIEVE file.
3129 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3130 computation of the current number of files was incorrect.
3132 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3133 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3136 4. Give more explanation in the error message when the command for a transport
3137 filter fails to execute.
3139 5. There are several places where Exim runs a non-Exim command in a
3140 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3141 was being done only for the command run by the queryprogram router. It is
3142 now done for all such subprocesses. The other cases are: ${run, transport
3143 filters, and the commands run by the lmtp and pipe transports.
3145 6. Added CONFIGURE_GROUP build-time option.
3147 7. Some older OS have a limit of 256 on the maximum number of file
3148 descriptors. Exim was using setrlimit() to set 1000 as a large value
3149 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3150 systems. I've change it so that if it can't get 1000, it tries for 256.
3152 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3153 was an oversight, and furthermore, ever since the addition of extra
3154 controls (e.g. 4.43/32), the checks on when to allow different forms of
3155 "control" were broken. There should now be diagnostics for all cases when a
3156 control that does not make sense is encountered.
3158 9. Added the /retain_sender option to "control=submission".
3160 10. $recipients is now available in the predata ACL (oversight).
3162 11. Tidy the search cache before the fork to do a delivery from a message
3163 received from the command line. Otherwise the child will trigger a lookup
3164 failure and thereby defer the delivery if it tries to use (for example) a
3165 cached ldap connection that the parent has called unbind on.
3167 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3168 of $address_data from the recipient verification was clobbered by the
3169 sender verification.
3171 13. The value of address_data from a sender verification is now available in
3172 $sender_address_data in subsequent conditions in the ACL statement.
3174 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3176 15. Added a new option "connect=<time>" to callout options, to set a different
3179 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3180 was its contents. (It was OK if the option was not defined at all.)
3182 17. A "Completed" log line is now written for messages that are removed from
3183 the spool by the -Mrm option.
3185 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3186 information about exactly what failed.
3188 19. Added -dd to debug only the daemon process.
3190 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3191 handles timeouts, both on the server side and network timeouts. Renamed the
3192 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3194 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3195 It is now set to "smtps".
3197 22. $host_address is now set to the target address during the checking of
3198 ignore_target_hosts.
3200 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3201 being passed; this would have caused $sender_host_name to have been used if
3202 matching the list had actually called for a host name (not very likely,
3203 since this list is usually IP addresses). A host name is now passed as
3206 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3207 code that specifies a non-restarting handler (typically sigaction() in
3208 modern systems) in an attempt to fix a rare and obscure crash bug.
3210 25. Narrowed the window for a race in the daemon that could cause it to ignore
3211 SIGCHLD signals. This is not a major problem, because they are used only to
3212 wake it up if nothing else does.
3214 26. A malformed maildirsize file could cause Exim to calculate negative values
3215 for the mailbox size or file count. Odd effects could occur as a result.
3216 The maildirsize information is now recalculated if the size or filecount
3219 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3220 support for a long time. Removed HAVE_SYS_VFS_H.
3222 28. Installed the latest version of exipick from John Jetmore.
3224 29. In an address list, if the pattern was not a regular expression, an empty
3225 subject address (from a bounce message) matched only if the pattern was an
3226 empty string. Non-empty patterns were not even tested. This was the wrong
3227 because it is perfectly reasonable to use an empty address as part of a
3228 database query. An empty address is now tested by patterns that are
3229 lookups. However, all the other forms of pattern expect the subject to
3230 contain a local part and a domain, and therefore, for them, an empty
3231 address still always fails if the pattern is not itself empty.
3233 30. Exim went into a mad DNS loop when attempting to do a callout where the
3234 host was specified on an smtp transport, and looking it up yielded more
3235 than one IP address.
3237 31. Re-factored the code for checking spool and log partition space into a
3238 function that finds that data and another that does the check. The former
3239 is then used to implement four new variables: $spool_space, $log_space,
3240 $spool_inodes, and $log_inodes.
3242 32. The RFC2047 encoding function was originally intended for short strings
3243 such as real names; it was not keeping to the 75-character limit for
3244 encoded words that the RFC imposes. It now respects the limit, and
3245 generates multiple encoded words if necessary. To be on the safe side, I
3246 have increased the buffer size for the ${rfc2047: expansion operator from
3249 33. It is now permitted to omit both strings after an "if" condition; if the
3250 condition is true, the result is "true". As before, when the second string
3251 is omitted, a false condition yields an empty string. This makes it less
3252 cumbersome to write custom ACL and router conditions.
3254 34. Failure to deliver a bounce message always caused it to be frozen, even if
3255 there was an errors_to setting on the router. The errors_to setting is now
3258 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3259 canonical form (fully expanded) before being placed in
3260 $sender_host_address.
3262 36. The table in the code that translates DNS record types into text (T_A to
3263 "A" for instance) was missing entries for NS and CNAME. It is just possible
3264 that this could have caused confusion if both these types were looked up
3265 for the same domain, because the text type is used as part of Exim's
3266 per-process caching. But the chance of anyone hitting this buglet seems
3269 37. The dnsdb lookup has been extended in a number of ways.
3271 (1) There is a new type, "zns", which walks up the domain tree until it
3272 finds some nameserver records. It should be used with care.
3274 (2) There is a new type, "mxh", which is like "mx" except that it returns
3275 just the host names, not the priorities.
3277 (3) It is now possible to give a list of domains (or IP addresses) to be
3278 looked up. The behaviour when one of the lookups defers can be
3279 controlled by a keyword.
3281 (4) It is now possible to specify the separator character for use when
3282 multiple records are returned.
3284 38. The dnslists ACL condition has been extended: it is now possible to supply
3285 a list of IP addresses and/or domains to be looked up in a particular DNS
3288 39. Added log_selector=+queue_time_overall.
3290 40. When running the queue in the test harness, wait just a tad after forking a
3291 delivery process, to get repeatability of debugging output.
3293 41. Include certificate and key file names in error message when GnuTLS fails
3294 to set them up, because the GnuTLS error message doesn't include the name
3295 of the failing file when there is a problem reading it.
3297 42. Allow both -bf and -bF in the same test run.
3299 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3301 44. The "Exiscan patch" is now merged into the mainline Exim source.
3303 45. Sometimes the final signoff response after QUIT could fail to get
3304 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3305 < 0 before doing a fflush(). This bug looks as though it goes back to the
3306 introduction of TLS in release 3.20, but "sometimes" must have been rare
3307 because the tests only now provoked it.
3309 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3310 (this can affect the format of dates).
3312 47. exim_tidydb, when checking for the continued existence of a message for
3313 which it has found a message-specific retry record, was not finding
3314 messages that were in split spool directories. Consequently, it was
3315 deleting retry records that should have stayed in existence.
3317 48. Steve fixed some bugs in eximstats.
3319 49. The SPA authentication driver was not abandoning authentication and moving
3320 on to the next authenticator when an expansion was forced to fail,
3321 contradicting the general specification for all authenticators. Instead it
3322 was generating a temporary error. It now behaves as specified.
3324 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3325 (the order specifies the preference for clients). The order is now AES256,
3326 AES128, 3DES, ARCFOUR128.
3328 51. Small patch to Sieve code - explicitly set From: when generating an
3331 52. Exim crashed if a remote delivery caused a very long error message to be
3332 recorded - for instance if somebody sent an entire SpamAssassin report back
3333 as a large number of 550 error lines. This bug was coincidentally fixed by
3334 increasing the size of one of Exim's internal buffers (big_buffer) that
3335 happened as part of the Exiscan merge. However, to be on the safe side, I
3336 have made the code more robust (and fixed the comments that describe what
3339 53. Now that there can be additional text after "Completed" in log lines (if
3340 the queue_time_overall log selector is set), a one-byte patch to exigrep
3341 was needed to allow it to recognize "Completed" as not the last thing in
3344 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3345 patch that reportedly fixes this has been added. I am not expert enough to
3346 create a test for it. This is what the patch creator wrote:
3348 "I found a little strange behaviour of ldap code when working with
3349 Windows 2003 AD Domain, where users was placed in more than one
3350 Organization Units. When I tried to give exim partial DN, the exit code
3351 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3352 But simultaneously result of request was absolutely normal ldap result,
3353 so I produce this patch..."
3355 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3356 REFERENCE, so I have modified the code to exclude the patch when that macro
3359 55. Some experimental protocols are using DNS PTR records for new purposes. The
3360 keys for these records are domain names, not reversed IP addresses. The
3361 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3362 leaves it alone. Component reversal etc. now happens only for IP addresses.
3365 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3367 57. Double the size of the debug message buffer (to 2048) so that more of very
3368 long debug lines gets shown.
3370 58. The exicyclog utility now does better if the number of log files to keep
3371 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3373 59. Two changes related to the smtp_active_hostname option:
3375 (1) $smtp_active_hostname is now available as a variable.
3376 (2) The default for smtp_banner uses $smtp_active_hostname instead
3377 of $primary_hostname.
3379 60. The host_aton() function is supposed to be passed a string that is known
3380 to be a valid IP address. However, in the case of IPv6 addresses, it was
3381 not checking this. This is a hostage to fortune. Exim now panics and dies
3382 if the condition is not met. A case was found where this could be provoked
3383 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3384 components; fortuitously, this particular loophole had already been fixed
3385 by change 4.50/55 above.
3387 If there are any other similar loopholes, the new check in host_aton()
3388 itself should stop them being exploited. The report I received stated that
3389 data on the command line could provoke the exploit when Exim was running as
3390 exim, but did not say which command line option was involved. All I could
3391 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3392 running as the user.
3395 61. There was a buffer overflow vulnerability in the SPA authentication code
3396 (which came originally from the Samba project). I have added a test to the
3397 spa_base64_to_bits() function which I hope fixes it.
3400 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3401 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3403 63. The daemon start-up calls getloadavg() while still root for those OS that
3404 need the first call to be done as root, but it missed one case: when
3405 deliver_queue_load_max is set with deliver_drop_privilege. This is
3406 necessary for the benefit of the queue runner, because there is no re-exec
3407 when deliver_drop_privilege is set.
3409 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3410 This has been fixed.
3412 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3413 was in use, was not putting the data itself into the right store pool;
3414 consequently, it could be overwritten for a subsequent message in the same
3415 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3418 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3420 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3421 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3422 the function were treating the return as a boolean value, which happened to
3423 work because 0=false and not-0=true, but is not correct code.
3425 68. The host_aton() function was not handling scoped IPv6 addresses (those
3426 with, for example, "%eth0" on the end) correctly.
3428 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3429 negated items (that is, ~something) in unsigned ints. Some compilers
3430 apparently mutter when there is no cast.
3432 70. If an address verification called from an ACL failed, and did not produce a
3433 user-specific message (i.e. there was only a "system" message), nothing was
3434 put in $acl_verify_message. In this situation, it now puts the system
3437 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3438 session; change 4.31/43 added the unwanted input to the log line - except
3439 that it did not do this in the start of session case. It now does.
3441 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3442 This is wrong and can cause the other end to generate a synchronization
3443 error if it is another Exim or anything else that does the synchronization
3444 check. A QUIT command is no longer sent after a timeout.
3446 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3447 during host lookups.
3449 74. The defer_ok option of callout verification was not working if it was used
3450 when verifying addresses in header lines, that is, for this case:
3452 verify = header_sender/callout=defer_ok
3454 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3455 those file descriptors could be used for SMTP connections. If anything
3456 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3457 could be sent to the SMTP client, causing chaos. The daemon now opens
3458 stdin, stdout, and stderr to /dev/null when it puts itself into the
3461 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3462 log by default. The user can override this with suitable Perl magic.
3464 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3465 the log message when discard triggers, was not working for the DATA ACL or
3466 for the non-SMTP ACL.
3468 78. Error message wording change in sieve.c.
3470 79. If smtp_accept_max_per_host was set, the number of connections could be
3471 restricted to fewer than expected, because the daemon was trying to set up
3472 a new connection before checking whether the processes handling previous
3473 connections had finished. The check for completed processes is now done
3474 earlier. On busy systems, this bug wouldn't be noticed because something
3475 else would have woken the daemon, and it would have reaped the completed
3478 80. If a message was submitted locally by a user whose login name contained one
3479 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3480 It caused a spool format error. I have fixed the spool reading code. A
3481 related problem was that the "from" clause in the Received: line became
3482 illegal because of the space(s). It is now covered by ${quote_local_part.
3484 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3487 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3489 83. Patch from Sieve maintainer to fix unterminated string problem in
3490 "vacation" handling.
3492 84. Some minor changes to the Linux configuration files to help with other
3493 OS variants using glibc.
3495 85. One more patch for Sieve to update vacation handling to latest spec.
3498 ----------------------------------------------------
3499 See the note above about the 4.44 and 4.50 releases.
3500 ----------------------------------------------------
3506 1. Change 4.43/35 introduced a bug that caused file counts to be
3507 incorrectly computed when quota_filecount was set in an appendfile
3510 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3511 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3514 3. Give more explanation in the error message when the command for a transport
3515 filter fails to execute.
3517 4. There are several places where Exim runs a non-Exim command in a
3518 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3519 was being done only for the command run by the queryprogram router. It is
3520 now done for all such subprocesses. The other cases are: ${run, transport
3521 filters, and the commands run by the lmtp and pipe transports.
3523 5. Some older OS have a limit of 256 on the maximum number of file
3524 descriptors. Exim was using setrlimit() to set 1000 as a large value
3525 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3526 systems. I've change it so that if it can't get 1000, it tries for 256.
3528 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3529 was an oversight, and furthermore, ever since the addition of extra
3530 controls (e.g. 4.43/32), the checks on when to allow different forms of
3531 "control" were broken. There should now be diagnostics for all cases when a
3532 control that does not make sense is encountered.
3534 7. $recipients is now available in the predata ACL (oversight).
3536 8. Tidy the search cache before the fork to do a delivery from a message
3537 received from the command line. Otherwise the child will trigger a lookup
3538 failure and thereby defer the delivery if it tries to use (for example) a
3539 cached ldap connection that the parent has called unbind on.
3541 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3542 of $address_data from the recipient verification was clobbered by the
3543 sender verification.
3545 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3546 was its contents. (It was OK if the option was not defined at all.)
3548 11. A "Completed" log line is now written for messages that are removed from
3549 the spool by the -Mrm option.
3551 12. $host_address is now set to the target address during the checking of
3552 ignore_target_hosts.
3554 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3555 being passed; this would have caused $sender_host_name to have been used if
3556 matching the list had actually called for a host name (not very likely,
3557 since this list is usually IP addresses). A host name is now passed as
3560 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3561 code that specifies a non-restarting handler (typically sigaction() in
3562 modern systems) in an attempt to fix a rare and obscure crash bug.
3564 15. Narrowed the window for a race in the daemon that could cause it to ignore
3565 SIGCHLD signals. This is not a major problem, because they are used only to
3566 wake it up if nothing else does.
3568 16. A malformed maildirsize file could cause Exim to calculate negative values
3569 for the mailbox size or file count. Odd effects could occur as a result.
3570 The maildirsize information is now recalculated if the size or filecount
3573 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3574 support for a long time. Removed HAVE_SYS_VFS_H.
3576 18. Updated exipick to current release from John Jetmore.
3578 19. Allow an empty sender to be matched against a lookup in an address list.
3579 Previously the only cases considered were a regular expression, or an
3582 20. Exim went into a mad DNS lookup loop when doing a callout where the
3583 host was specified on the transport, if the DNS lookup yielded more than
3586 21. The RFC2047 encoding function was originally intended for short strings
3587 such as real names; it was not keeping to the 75-character limit for
3588 encoded words that the RFC imposes. It now respects the limit, and
3589 generates multiple encoded words if necessary. To be on the safe side, I
3590 have increased the buffer size for the ${rfc2047: expansion operator from
3593 22. Failure to deliver a bounce message always caused it to be frozen, even if
3594 there was an errors_to setting on the router. The errors_to setting is now
3597 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3598 canonical form (fully expanded) before being placed in
3599 $sender_host_address.
3601 24. Updated eximstats to version 1.33
3603 25. Include certificate and key file names in error message when GnuTLS fails
3604 to set them up, because the GnuTLS error message doesn't include the name
3605 of the failing file when there is a problem reading it.
3607 26. Expand error message when OpenSSL has problems setting up cert/key files.
3610 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3611 (this can affect the format of dates).
3613 28. exim_tidydb, when checking for the continued existence of a message for
3614 which it has found a message-specific retry record, was not finding
3615 messages that were in split spool directories. Consequently, it was
3616 deleting retry records that should have stayed in existence.
3618 29. eximstats updated to version 1.35
3619 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3620 1.35 - bugfix such that pie charts by volume are generated correctly
3622 30. The SPA authentication driver was not abandoning authentication and moving
3623 on to the next authenticator when an expansion was forced to fail,
3624 contradicting the general specification for all authenticators. Instead it
3625 was generating a temporary error. It now behaves as specified.
3627 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3628 (the order specifies the preference for clients). The order is now AES256,
3629 AES128, 3DES, ARCFOUR128.
3631 31. Small patch to Sieve code - explicitly set From: when generating an
3634 32. Exim crashed if a remote delivery caused a very long error message to be
3635 recorded - for instance if somebody sent an entire SpamAssassin report back
3636 as a large number of 550 error lines. This bug was coincidentally fixed by
3637 increasing the size of one of Exim's internal buffers (big_buffer) that
3638 happened as part of the Exiscan merge. However, to be on the safe side, I
3639 have made the code more robust (and fixed the comments that describe what
3642 33. Some experimental protocols are using DNS PTR records for new purposes. The
3643 keys for these records are domain names, not reversed IP addresses. The
3644 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3645 leaves it alone. Component reversal etc. now happens only for IP addresses.
3648 34. The host_aton() function is supposed to be passed a string that is known
3649 to be a valid IP address. However, in the case of IPv6 addresses, it was
3650 not checking this. This is a hostage to fortune. Exim now panics and dies
3651 if the condition is not met. A case was found where this could be provoked
3652 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3653 components; fortuitously, this particular loophole had already been fixed
3654 by change 4.50/55 or 4.44/33 above.
3656 If there are any other similar loopholes, the new check in host_aton()
3657 itself should stop them being exploited. The report I received stated that
3658 data on the command line could provoke the exploit when Exim was running as
3659 exim, but did not say which command line option was involved. All I could
3660 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3661 running as the user.
3664 35. There was a buffer overflow vulnerability in the SPA authentication code
3665 (which came originally from the Samba project). I have added a test to the
3666 spa_base64_to_bits() function which I hope fixes it.
3669 36. The daemon start-up calls getloadavg() while still root for those OS that
3670 need the first call to be done as root, but it missed one case: when
3671 deliver_queue_load_max is set with deliver_drop_privilege. This is
3672 necessary for the benefit of the queue runner, because there is no re-exec
3673 when deliver_drop_privilege is set.
3675 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3676 was in use, was not putting the data itself into the right store pool;
3677 consequently, it could be overwritten for a subsequent message in the same
3678 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3681 38. Sometimes the final signoff response after QUIT could fail to get
3682 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3683 < 0 before doing a fflush(). This bug looks as though it goes back to the
3684 introduction of TLS in release 3.20, but "sometimes" must have been rare
3685 because the tests only now provoked it.
3691 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3692 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3693 Now it returns an integer. A number of calls were still expecting a T/F
3694 return. Fortuitously, in all cases, the tests worked in OK situations,
3695 which is the norm. However, things would have gone wrong on any write
3696 failures on the smtp file descriptor. This function is used when sending
3697 messages over SMTP and also when doing verify callouts.
3699 2. When Exim is called to do synchronous delivery of a locally submitted
3700 message (the -odf or -odi options), it no longer closes stderr before doing
3703 3. Implemented the mua_wrapper option.
3705 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3707 5. Implemented the functions header_remove(), header_testname(),
3708 header_add_at_position(), and receive_remove_recipient(), and exported them
3711 6. If an ACL "warn" statement specified the addition of headers, Exim already
3712 inserted X-ACL-Warn: at the start if there was no header name. However, it
3713 was not making this test for the second and subsequent header lines if
3714 there were newlines in the string. This meant that an invalid header could
3715 be inserted if Exim was badly configured.
3717 7. Allow an ACL "warn" statement to add header lines at the start or after all
3718 the Received: headers, as well as at the end.
3720 8. Added the rcpt_4xx retry error code.
3722 9. Added postmaster_mailfrom=xxx to callout verification option.
3724 10. Added mailfrom=xxxx to the callout verification option, for verify=
3727 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3728 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3730 12. Inserted some casts to stop certain compilers warning when using pointer
3731 differences as field lengths or precisions in printf-type calls (mostly
3732 affecting debugging statements).
3734 13. Added optional readline() support for -be (dynamically loaded).
3736 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3737 same clock tick as a message's arrival, so that its received time was the
3738 same as the "first fail" time on the retry record, and that message
3739 remained on the queue past the ultimate address timeout, every queue runner
3740 would try a delivery (because it was past the ultimate address timeout) but
3741 after another failure, the ultimate address timeout, which should have then
3742 bounced the address, did not kick in. This was a "< instead of <=" error;
3743 in most cases the first failure would have been in the next clock tick
3744 after the received time, and all would be well.
3746 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3747 being recognized when the domain list was tested by the match_domain
3748 condition in an expansion string.
3750 16. Added the ${str2b64: operator.
3752 17. Exim was always calling setrlimit() to set a large limit for the number of
3753 processes, without checking whether the existing limit was already
3754 adequate. (It did check for the limit on file descriptors.) Furthermore,
3755 errors from getrlimit() and setrlimit() were being ignored. Now they are
3756 logged to the main and panic logs, but Exim does carry on, to try to do its
3757 job under whatever limits there are.
3759 18. Imported PCRE 5.0.
3761 19. Trivial typo in log message " temporarily refused connection" (the leading
3764 20. If the log selector return_path_on_delivery was set and an address was
3765 redirected to /dev/null, the delivery process crashed because it assumed
3766 that a return path would always be set for a "successful" delivery. In this
3767 case, the whole delivery is bypassed as an optimization, and therefore no
3770 21. Internal re-arrangement: the function for sending a challenge and reading
3771 a response while authentication was assuming a zero-terminated challenge
3772 string. It's now changed to take a pointer and a length, to allow for
3773 binary data in such strings.
3775 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3777 23. Exim was not respecting finduser_retries when seeking the login of the
3778 uid under which it was called; it was always trying 10 times. (The default
3779 setting of finduser_retries is zero.) Also, it was sleeping after the final
3780 failure, which is pointless.
3782 24. Implemented tls_on_connect_ports.
3784 25. Implemented acl_smtp_predata.
3786 26. If the domain in control=submission is set empty, Exim assumes that the
3787 authenticated id is a complete email address when it generates From: or
3788 Sender: header lines.
3790 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3791 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3792 chown and chgrp in /bin and hostname in /usr/bin.
3794 28. Exim was keeping the "process log" file open after each use, just as it
3795 does for the main log. This opens the possibility of it remaining open for
3796 long periods when the USR1 signal hits a daemon. Occasional processlog
3797 errors were reported, that could have been caused by this. Anyway, it seems
3798 much more sensible not to leave this file open at all, so that is what now
3801 29. The long-running daemon process does not normally write to the log once it
3802 has entered its main loop, and it closes the log before doing so. This is
3803 so that log files can straightforwardly be renamed and moved. However,
3804 there are a couple of unusual error situations where the daemon does write
3805 log entries, and I had neglected to close the log afterwards.
3807 30. The text of an SMTP error response that was received during a remote
3808 delivery was being truncated at 512 bytes. This is too short for some of
3809 the long messages that one sometimes sees. I've increased the limit to
3812 31. It is now possible to make retry rules that apply only when a message has a
3813 specific sender, in particular, an empty sender.
3815 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3816 it possible to be selective about when SMTP synchronization is enforced.
3818 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3820 32. Implemented hosts_connection_nolog.
3822 33. Added an ACL for QUIT.
3824 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3827 35. Added mailbox_size and mailbox_filecount to appendfile.
3829 36. Added control = no_multiline_responses to ACLs.
3831 37. There was a bug in the logic of the code that waits for the clock to tick
3832 in the case where the clock went backwards by a substantial amount such
3833 that the microsecond fraction of "now" was more than the microsecond
3834 fraction of "then" (but the whole seconds number was less).
3836 38. Added support for the libradius Radius client library this is found on
3837 FreeBSD (previously only the radiusclient library was supported).
3843 1. When certain lookups returned multiple values in the form name=value, the
3844 quoting of the values was not always being done properly. Specifically:
3845 (a) If the value started with a double quote, but contained no whitespace,
3847 (b) If the value contained whitespace other than a space character (i.e.
3848 tabs or newlines or carriage returns) it was not quoted.
3849 This fix has been applied to the mysql and pgsql lookups by writing a
3850 separate quoting function and calling it from the lookup code. The fix
3851 should probably also be applied to nisplus, ibase and oracle lookups, but
3852 since I cannot test any of those, I have not disturbed their existing code.
3854 2. A hit in the callout cache for a specific address caused a log line with no
3855 reason for rejecting RCPT. Now it says "Previous (cached) callout
3856 verification failure".
3858 3. There was an off-by-one bug in the queryprogram router. An over-long
3859 return line was truncated at 256 instead of 255 characters, thereby
3860 overflowing its buffer with the terminating zero. As well as fixing this, I
3861 have increased the buffer size to 1024 (and made a note to document this).
3863 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3864 when Exim is waiting for an SMTP response from a remote server, Exim
3865 restarts its select() call on the socket, thereby resetting its timeout.
3866 This is not a problem when such interrupts are rare. Somebody set up a cron
3867 job to run exiwhat every 2 minutes, which is less than the normal select()
3868 timeout (5 or 10 minutes). This meant that the select() timeout never
3869 kicked in because it was always reset. I have fixed this by comparing the
3870 time when an interrupt arrives with the time at the start of the first call
3871 to select(). If more time than the timeout has elapsed, the interrupt is
3872 treated as a timeout.
3874 5. Some internal re-factoring in preparation for the addition of Sieve
3875 extensions (by MH). In particular, the "personal" test is moved to a
3876 separate function, and given an option for scanning Cc: and Bcc: (which is
3877 not set for Exim filters).
3879 6. When Exim created an email address using the login of the caller as the
3880 local part (e.g. when creating a From: or Sender: header line), it was not
3881 quoting the local part when it contained special characters such as @.
3883 7. Installed new OpenBSD configuration files.
3885 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3886 try to make them clearer.
3888 9. Callout options, other than the timeout value, were being ignored when
3889 verifying sender addresses in header lines. For example, when using
3891 verify = header_sender/callout=no_cache
3893 the cache was (incorrectly) being used.
3895 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3896 only the Cygwin environment.
3898 11. When return_path_on_delivery was set as a log selector, if different remote
3899 addresses in the same message used different return paths and parallel
3900 remote delivery occurred, the wrong values would sometimes be logged.
3901 (Whenever a remote delivery process finished, the return path value from
3902 the most recently started remote delivery process was logged.)
3904 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3905 lines when AUTH and/or TLS are in use. This is the "received protocol"
3906 field. Exim used to use "asmtp" for authenticated SMTP, without any
3907 indication (in the protocol name) for TLS use. Now it follows the RFC and
3908 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3909 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3910 names appear in log lines as well as in Received: header lines.
3912 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3913 extensions, and comparison tests, and to fix some bugs.
3915 14. Changes to the "personal" filter test:
3917 (1) The test was buggy in that it was just doing the equivalent of
3918 "contains" tests on header lines. For example, if a user's address was
3919 anne@some.where, the "personal" test would incorrectly be true for
3921 To: susanne@some.where
3923 This test is now done by extracting each address from the header in turn,
3924 and checking the entire address. Other tests that are part of "personal"
3925 are now done using regular expressions (for example, to check local parts
3926 of addresses in From: header lines).
3928 (2) The list of non-personal local parts in From: addresses has been
3929 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3930 taken from the Sieve specification recommendations.
3932 (3) If the message contains any header line starting with "List-" it is
3933 treated as non-personal.
3935 (4) The test for "circular" in the Subject: header line has been removed
3936 because it now seems ill-conceived.
3938 15. Minor typos in src/EDITME comments corrected.
3940 16. Installed latest exipick from John Jetmore.
3942 17. If headers_add on a router specified a text string that was too long for
3943 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3944 of string_sprintf() is now avoided.
3946 18. $message_body_size was not set (it was always zero) when running the DATA
3947 ACL and the local_scan() function.
3949 19. For the "mail" command in an Exim filter, no default was being set for
3950 the once_repeat time, causing a random time value to be used if "once" was
3951 specified. (If the value happened to be <= 0, no repeat happened.) The
3952 default is now 0s, meaning "never repeat". The "vacation" command was OK
3953 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3954 (I found it when inspecting the code).
3956 20. There is now an overall timeout for performing a callout verification. It
3957 defaults to 4 times the callout timeout, which applies to individual SMTP
3958 commands during the callout. The overall timeout applies when there is more
3959 than one host that can be tried. The timeout is checked before trying the
3960 next host. This prevents very long delays if there are a large number of
3961 hosts and all are timing out (e.g. when the network connections are timing
3962 out). The value of the overall timeout can be changed by specifying an
3963 additional sub-option for "callout", called "maxwait". For example:
3965 verify = sender/callout=5s,maxwait=20s
3967 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3968 to the end before writing, but this should make it even safer).
3970 22. Exim was forgetting that it had advertised PIPELINING for the second and
3971 subsequent messages on an SMTP connection. It was also not resetting its
3972 memory on STARTTLS and an internal HELO.
3974 23. When Exim logs an SMTP synchronization error within a session, it now
3975 records whether PIPELINING has been advertised or not.
3977 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3978 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3979 rather than long int.
3981 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3983 26. Added the never_mail option to autoreply.
3989 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3990 crash if the getsockname() call failed; this can happen if a connection is
3991 closed very soon after it is established. The problem was simply in the
3992 order in which certain operations were done, causing Exim to try to write
3993 to the SMTP stream before it had set up the file descriptor. The bug has
3994 been fixed by making things happen in the correct order.
4000 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4001 before the connection was closed, thus losing the rejection response.
4003 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4004 some early Solaris releases, but causes trouble in current releases where
4005 socklen_t is defined.
4007 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4010 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4013 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4014 that releasing the top of it at the end releases what was used for sub-
4015 expansions (unless the block got too big). However, discard this block if
4016 the first thing is a variable or header, so that we can use its block when
4017 it is dynamic (useful for very large $message_headers, for example).
4019 6. Lookups now cache *every* query, not just the most recent. A new, separate
4020 store pool is used for this. It can be recovered when all lookup caches are
4021 flushed. Lookups now release memory at the end of their result strings.
4022 This has involved some general refactoring of the lookup sources.
4024 7. Some code has been added to the store_xxx() functions to reduce the amount
4025 of flapping under certain conditions.
4027 8. log_incoming_interface used to affect only the <= reception log lines. Now
4028 it causes the local interface and port to be added to several more SMTP log
4029 lines, for example "SMTP connection from", and rejection lines.
4031 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4033 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4035 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4036 could be overwritten at the end of the current message (or the start of a
4037 new message if it was set in a HELO ACL). The value is now preserved for
4038 the duration of the SMTP connection.
4040 12. If a transport had a headers_rewrite setting, and a matching header line
4041 contained an unqualified address, that address was qualified, even if it
4042 did not match any rewriting rules. The underlying bug was that the values
4043 of the flags that permit the existence of unqualified sender and recipient
4044 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4045 non-local messages, and by -bnq for local messages) were not being
4046 preserved with the message after it was received.
4048 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4049 "next input=" as part of the text comprising the host identity instead of
4050 the correct text. The code was using the same buffer for two different
4051 strings. However, depending on which order the printing function evaluated
4052 its arguments, the bug did not always show up. Under Linux, for example, my
4053 test suite worked just fine.
4055 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4056 doesn't work with some older versions of Perl. It has been changed to "my",
4057 which in any case is probably the better facility to use.
4059 15. A really picky compiler found some instances of statements for creating
4060 error messages that either had too many or two few arguments for the format
4063 16. The size of the buffer for calls to the DNS resolver has been increased
4064 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4065 for addresses that have a lot of PTR records. This alleviates a problem; it
4066 does not fully solve it.
4068 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4069 buffer now truncates the list and logs the incident, which is the same
4070 action as happens when Exim is looking up a host name and its aliases.
4071 Previously in this situation something unpredictable would happen;
4072 sometimes it was "internal error: store_reset failed".
4074 18. If a server dropped the connection unexpectedly when an Exim client was
4075 using GnuTLS and trying to read a response, the client delivery process
4076 crashed while trying to generate an error log message.
4078 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4079 string, for example:
4081 warn message = H1: something\nH2: something
4083 the text was added as a single header line from Exim's point of view
4084 though it ended up OK in the delivered message. However, searching for the
4085 second and subsequent header lines using $h_h2: did not work. This has been
4086 fixed. Similarly, if a system filter added multiple headers in this way,
4087 the routers could not see them.
4089 20. Expanded the error message when iplsearch is called with an invalid key to
4090 suggest using net-iplsearch in a host list.
4092 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4093 ACLs are no longer actually imposed (and a message to that effect is
4096 22. If a "gecos" field in a passwd entry contained escaped characters, in
4097 particular, if it contained a \" sequence, Exim got it wrong when building
4098 a From: or a Sender: header from that name. A second bug also caused
4099 incorrect handling when an unquoted " was present following a character
4100 that needed quoting.
4102 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4103 was not being matched caselessly.
4105 24. Arranged for all hyphens in the exim.8 source to be escaped with
4108 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4109 properly. Recipient callout cache records were still being keyed to include
4110 the sender, even when use_sender was set false. This led to far more
4111 callouts that were necessary. The sender is no longer included in the key
4112 when use_sender is false.
4114 26. Added "control = submission" modifier to ACLs.
4116 27. Added the ${base62d: operator to decode base 62 numbers.
4118 28. dnsdb lookups can now access SRV records.
4120 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4121 the configuration file.
4123 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4124 (-v) mode. This makes the output for a verbose queue run more intelligible.
4126 31. Added a use_postmaster feature to recipient callouts.
4128 32. Added the $body_zerocount variable, containing the number of binary zero
4129 bytes in the message body.
4131 33. The time of last modification of the "new" subdirectory is now used as the
4132 "mailbox time last read" when there is a quota error for a maildir
4135 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4137 35. Added +ignore_unknown as a special item in host lists.
4139 36. Code for decoding IPv6 addresses in host lists is now included, even if
4140 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4141 address was recognized as an IP address, but was then not correctly decoded
4142 into binary, causing unexpected and incorrect effects when compared with
4149 1. Very minor rewording of debugging text in manualroute to say "list of
4150 hosts" instead of "hostlist".
4152 2. If verify=header_syntax was set, and a header line with an unqualified
4153 address (no domain) and a large number of spaces between the end of the
4154 name and the colon was received, the reception process suffered a buffer
4155 overflow, and (when I tested it) crashed. This was caused by some obsolete
4156 code that should have been removed. The fix is to remove it!
4158 3. When running in the test harness, delay a bit after writing a bounce
4159 message to get a bit more predictability in the log output.
4161 4. Added a call to search_tidyup() just before forking a reception process. In
4162 theory, someone could use a lookup in the expansion of smtp_accept_max_
4163 per_host which, without the tidyup, could leave open a database connection.
4165 5. Added the variables $recipient_data and $sender_data which get set from a
4166 lookup success in an ACL "recipients" or "senders" condition, or a router
4167 "senders" option, similar to $domain_data and $local_part_data.
4169 6. Moved the writing of debug_print from before to after the "senders" test
4172 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4173 problems for message scanning, either using a data ACL, or using
4174 local_scan() because the Received: header was not generated till after they
4175 were called (in order to set the time as the time of reception completion).
4176 I have revised the way this works. The header is now generated after the
4177 body is received, but before the ACL or local_scan() are called. After they
4178 are run, the timestamp in the header is updated.
4184 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4185 before starting a queue runner without re-exec. This happened only when
4186 deliver_drop_privilege was set or when the Exim user was set to root. The
4187 effect of the bug was that timeouts during subsequent deliveries caused
4188 crashes instead of being properly handled. The handler is now left at its
4189 default (and expected) setting.
4191 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4192 message, again when deliver_drop_privilege is set or Exim is run as root.
4193 The bug described in (1) was not present in this case, but the tidying up
4194 of the other signals was missing. I have made the two cases consistent.
4196 3. The ignore_target_hosts setting on a manualroute router was being ignored
4197 for hosts that were looked up using the /MX notation.
4199 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4202 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4203 operated on the sender address. After changing the $sender_address to <>
4204 for the sender address verify, Exim was re-instated it as the original
4205 (before rewriting) address, but remembering that it had rewritten it, so it
4206 wasn't rewriting it again. This bug also had the effect of breaking the
4207 sender address verification caching when the sender address was rewritten.
4209 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4210 This has been changed so that if the ip literal address matches
4211 ignore_target_hosts, the router declines.
4213 7. Added expansion conditions match_domain, match_address, and match_local_
4214 part (NOT match_host).
4216 8. The placeholder for the Received: header didn't have a length field set.
4218 9. Added code to Exim itself and to exim_lock to test for a specific race
4219 condition that could lead to file corruption when using MBX delivery. The
4220 issue is with the lockfile that is created in /tmp. If this file is removed
4221 after a process has opened it but before that process has acquired a lock,
4222 there is the potential for a second process to recreate the file and also
4223 acquire a lock. This could lead to two Exim processes writing to the file
4224 at the same time. The added code performs the same test as UW imapd; it
4225 checks after acquiring the lock that its file descriptor still refers to
4226 the same named file.
4228 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4229 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4232 11. Added the smtp_active_hostname option. If used, this will typically be made
4233 to depend on the incoming interface address. Because $interface_address is
4234 not set up until the daemon has forked a reception process, error responses
4235 that can happen earlier (such as "too many connections") no longer contain
4238 12. If an expansion in a condition on a "warn" statement fails because a lookup
4239 defers, the "warn" statement is abandoned, and the next ACL statement is
4240 processed. Previously this caused the whole ACL to be aborted.
4242 13. Added the iplsearch lookup type.
4244 14. Added ident_timeout as a log selector.
4246 15. Added tls_certificate_verified as a log selector.
4248 16. Added a global option tls_require_ciphers (compare the smtp transport
4249 option of the same name). This controls incoming TLS connections.
4251 17. I finally figured out how to make tls_require_ciphers do a similar thing
4252 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4253 before starting the TLS session.
4255 18. Tabs are now shown as \t in -bP output.
4257 19. If the log selector return_path_on_delivery was set, Exim crashed when
4258 bouncing a message because it had too many Received: header lines.
4260 20. If two routers both had headers_remove settings, and the first one included
4261 a superfluous trailing colon, the final name in the first list and the
4262 first name in the second list were incorrectly joined into one item (with a
4263 colon in the middle).
4269 1. Added -C and -D options to the exinext utility, mainly to make it easier
4270 to include in the automated testing, but these could be helpful when
4271 multiple configurations are in use.
4273 2. The exinext utility was not formatting the output nicely when there was
4274 an alternate port involved in the retry record key, nor when there was a
4275 message id as well (for retries that were specific to a specific message
4276 and a specific host). It was also confused by IPv6 addresses, because of
4277 the additional colons they contain. I have fixed the IPv4 problem, and
4278 patched it up to do a reasonable job for IPv6.
4280 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4281 delivery, the log line now contains "pipelined" if PIPELINING was used.
4283 4. An SMTP transport process used to panic and die if the bind() call to set
4284 an explicit outgoing interface failed. This has been changed; it is now
4285 treated in the same way as a connect() failure.
4287 5. A reference to $sender_host_name in the part of a conditional expansion
4288 that was being skipped was still causing a DNS lookup. This no longer
4291 6. The def: expansion condition was not recognizing references to header lines
4292 that used bh_ and bheader_.
4294 7. Added the _cache feature to named lists.
4296 8. The code for checking quota_filecount in the appendfile transport was
4297 allowing one more file than it should have been.
4299 9. For compatibility with Sendmail, the command line option
4307 and sets the incoming protocol and host name (for trusted callers). The
4308 host name and its colon can be omitted when only the protocol is to be set.
4309 Note the Exim already has two private options, -pd and -ps, that refer to
4310 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4311 "s", but I don't think that's a major issue.
4313 10. A number of refactoring changes to the code, none of which should affect
4316 (a) The number of logging options was getting close to filling up the
4317 32-bit word that was used as a bit map. I have split them into two classes:
4318 those that are passed in the argument to log_write(), and those that are
4319 only ever tested independently outside of that function. These are now in
4320 separate 32-bit words, so there is plenty of room for expansion again.
4321 There is no change in the user interface or the logging behaviour.
4323 (b) When building, for example, log lines, the code previously used a
4324 macro that called string_cat() twice, in order to add two strings. This is
4325 not really sufficiently general. Furthermore, there was one instance where
4326 it was actually wrong because one of the argument was used twice, and in
4327 one call a function was used. (As it happened, calling the function twice
4328 did not affect the overall behaviour.) The macro has been replaced by a
4329 function that can join an arbitrary number of extra strings onto a growing
4332 (c) The code for expansion conditions now uses a table and a binary chop
4333 instead of a serial search (which was left over from when there were very
4334 few conditions). Also, it now recognizes conditions like "pam" even when
4335 the relevant support is not compiled in: a suitably worded error message is
4336 given if an attempt is made to use such a condition.
4338 11. Added ${time_interval:xxxxx}.
4340 12. A bug was causing one of the ddress fields not to be passed back correctly
4341 from remote delivery subprocesses. The field in question was not being
4342 subsequently used, so this caused to problems in practice.
4344 13. Added new log selectors queue_time and deliver_time.
4346 14. Might have fixed a bug in maildirsizefile handling that threw up
4347 "unexpected character" debug warnings, and recalculated the data
4348 unnecessarily. In any case, I expanded the warning message to give more
4351 15. Added the message "Restricted characters in address" to the statements in
4352 the default ACL that block characters like @ and % in local parts.
4354 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4355 Three changes have been made:
4357 (a) There was a serious bug; a negative response to MAIL caused the whole
4358 recipient domain to be cached as invalid, thereby blocking all messages
4359 to all local parts at the same domain, from all senders. This bug has
4360 been fixed. The domain is no longer cached after a negative response to
4361 MAIL if the sender used is not empty.
4363 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4366 (c) A new callout option, "use_sender" has been added for people who want
4367 the modified behaviour.
4373 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4376 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4377 indeed breaks things for older releases.
4379 3. Added additional logging to the case where there is a problem reading data
4380 from a filter that is running in a subprocess using a pipe, in order to
4381 try to track down a specific problem.
4383 4. Testing facility fudge: when running in the test harness and attempting
4384 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4385 getting "No route to host". Convert this to a timeout.
4387 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4390 6. Some OS don't have socklen_t but use size_t instead. This affects the
4391 fifth argument of getsockopt() amongst other things. This is now
4392 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4393 can be set for individual OS. I have set it for SunOS5, OSF1, and
4394 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4395 some earlier ones do not.
4397 7. Change 4.30/15 was not doing the test caselessly.
4399 8. The standard form for an IPv6 address literal was being rejected by address
4400 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4401 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4402 this, as well as the form without the "IPv6" on the front (but only when
4403 address literals are enabled, of course).
4405 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4407 10. Exim crashed if a message with an empty sender address specified by -f
4408 encountered a router with an errors_to setting. This could be provoked only
4409 by a command such as
4413 where an empty string was supplied; "<>" did not hit this bug.
4415 11. Installed PCRE release 4.5.
4417 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4418 remained set. It is now erased.
4420 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4421 times from message ids (which are base 36 rather than the normal 62).
4423 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4424 were being counted as actual protocol errors, and logged if the log
4425 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4426 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4427 and DATA following a set of rejected RCPTs do not count as protocol errors.
4428 In other words, Exim assumes they were pipelined, though this may not
4429 actually be the case. Of course, in all cases the client gets an
4430 appropriate error code.
4432 15. If a lookup fails in an ACL condition, a message about the failure may
4433 be available; it is used if testing the ACL cannot continue, because most
4434 such messages specify what the cause of the deferral is. However, some
4435 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4436 that caused an old message to be retained and used if a later statement
4437 caused a defer, replacing the real cause of the deferral.
4439 16. If an IP address had so many PTR records that the DNS lookup buffer
4440 was not large enough to hold them, Exim could crash while trying to process
4441 the truncated data. It now detects and logs this case.
4443 17. Further to 4.21/58, another change has been made: if (and only if) the
4444 first line of a message (the first header line) ends with CRLF, a bare LF
4445 in a subsequent header line has a space inserted after it, so as not to
4446 terminate the header.
4448 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4449 unnecessarily, used atoi() instead of strtol(), and didn't check the
4450 termination when getting file sizes from file names by regex.
4452 19. Completely re-implemented the support for maildirsize files, in the light
4453 of a number of problems with the previous contributed implementation
4454 (4.30/29). In particular:
4456 . If the quota is zero, the maildirsize file is maintained, but no quota is
4459 . If the maildir directory does not exist, it is created before any attempt
4460 to write a maildirsize file.
4462 . The quota value in the file is just a cache; if the quota is changed in
4463 the transport, the new value overrides.
4465 . A regular expression is available for excluding directories from the
4468 20. The autoreply transport checks the characters in options that define the
4469 message's headers; it allows continued headers, but it was checking with
4470 isspace() after an embedded newline instead of explicitly looking for a
4473 21. If all the "regular" hosts to which an address was routed had passed their
4474 expiry times, and had not reached their retry times, the address was
4475 bounced, even if fallback hosts were defined. Now Exim should go on to try
4478 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4479 equivalent code in the SMTP transport. Some hosts send humungous responses
4480 to HELO/EHLO, more than 1024 it seems.
4482 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4483 gives compiler warnings in some environments. I've now done it "properly",
4486 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4487 (because of problems with the built-in one) was declared to return uschar *
4488 instead of char *, causing compiler failure.
4490 25. Fixed a file descriptor leak when processing alias/forward files.
4492 26. Fixed a minor format string issue in dbfn.c.
4494 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4496 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4497 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4500 29. When a sender address is verified, it is cached, to save repeating the test
4501 when there is more than one recipient in a message. However, when the
4502 verification involves a callout, it is possible for different callout
4503 options to be set for different recipients. It is too complicated to keep
4504 track of this in the cache, so now Exim always runs a verification when a
4505 callout is required, relying on the callout cache for the optimization.
4506 The overhead is duplication of the address routing, but this should not be
4509 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4510 to be verified with callout=postmaster, and the main callout worked but the
4511 postmaster check failed, the verification correctly failed. However, if a
4512 subsequent RCPT command asked for sender verification *without* the
4513 postmaster check, incorrect caching caused this verification also to fail,
4516 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4517 it was not caching the DNS options (qualify_single, search_parents) that
4518 were used when the lookup failed. A subsequent lookup with different
4519 options therefore always gave the same answer, though there were cases
4520 where it should not have. (Example: a "domains = !$mx_any" option on a
4521 dnslookup router: the "domains" option is always processed without any
4522 widening, but the router might have qualify_single set.) Now Exim uses the
4523 cached value only when the same options are set.
4525 32. Added John Jetmore's "exipick" utility to the distribution.
4527 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4528 than a timeout (e.g. a certificate is required, and is not provided), an
4529 Exim server now closes the connection immediately. Previously it waited for
4530 the client to close - but if the client is SSL, it seems that they each
4531 wait for each other, leading to a delay before one of them times out.
4533 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4534 maintained 0.8.x compatibility because I don't think many are using it, and
4535 it is clearly obsolete.
4537 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4540 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4541 client certificate was expired. A simple patch fixes this, though I don't
4542 understand the full logic of why the verify callback is called multiple
4545 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4546 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4547 which causes problems with some clients (such as the Certicom SSL Plus
4548 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4549 disables the coutermeasure allowing Eudora to connect."
4551 38. Exim was not checking that a write() to a log file succeeded. This could
4552 lead to Bad Things if a log got too big, in particular if it hit a file
4553 size limit. Exim now panics and dies if it cannot write to a log file, just
4554 as it does if it cannot open a log file.
4556 39. Modified OS/Makefile-Linux so that it now contains
4558 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4560 The two -D definitions ensure that Exim is compiled with large file
4561 support, which makes it possible to handle log files that are bigger than
4564 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4565 instance) a domain was checked against a named list that involved a lookup,
4566 causing $domain_data to be set, then another domain was checked against the
4567 same list, then the first domain was re-checked, the value of $domain_data
4568 after the final check could be wrong. In particular, if the second check
4569 failed, it could be set empty. This bug probably also applied to
4572 41. The strip_trailing_dot option was not being applied to the address given
4573 with the -f command-line option.
4575 42. The code for reading a message's header from the spool was incrementing
4576 $received_count, but never initializing it. This meant that the value was
4577 incorrect (doubled) while delivering a message in the same process in which
4578 it was received. In the most common configuration of Exim, this never
4579 happens - a fresh exec is done - but it can happen when
4580 deliver_drop_privilege is set.
4582 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4583 it now includes up to 150 characters of the unexpected data in the log
4586 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4587 and building data strings. The size of both of these buffers was 10 000
4588 bytes - far larger than anybody would *ever* want, thought I. Needless to
4589 say, somebody hit the limit. I have increased the maximum line length to
4590 20 000 and the maximum data length of concatenated lines to 100 000. I have
4591 also fixed two bugs, because there was no checking on these buffers. Tsk,
4592 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4593 buffer is too small.
4595 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4596 lsearch lookups. Now it does.
4598 46. When parsing a route_list item in a manualroute router, a fixed-length
4599 buffer was used for the list of hosts. I made this 1024 bytes long,
4600 thinking that nobody would ever have a list of hosts that long. Wrong.
4601 Somebody had a whole pile of complicated expansion conditions, and the
4602 string was silently truncated, leading to an expansion error. It turns out
4603 that it is easier to change to an unlimited length (owing to other changes
4604 that have happened since this code was originally written) than to build
4605 structure for giving a limitation error. The length of the item that
4606 expands into the list of hosts is now unlimited.
4608 47. The lsearch lookup could not handle data where the length of text line was
4609 more than 4095 characters. Such lines were truncated, leading to shortened
4610 data being returned. It should now handle lines of any length.
4612 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4613 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4616 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4617 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4618 no longer happens in comment lines. A list of replacements is now placed
4619 at the head of all of the source files, except those whose only change is
4620 to replace PERL_COMMAND in the very first #! line.
4622 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4623 messages on the queue, with a bottom-up merge sort, using code contributed
4624 by Michael Haardt. This should make operations like -bp somewhat faster on
4625 large queues. It won't affect queue runners, except when queue_run_in_order
4628 51. Installed eximstats 1.31 in the distribution.
4630 52. Added support for SRV lookups to the dnslookup router.
4632 53. If an ACL referred to $message_body or $message_body_end, the value was not
4633 reset for any messages that followed in the same SMTP session.
4635 54. The store-handling optimization for building very long strings was not
4636 differentiating between the different store pools. I don't think this
4637 actually made any difference in practice, but I've tidied it.
4639 55. While running the routers to verify a sender address, $sender_address
4640 was still set to the sender address. This is wrong, because when routing to
4641 send a bounce to the sender, it would be empty. Therefore, I have changed
4642 it so that, while verifying a sender address, $sender_address is set to <>.
4643 (There is no change to what happens when verifying a recipient address.)
4645 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4646 target A or AAAA records (if not already returned) without resetting the
4647 qualify_single or search_parents options of the DNS resolver. These are
4648 inappropriate in this case because the targets of MX and SRV records must
4649 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4650 target that, when qualified, matched something in the local domain. These
4651 two options are now turned off when doing these lookups.
4653 57. It seems that at least some releases of Reiserfs (which does not have the
4654 concept of a fixed number of inodes) returns zero and not -1 for the
4655 number of available inodes. This interacted badly with check_spool_inodes,
4656 which assumed that -1 was the "no such thing" setting. What I have done is
4657 to check that the total number of inodes is greater than zero before doing
4658 the test of how many are available.
4660 58. When a "warn" ACL statement has a log_message modifier, the message is
4661 remembered, and not repeated. This is to avoid a lot of repetition when a
4662 message has many recipients that cause the same warning to be written.
4663 Howewer, Exim was preserving the list of already written lines for an
4664 entire SMTP session, which doesn't seem right. The memory is now reset if a
4665 new message is started.
4667 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4668 debugging output unless log_rewrite was also set.
4670 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4671 of (int)(handle) when we know that handle contains (void *)(-1).
4673 61. The Exim daemon panic-logs an error return when it closes the incoming
4674 connection. However "connection reset by peer" seems to be common, and
4675 isn't really an error worthy of noting specially, so that particular error
4678 62. When Exim is trying to find all the local interfaces, it used to panic and
4679 die if the ioctl to get the interface flags failed. However, it seems that
4680 on at least one OS (Solaris 9) it is possible to have an interface that is
4681 included in the list of interfaces, but for which you get a failure error
4682 for this call. This happens when the interface is not "plumbed" into a
4683 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4684 failure of the "get flags" call assumes that the interface is down.
4686 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4687 makes life easier for people who are doing arithmetic on fields extracted
4688 from dates, where you often get leading zeros that should not be
4689 interpreted as octal.
4691 64. Added qualify_domain to the redirect router, to override the global
4694 65. If a pathologically long header line contained very many addresses (the
4695 report of this problem mentioned 10 000) and each of them was rewritten,
4696 Exim could use up a very large amount of memory. (It kept on making new
4697 copies of the header line as it rewrote, and never released the old ones.)
4698 At the expense of a bit more processing, the header rewriting function has
4699 been changed so that it no longer eats memory in this way.
4701 66. The generation of the Received: header has been moved from the time that a
4702 message starts to be received, to the time that it finishes. The timestamp
4703 in the Received: header should now be very close to that of the <= log
4704 line. There are two side-effects of this change:
4706 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4707 logged header lines no longer include the local Received: line, because
4708 it has not yet been created. The same applies to a copy of the message
4709 that is returned to a non-SMTP sender when a message is rejected.
4711 (b) When a filter file is tested using -bf, no additional Received: header
4712 is added to the test message. After some thought, I decided that this
4715 This change does not affect the value of $received_for. It is still set
4716 after address rewriting, but before local_scan() is called.
4718 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4720 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4721 gave an unhelpful panic error message, and a defer error. I have managed to
4722 change this behaviour so that it now rejects any supplied certificate,
4723 which seems right, as the list of acceptable certificates is empty.
4725 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4726 gave an unhelpful defer error. I have not managed to make this reject any
4727 supplied certificates, but the error message it gives is "no certificate
4728 supplied", which is not helpful.
4730 70. exigrep's output now also includes lines that are not associated with any
4731 message, but which match the given pattern. Implemented by a patch from
4732 Martin Sluka, which also tidied up the Perl a bit.
4734 71. Recipient callout verification, like sender verification, was using <> in
4735 the MAIL FROM command. This isn't really the right thing, since the actual
4736 sender may affect whether the remote host accepts the recipient or not. I
4737 have changed it to use the actual sender in the callout; this means that
4738 the cache record is now keyed on a recipient/sender pair, not just the
4739 recipient address. There doesn't seem to be a real danger of callout loops,
4740 since a callout by the remote host to check the sender would use <>.
4741 [SEE ABOVE: changed after hitting problems.]
4743 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4744 temporary errors. However, in the case of such a code being given after
4745 the end of a data transmission (i.e. after ".") Exim was failing to write
4746 a retry record for the message. (Yes, there was some broken host that was
4747 actually sending 8xx at this point.)
4749 73. An unknown lookup type in a host list could cause Exim to panic-die when
4750 the list was checked. (An example that provoked this was putting <; in the
4751 middle of a list instead of at the start.) If this happened during a DATA
4752 ACL check, a -D file could be left lying around. This kind of configuration
4753 error no longer causes Exim to die; instead it causes a defer errror. The
4754 incident is still logged to the main and panic logs.
4756 74. Buglet left over from Exim 3 conversion. The message "too many messages
4757 in one connection" was written to the rejectlog but not the mainlog, except
4758 when address rewriting (yes!) was being logged.
4760 75. Added write_rejectlog option.
4762 76. When a system filter was run not as root (that is, when system_filter_user
4763 was set), the values of the $n variables were not being returned to the
4764 main process; thus, they were not subsequently available in the $sn
4767 77. Added +return_path_on_delivery log selector.
4769 78. A connection timeout was being treated differently from recipients deferred
4770 when testing hosts_max_try with a message that was older than the host's
4771 retry timeout. (The host should not be counted, thus allowing all hosts to
4772 be tried at least once before bouncing.) This may have been the cause of an
4773 occasionally reported bug whereby a message would remain on the queue
4774 longer than the retry timeout, but would be bounced if a delivery was
4775 forced. I say "may" because I never totally pinned down the problem;
4776 setting up timeout/retry tests is difficult. See also the next item.
4778 79. The ultimate address timeout was not being applied to errors that involved
4779 a combination of host plus message (for example, a timeout on a MAIL
4780 command). When an address resolved to a number of possible hosts, and they
4781 were not all tried for each delivery (e.g. because of hosts_max_try), a
4782 message could remain on the queue longer than the retry timeout.
4784 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4787 81. Fixed an obscure SMTP outgoing bug which required at least the following
4788 conditions: (a) there was another message waiting for the same server;
4789 (b) the server returned 5xx to all RCPT commands in the first message so
4790 that the message was not completed; (c) the server dropped the connection
4791 or gave a negative response to the RSET that Exim sends to abort the
4792 transaction. The observed case was a dropped connection after DATA that had
4793 been sent in pipelining mode. That is, the server had advertised PIPELINING
4794 but was not implementing it correctly. The effect of the bug was incorrect
4795 behaviour, such as trying another host, and this could lead to a crash.
4801 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4802 and daemon.c were passed as pointers to ints; they should have been
4803 pointers to socklen_t variables (which are typically unsigned ints).
4805 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4808 3. Fixed a really odd bug that affected only the testing scheme; patching a
4809 certain fixed string in the binary changed the value of another string that
4810 happened to be identical to the end of the original first string.
4812 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4813 name", it returns that address as the IP address. On some operating
4814 systems (e.g. Solaris), it also passes back the IP address string as the
4815 "host name". However, on others (e.g. Linux), it passes back an empty
4816 string. Exim wasn't checking for this, and was changing the host name to an
4817 empty string, assuming it had been canonicized.
4819 5. Although rare, it is permitted to have more than one PTR record for a given
4820 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4821 all the names associated with an address, because they do in Solaris.
4822 However, it seems that they do not in Linux for data that comes from the
4823 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4824 I found this out when I moved to a new Linux workstation and tried to run
4825 the Exim test suite.
4827 To get round this problem I have changed the code so that it now does its
4828 own call to the DNS to look up PTR records when searching for a host name.
4829 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4830 addresses that are only in /etc/hosts are still found.
4832 This behaviour is, however, controlled by an option called host_lookup_
4833 order, which defaults to "bydns:byaddr". If people want to use the other
4834 order, or indeed, just use one or the other means of lookup, they can
4835 specify it in this variable.
4837 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4838 some operating systems, this comes back from gethostbyaddr() as an empty
4839 string, and this is what Exim used to test for. However, it seems that in
4840 other systems, "." is yielded. Exim now tests for this case too.
4842 7. The values of check_spool_space and check_log_space are now held internally
4843 as a number of kilobytes instead of an absolute number of bytes. If a
4844 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4845 kilobyte. This means that much larger values can be stored.
4847 8. Exim monitor: an attempt to get the action menu when not actually pointing
4848 at a message produces an empty menu entitled "No message selected". This
4849 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4850 no entries in it ("Shell widget menu has zero width and/or height"). So I
4851 have added a single, blank menu entry in this case.
4853 9. Added ${quote_local_part.
4855 10. MIME decoding is now applied to the contents of Subject: header lines when
4858 11. Now that a reference to $sender_host_address automatically causes a reverse
4859 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4860 host lookup before query-style lookups in lists that might use this
4861 variable. This has therefore been abolished, and the "net-" prefix is no
4862 longer necessary for query-style lookups.
4864 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4865 have been a typo for LFLAGS, so it has been changed.
4867 13. The install script calls Exim with "-C /dev/null" in order to find the
4868 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4869 to be output. Howeve, since Exim outputs its version number before the
4870 error, it didn't break the script. It just looked ugly. I fixed this by
4871 always allowing "-C /dev/null" if the caller is root.
4873 14. Ignore overlarge ACL variable number when reading spool file - insurance
4874 against a later release with more variables having written the file.
4876 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4877 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4878 this, as well as the form without the "IPv6" on the front.
4880 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4881 OS/Makefile-Darwin file.
4883 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4884 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4886 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4887 variable. However, some OS use ints for inodes. I've added cast to long int
4888 to get rid of the compiler warning.
4890 19. I had forgotten to lock out "/../" in configuration file names when
4891 ALT_CONFIG_PREFIX was set.
4893 20. Routers used for verification do not need to specify transports. However,
4894 if such a router generated a host list, and callout was configured, Exim
4895 crashed, because it could not find a port number from the (non-existent)
4896 transport. It now assumes port 25 in this circumstance.
4898 21. Added the -t option to exigrep.
4900 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4901 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4902 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4903 them is set, LOOKUP_LSEARCH is forced.
4905 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4907 24. Added sender and host information to the "rejected by local_scan()" log
4908 line; previously there was no indication of these.
4910 25. Added .include_if_exists.
4912 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4913 Linux. It turns out that not all file systems support this. Apparently some
4914 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4915 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4916 not supported on the file descriptor, is now ignored when Exim is trying to
4917 sync a directory. This applies only to Linux.
4919 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4921 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4922 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4925 29. Added support for maildirsize files from supplied patch (modified a bit).
4927 30. The use of :fail: followed by an empty string could lead Exim to respond to
4928 sender verification failures with (e.g.):
4930 550 Verification failed for <xxx>
4931 550 Sender verify failed
4933 where the first response line was missing the '-' that indicates it is not
4934 the final line of the response.
4936 31. The loop for finding the name of the user that called Exim had a hardwired
4937 limit of 10; it now uses the value of finduser_retries, which is used for
4938 all other user lookups.
4940 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4943 33. Exim was neglecting to zero errno before one call of strtol() when
4944 expanding a string and expecting an integer value. On some systems this
4945 resulted in spurious "integer overflow" errors. Also, it was casting the
4946 result into an int without checking.
4948 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4949 did not work. The code looks as if it has *never* worked, though it appears
4950 to have been documented since at least releast 1.62. I have made it work.
4952 35. The "timeout_DNS" error in retry rules, also documented since at least
4953 1.62, also never worked. As it isn't clear exactly what this means, and
4954 clearly it isn't a major issue, I have abolished the feature by treating it
4955 as "timeout", and writing a warning to the main and panic logs.
4957 36. The display of retry rules for -brt wasn't always showing the error code
4960 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4961 timeout_connect_A, timeout_connect_MX.
4963 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4964 to the empty sender.
4966 39. The daemon was not analysing the content of -oX till after it had closed
4967 stderr and disconnected from the controlling terminal. This meant that any
4968 syntax errors were only noted on the panic log, and the return code from
4969 the command was 0. By re-arranging the code a little, I've made the
4970 decoding happen first, so such errors now appear on stderr, and the return
4971 code is 1. However, the actual setting up of the sockets still happens in
4972 the disconnected process, so errors there are still only recorded on the
4975 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4976 connections (as happens on some IP stacks) was logged at start up time as
4977 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4978 it from "IPv6 and IPv4", which means that two separate sockets are being
4981 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4982 says whether AF_INET or AF_INET6 was passed as an argument.
4984 42. Exiwhat output was messed up when time zones were included in log
4987 43. Exiwhat now gives more information about the daemon's listening ports,
4988 and whether -tls-on-connect was used.
4990 44. The "port" option of the smtp transport is now expanded.
4992 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4993 silently ignored. Now an error message is written to the main and panic
4996 46. There's a new ACL modifier called "logwrite" which writes to a log file
4997 as soon as it is encountered.
4999 47. Added $local_user_uid and $local_user_gid at routing time.
5001 48. Exim crashed when trying to verify a sender address that was being
5004 49. Exim was recognizing only a space character after ".include". It now also
5005 recognizes a tab character.
5007 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5008 extracting the relevant information from the specification. The man page no
5009 longer contains scrambled data for the -d option, and I've added a section
5010 at the front about calling Exim under different names.
5012 51. Added "extra_headers" argument to the "mail" command in filter files.
5014 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5017 53. Installed eximstats 1.29.
5019 54. Added transport_filter_timeout as a generic transport option.
5021 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5022 Cc: header lines. This was required by RFC 822, but it not required by RFC
5025 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5026 incoming messages that did not have them. Now it does so only if the
5027 message originates locally, that is, if there is no associated remote host
5028 address. When Resent- header lines are present, this applies to the Resent-
5029 lines rather than the non-Resent- lines.
5031 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5032 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5034 58. Messages for configuration errors now include the name of the main
5035 configuration file - useful now that there may be more than one file in a
5036 list (.included file names were always shown).
5038 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5039 for those rare installations that do not start the daemon as root or run it
5040 setuid root. I've cut out the call to initgroups() if the daemon is not
5043 60. The Exim user and group can now be bound into the binary as text strings
5044 that are looked up at the start of Exim's processing.
5046 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5048 62. Added $mailstore_basename variable.
5050 63. Installed patch to sieve.c from Michael Haardt.
5052 64. When Exim failed to open the panic log after failing to open the main log,
5053 the original message it was trying to log was written to stderr and debug
5054 output, but if they were not available (the usual case in production), it
5055 was lost. Now it is written to syslog before the two lines that record the
5056 failures to open the logs.
5058 65. Users' Exim filters run in subprocesses under the user's uid. It is
5059 possible for a "deliver" command or an alias in a "personal" command to
5060 provoke an address rewrite. If logging of address rewriting is configured,
5061 this fails because the process is not running as root or exim. There may be
5062 a better way of dealing with this, but for the moment (because 4.30 needs
5063 to be released), I have disabled address rewrite logging when running a
5064 filter in a non-root, non-exim process.
5070 1. The buildconfig auxiliary program wasn't quoting the value set for
5071 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5072 not defined. This bug was masked in 4.22 by the effect that was fixed in
5075 2. Some messages that were rejected after a message id was allocated were
5076 shown as "incomplete" by exigrep. It no longer does this for messages that
5077 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5079 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5080 have allow_domain_literals set, the ID did not get logged in the <= line.
5081 Domain literals are now always recognized in Message-ID: header lines.
5083 4. The first argument for a ${extract expansion item is the key name or field
5084 number. Leading and trailing spaces in this item were not being ignored,
5085 causing some misleading effects.
5087 5. When deliver_drop_privilege was set, single queue runner processes started
5088 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5089 same command in the process it spins off) were not dropping privilege.
5091 6. When the daemon running as "exim" started a queue runner, it always
5092 re-executed Exim in the spun-off process. This is a waste of effort when
5093 deliver_drop_privilege is set. The new process now just calls the
5094 queue-runner function directly.
5100 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5103 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5104 ignored. Though the use of -oP was forcing the writing of a pid file, it
5105 was always written to the default place.
5107 3. If the message "no IP address found for host xxxx" is generated during
5108 incoming verification, it is now followed by identification of the incoming
5109 connection (so you can more easily find what provoked it).
5111 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5113 5. Added some features to "harden" Exim a bit more against certain attacks:
5115 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5116 be put in Local/Makefile. This is like the never_users runtime option,
5117 but it cannot be overridden. The default setting is "root".
5119 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5120 prefix string with which any file named in a -C command line option
5123 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5124 is retained for -C and -D only if the caller of Exim is root. Without
5125 it, the exim user may also use -C and -D and retain privilege.
5127 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5128 command line option is disabled.
5130 6. Macro names set by the -D option must start with an upper case letter, just
5131 like macro names defined in the configuration file.
5133 7. Added "dereference=" facility to LDAP.
5135 8. Two instances of the typo "uknown" in the source files are fixed.
5137 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5138 the Configure-Makefile script screwed up while processing it.
5140 10. Incorporated PCRE 4.4.
5142 11. The SMTP synchronization check was not operating right at the start of an
5143 SMTP session. For example, it could not catch a HELO sent before the client
5144 waited for the greeting. There is now a check for outstanding input at the
5145 point when the greeting is written. Because of the duplex, asynchronous
5146 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5147 way, but not yet received, when the check is performed.
5149 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5150 on TCP/IP sockets, because this apparently causes some broken clients to
5153 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5154 unchanged) from the Cygwin maintainer.
5156 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5157 was supported instead of testing for mbx. Effectively a typo.
5159 15. The spa authenticator server code was not checking that the input it
5160 received was valid base64.
5162 16. The debug output line for the "set" modifier in ACLs was not showing the
5163 name of the variable that was being set.
5165 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5167 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5168 lookup on its own. Something else was needed to trigger the lookup. For
5169 example, a match in host_lookup or the need for a host name in a host list.
5170 Now, if $sender_host_name is referenced and the host name has not yet been
5171 looked up, a lookup is performed. If the lookup fails, the variable remains
5172 empty, and $host_lookup_failed is set to "1".
5174 19. Added "eqi" as a case-independent comparison operator.
5176 20. The saslauthd authentication condition could segfault if neither service
5177 nor realm was specified.
5179 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5180 error message that was logged was misleading, and incoming SMTP
5181 connections were dropped. The message is now more accurate, and temporary
5182 errors are given to SMTP connections.
5184 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5185 (except RSET) with a 421 error, until QUIT is received. However, it was
5186 failing to send a response to QUIT.
5188 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5189 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5190 ACL is now run after the helo_try_verify_hosts code.
5192 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5193 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5194 case-independent, so other case variants are also recognized). Apparently
5195 some systems use these upper case variants.
5197 25. If more than two messages were waiting for the same host, and a transport
5198 filter was specified for the transport, Exim sent two messages over the
5199 same TCP/IP connection, and then failed with "socket operation on non-
5200 socket" when it tried to send the third.
5202 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5204 27. The extern definition of crypt16() in expand.c was not being excluded when
5205 the OS had its own crypt16() function.
5207 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5208 as a preferred synonym for return_size_limit, both as an option and as an
5211 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5213 30. Changed the default configuration ACL to relax the local part checking rule
5214 for addresses that are not in any local domains. For these addresses,
5215 slashes and pipe symbols are allowed within local parts, but the sequence
5216 /../ is explicitly forbidden.
5218 31. SPA server authentication was not clearing the challenge buffer before
5221 32. log_message in a "warn" ACL statement was writing to the reject log as
5222 well as to the main log, which contradicts the documentation and doesn't
5223 seem right (because no rejection is happening). So I have stopped it.
5225 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5226 However, I am unable to do any testing of this.
5228 34. Fixed an infelicity in the appendfile transport. When checking directories
5229 for a mailbox, to see if any needed to be created, it was accidentally
5230 using path names with one or more superfluous leading slashes; tracing
5231 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5233 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5234 contents are added to the log line that is written for every discarded
5235 recipient. (Previously a log_message setting was ignored.)
5237 36. The ${quote: operator now quotes the string if it is empty.
5239 37. The install script runs exim in order to find its version number. If for
5240 some reason other than non-existence or emptiness, which it checks, it
5241 could not run './exim', it was installing it with an empty version number,
5242 i.e. as "exim-". This error state is now caught, and the installation is
5245 38. An argument was missing from the function that creates an error message
5246 when Exim fails to connect to the socket for saslauthd authentication.
5247 This could cause Exim to crash, or give a corrupted message.
5249 39. Added isip, isip4, and isip6 to ${if conditions.
5251 40. The ACL variables $acl_xx are now saved with the message, and can be
5252 accessed later in routers, transports, and filters.
5254 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5255 strings in the file are not string-expanded.
5257 42. If a MAIL command specified a SIZE value that was too large to fit into an
5258 int variable, the check against message_size_limit failed. Such values are
5259 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5260 day this will have to be increased, but I don't think I want to be around
5261 when emails are that large.
5268 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5269 iconv() is not standard in FreeBSD.
5271 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5272 IPv6 enabled. The observed symptom was a segmentation fault on return from
5273 the function os_common_find_running_interfaces() in src/os.c.
5275 3. In the check_special_case() function in daemon.c I had used "errno" as an
5276 argument name, which causes warnings on some systems. This was basically a
5277 typo, since it was named "eno" in the comments!
5279 4. The code that waits for the clock to tick (at a resolution of some fraction
5280 of a second) so as to ensure message-id uniqueness was always waiting for
5281 at least one whole tick, when it could have waited for less. [This is
5282 almost certainly not relevant at current processor speeds, where it is
5283 unlikely to ever wait at all. But we try to future-proof.]
5285 5. The function that sleeps for a time interval that includes fractions of a
5286 second contained a race. It did not block SIGALRM between setting the
5287 timer, and suspending (a couple of lines later). If the interval was short
5288 and the sigsuspend() was delayed until after it had expired, the suspension
5289 never ended. On busy systems this could lead to processes getting stuck for
5292 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5293 process, before it forks any delivery processes. The open lookup caching
5294 mechanism meant that the open file or database connection was passed into
5295 the delivery process. The problem was that delivery processes always tidy
5296 up cached lookup data. This could cause a problem for the next delivery
5297 process started by the queue runner, because the external queue runner
5298 process does not know about the closure. So the next delivery process
5299 still has data in the lookup cache. In the case of a file lookup, there was
5300 no problem because closing a file descriptor in a subprocess doesn't affect
5301 the parent. However, if the lookup was caching a connection to a database,
5302 the connection was closed, and the second delivery process was likely to
5303 see errors such as "PGSQL: query failed: server closed the connection
5304 unexpectedly". The problem has been fixed by closing all cached lookups
5305 in a queue runner before running a delivery process.
5307 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5308 seem to have the "const" qualifier which it has on other OS. I've
5311 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5312 *of the same type* (client or server) with the same public name that an
5313 error should be diagnosed.
5315 9. When Exim looked up a host name for an IP address, but failed to find the
5316 original IP address when looking up the host name (a safety check), it
5317 output the message "<ip address> does not match any IP for NULL", which was
5318 confusing, to say the least. The bug was that the host name should have
5319 appeared instead of "NULL".
5321 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5322 user that is built into the binary, and the -C or -D options is used, root
5323 privilege is dropped before the configuration file is read. In addition,
5324 logging is switched to stderr instead of the normal log files. If the
5325 configuration then re-defines the Exim user, the unprivileged environment
5326 is probably not what is expected, so Exim logs a panic warning message (but
5329 However, if deliver_drop_privilege is set, the unprivileged state may well
5330 be exactly what is intended, so the warning has been cut out in that case,
5331 and Exim is allowed to try to write to its normal log files.
5337 1. smtp_return_error_details was not giving details for temporary sender
5338 or receiver verification errors.
5340 2. Diagnose a configuration error if two authenticators have the same public
5343 3. Exim used not to create the message log file for a message until the first
5344 delivery attempt. This could be confusing when incoming messages were held
5345 for policy or load reasons. The message log file is now created at the time
5346 the message is received, and an initial "Received" line is written to it.
5348 4. The automatically generated man page for command line options had a minor
5349 bug that caused no ill effects; however, a more serious problem was that
5350 the procedure for building the man page automatically didn't always
5351 operate. Consequently, release 4.20 contains an out-of-date version. This
5352 shouldn't happen again.
5354 5. When building Exim with embedded Perl support, the script that builds the
5355 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5356 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5358 6. The freeze_tell option was not being used for messages that were frozen on
5359 arrival, either by an ACL or by local_scan().
5361 7. Added the smtp_incomplete_transaction log selector.
5363 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5364 was accepting AUTH without a new EHLO.
5366 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5367 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5370 10. Logging of TCP/IP connections (when configured) now happens in the main
5371 daemon process instead of the child process, so that the TCP/IP connection
5372 count is more accurate (but it can never be perfect).
5374 11. The use of "drop" in a nested ACL was not being handled correctly in the
5375 outer ACL. Now, if condition failure induced by the nested "drop" causes
5376 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5377 or "require"), the connection is dropped.
5379 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5380 that yield "discard" can now be used with an "accept" or a "discard" verb,
5381 but an error is generated for any others (because I can't see a useful way
5382 to define what should happen).
5384 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5385 are now processed in the same way as lines in the Exim configuration file.
5386 In particular, continuation lines are supported.
5388 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5390 15. Added -ti meaning -t -i.
5392 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5393 domains, and warn by logging if others are found.
5395 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5396 returned from GIFCONF when Exim is trying to find the list of interfaces on
5397 a host. The code in os.c has been modified to copy each ifreq to an aligned
5398 structure in all cases.
5400 Also, in some cases, the returned ifreq's were being copied to a 'struct
5401 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5402 means the last couple of bytes of an IPv6 address could be chopped if the
5403 ifreq contained only a normal sockaddr (14 bytes storage).
5405 18. Named domain lists were not supported in the hosts_treat_as_local option.
5406 An entry such as +xxxx was not recognized, and was treated as a literal
5409 19. Ensure that header lines added by a DATA ACL are included in the reject log
5410 if the ACL subsequently rejects the message.
5412 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5413 MD5 (which is deprecated).
5415 21. When testing a filter file using -bf, Exim was writing a message when it
5416 took the sender from a "From " line in the message, but it was not doing so
5417 when it took $return_path from a Return-Path: header line. It now does.
5419 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5420 with a valid header line field name (a series of printing characters
5421 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5423 23. Changed "disc" in the source to "disk" to conform to the documentation and
5424 the book and for uniformity.
5426 24. Ignore Sendmail's -Ooption=value command line item.
5428 25. When execve() failed while trying to run a command in a pipe transport,
5429 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5430 could be confused with a return value of 69 from the command itself. This
5431 has been changed to 127, the value the shell returns if it is asked to run
5432 a non-existent command. The wording for the related log line suggests a
5433 non-existent command as the problem.
5435 26. If received_header_text expands to an empty string, do not add a Received:
5436 header line to the message. (Well, it adds a token one on the spool, but
5437 marks it "old" so that it doesn't get used or transmitted.)
5439 27. Installed eximstats 1.28 (addition of -nt option).
5441 28. There was no check for failure on the call to getsockname() in the daemon
5442 code. This can fail if there is a shortage of resources on the system, with
5443 ENOMEM, for example. A temporary error is now given on failure.
5445 29. Contrary to the C standard, it seems that in some environments, the
5446 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5447 program. Exim now does this explicitly; it affects the formatting of
5448 timestamps using strftime().
5450 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5451 complaints. I've now initialized all the variables, to avoid this.
5453 32. Header lines added by a system filter were not being "seen" during
5454 transport-time rewrites.
5456 33. The info_callback() function passed to OpenSSL is set up with type void
5457 (*)(SSL *, int, int), as described somewhere. However, when calling the
5458 function (actually a macro) that sets it up, the type void(*)() is
5459 expected. I've put in a cast to prevent warnings from picky compilers.
5461 34. If a DNS black list lookup found a CNAME record, but there were no A
5462 records associated with the domain it pointed at, Exim crashed.
5464 35. If a DNS black list lookup returned more than one A record, Exim ignored
5465 all but the first. It now scans all returned addresses if a particular IP
5466 value is being sought. In this situation, the contents of the
5467 $dnslist_value variable are a list of all the addresses, separated by a
5470 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5471 to accept a host name and all its aliases if the forward lookup for any of
5472 them yielded the IP address of the incoming connection. Now it accepts only
5473 those names whose forward lookup yields the correct IP address. Any other
5474 names are discarded. This closes a loophole whereby a rogue DNS
5475 administrator could create reverse DNS records to break through a
5476 wildcarded host restriction in an ACL.
5478 37. If a user filter or a system filter that ran in a subprocess used any of
5479 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5480 the wrong values were passed to the pipe command ($thisaddress had the
5481 value of $0, $0 had the value of $1, etc). This bug was introduced by
5482 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5484 38. Improved the line breaking for long SMTP error messages from ACLs.
5485 Previously, if there was no break point between 40 and 75 characters, Exim
5486 left the rest of the message alone. Two changes have been made: (a) I've
5487 reduced the minimum length to 35 characters; (b) if it can't find a break
5488 point between 35 and 75 characters, it looks ahead and uses the first one
5489 that it finds. This may give the occasional overlong line, but at least the
5490 remaining text gets split now.
5492 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5493 file descriptors might be low, and that setting 1000 would always raise it.
5494 It turns out that in some environments, the limit is already over 1000 and
5495 that lowering it causes trouble. So now Exim takes care not to decrease it.
5497 40. When delivering a message, the value of $return_path is set to $sender_
5498 address at the start of routing (routers may change the value). By an
5499 oversight, this default was not being set up when an address was tested by
5500 -bt or -bv, which affected the outcome if any router or filter referred to
5503 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5504 log only when "message" or "log_message" are set. However, if one of the
5505 conditions was an address verification, or a call to a nested ACL, the
5506 messages generated by the underlying test were being passed through. This
5507 no longer happens. The underlying message is available in $acl_verify_
5508 message for both "message" and "log_message" expansions, so it can be
5509 passed through if needed.
5511 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5512 new expansion $bh_ to give the encoded byte string without charset
5513 translation. Translation happens only if iconv() is available; HAVE_ICONV
5514 indicates this at build time. HEADERS_CHARSET gives the charset to
5515 translate to; headers_charset can change it in the configuration, and
5516 "headers charset" can change it in an individual filter file.
5518 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5519 that creates RFC 2047 encoded "words" labels them as that charset instead
5520 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5521 expansion operator; (ii) when Exim creates a From: line for a local
5522 message; (iii) when a header line is rewritten to include a "phrase" part.
5524 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5525 buggy, causing it to skip the first lines of messages whose message ID
5526 ended in 'D'. This would not have bitten before Exim release 4.14, because
5527 message IDs were unlikely to end in 'D' before then. The effect was to have
5528 incorrect size information for certain domains.
5530 45. #include "config.h" was missing at the start of the crypt16.c module. This
5531 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5534 46. If there was a timeout during a "random" callout check, Exim treated it as
5535 a failure of the random address, and carried on sending RSET and the real
5536 address. If the delay was just some slowness somewhere, the response to the
5537 original RCPT would be taken as a response to RSET and so on, causing
5538 mayhem of various kinds.
5540 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5541 when I implemented it. It didn't allow for the fact that some option values
5542 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5543 the right test for positive values.
5545 48. Domain names in DNS records are case-independent. Exim always looks them up
5546 in lower case. Some resolvers return domain names in exactly the case they
5547 appear in the zone file, that is, they may contain uppercase letters. Not
5548 all resolvers do this - some return always lower case. Exim was treating a
5549 change of case by a resolver as a change of domain, similar to a widening
5550 of a domain abbreviation. This triggered its re-routing code and so it was
5551 trying to route what was effectively the same domain again. This normally
5552 caused routing to fail (because the router wouldn't handle the domain
5553 twice). Now Exim checks for this case specially, and just changes the
5554 casing of the domain that it ultimately uses when it transmits the message
5557 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5560 50. If a filter generated a file delivery with a non-absolute name (possible if
5561 no home directory exists for the router), the forbid_file option was not
5564 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5565 the existing equality matching.
5567 52. Exim was using ints instead of ino_t variables in some places where it was
5568 dealing with inode numbers.
5570 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5571 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5572 called TMPDIR, and if it finds it is different, it changes its value.
5574 54. The smtp_printf() function is now made available to local_scan() so
5575 additional output lines can be written before returning. There is also an
5576 smtp_fflush() function to enable the detection of a dropped connection.
5577 The variables smtp_input and smtp_batched_input are exported to
5580 55. Changed the default runtime configuration: the message "Unknown user"
5581 has been removed from the ACL, and instead placed on the localuser router,
5582 using the cannot_route_message feature. This means that any verification
5583 failures that generate their own messages won't get overridden. Similarly,
5584 the "Unrouteable address" message that was in the ACL for unverifiable
5585 relay addresses has also been removed.
5587 56. Added hosts_avoid_esmtp to the smtp transport.
5589 57. The exicyclog script was not checking for the esoteric option
5590 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5591 will work only if exicyclog is run under the appropriate euid.
5593 58. Following a discussion on the list, the rules by which Exim recognises line
5594 endings on incoming messages have been changed. The -dropcr and drop_cr
5595 options are now no-ops, retained only for backwards compatibility. The
5596 following line terminators are recognized: LF CRLF CR. However, special
5597 processing applies to CR:
5599 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5600 nor a local message in the state where . is a terminator.
5602 (ii) If a bare CR is encountered in a header line, an extra space is added
5603 after the line terminator so as not to end the header. The reasoning
5604 behind this is that bare CRs in header lines are most likely either
5605 to be mistakes, or people trying to play silly games.
5607 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5608 was being incorrectly given as 18 bytes larger than it should have been.
5609 This is a VOB (very old bug).
5611 60. This may never have affected anything current, but just in case it has:
5612 When the local host is found other than at the start of a list of hosts,
5613 the local host, those with the same MX, and any that follow, are discarded.
5614 When the list in question was part of a longer list of hosts, the following
5615 hosts (not currently being processed) were also being discarded. This no
5616 longer happens. I'm not sure if this situation could ever has previously
5619 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5622 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5623 header. This is something that is recommended in a new Internet Draft, and
5624 is something that is documented as being done by Sendmail. There are two
5625 possible values. For messages generated by the autoreply transport, Exim
5628 Auto-Submitted: auto-replied
5630 whereas for all other generated messages (e.g. bounces) it adds
5632 Auto-Submitted: auto-generated
5634 63. The "personal" condition in filters now includes a test for the
5635 Auto-Submitted: header. If it contains the string "auto-" the message it
5636 not considered personal.
5638 64. Added rcpt_include_affixes as a generic transport option.
5640 65. Added queue_only_override (default true).
5642 66. Added the syslog_duplication option.
5644 67. If what should have been the first header line of a message consisted of
5645 a space followed by a colon, Exim was mis-interpreting it as a header line.
5646 It isn't of course - it is syntactically invalid and should therefore be
5647 treated as the start of the message body. The misbehaviour could have
5648 caused a number of strange effects, including loss of data in subsequent
5649 header lines, and spool format errors.
5651 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5652 client host had authenticated. This control can now be exercised by an ACL
5653 for more flexibility.
5655 69. By default, callouts do not happen when testing with -bh. There is now a
5656 variant, -bhc, which does actually run the callout code, including
5657 consulting and updating the callout cache.
5659 70. Added support for saslauthd authentication, courtesy of Alexander
5662 71. If statvfs() failed on the spool or log directories while checking their
5663 size for availability, Exim confusingly gave the error "space shortage".
5664 Furthermore, in debugging mode it crashed with a floating point exception.
5665 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5666 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5667 serious problem, Exim now writes to the main and panic logs when this
5668 happens, with details of the failure. It then refuses to accept the
5669 incoming message, giving the message "spool directory problem" or "log
5670 directory problem" with a 421 code for SMTP messages.
5672 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5673 0, 1, and 2 exist, because some OS complain for execs without them (see
5674 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5675 descriptors. However, the code omitted to check that the open succeeded,
5676 causing mysterious errors if for some reason the permissions on /dev/null
5677 got screwed. Now Exim writes a message to the main and panic logs, and
5678 bombs out if it can't open /dev/null.
5680 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5681 interact so that it is all more flexible. It is supposed to remain
5682 backwards compatible. Also added extra_local_interfaces.
5684 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5685 to bomb out with an assertion failure - to the client this appears as a
5686 connection drop. This problem occurs in the part of the code that was taken
5687 from the Samba project. Fortunately, the assertion is in a very simple
5688 function, so I have fixed this by reproducing the function inline in the
5689 one place where it is called, and arranging for authentication to fail
5690 instead of killing the process with assert().
5692 75. The SPA client code was not working when the server requested OEM rather
5693 than Unicode encoding.
5695 76. Added code to make require_files with a specific uid setting more usable in
5696 the case where statting the file as root fails - usually a non-root-mounted
5697 NFS file system. When this happens and the failure is EACCES, Exim now
5698 forks a subprocess and does the per-uid checking as the relevant uid.
5700 77. Added process_log_path.
5702 78. If log_file_path was not explicitly set, a setting of check_log_space or
5703 check_log_inodes was ignored.
5705 79. If a space check for the spool or log partitions fails, the incident is now
5706 logged. Of course, in the latter case the data may get lost...
5708 80. Added the %p formatting code to string_format() so that it can be used to
5709 print addresses in debug_print(). Adjusted all the address printing in the
5710 debugging in store.c to use %p rather than %d.
5712 81. There was a concern that a line of code in smtp_in.c could overflow a
5713 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5714 initially expressed, the concern was not well-founded, because trailing
5715 spaces are removed early. However, if the trailing spaces were followed by
5716 a NULL, they did not get removed, so the overflow was possible. Two fixes
5719 (a) I re-wrote the offending code in a cleaner fashion.
5720 (b) If an incoming SMTP command contains a NULL character, it is rejected
5723 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5724 runs initgroups(), so that if the Exim user is in any additional groups,
5725 they will be used during message reception.
5731 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob