1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
9 PP/02 Make -n do something, by making it not do something.
10 When combined with -bP, the name of an option is not output.
12 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
15 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
16 $sender_host_name and config options to manage this, and basic check
19 PP/05 DSCP support for outbound connections and control modifier for inbound.
21 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
22 (Only plugin which currently uses this is kerberos4, which nobody should
23 be using, but we should make it available and other future plugins might
24 conceivably use it, even though it would break NAT; stuff *should* be
25 using channel bindings instead).
27 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
28 name; added for Sendmail compatibility; requires admin caller.
29 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
30 just ignore it); requires trusted caller.
31 Also parse but ignore: -Ac -Am -X<logfile>
34 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
36 TL/02 Add +smtp_confirmation as a default logging option.
38 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
39 Patch by Magnus Holmgren from 2007-02-20.
41 TL/04 Bugzilla 1281 - Spec typo.
42 Bugzilla 1283 - Spec typo.
43 Bugzilla 1290 - Spec grammar fixes.
45 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
47 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
49 JH/02 Support "G" suffix to numbers in ${if comparisons.
51 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
53 NM/01 Bugzilla 1197 - Spec typo
54 Bugzilla 1196 - Spec examples corrections
56 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
58 PP/09 Add gnutls_enable_pkcs11 option.
60 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
61 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
63 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
64 "acl {{name}{arg}...}", and optional args on acl condition
67 JH/05 Permit multiple router/transport headers_add/remove lines.
69 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
71 JH/07 Avoid using a waiting database for a single-message-only transport.
72 Performance patch from Paul Fisher. Bugzilla 1262.
74 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
77 JH/09 Add $headers_added variable, with content from use of ACL modifier
78 add_header (but not yet added to the message). Bugzilla 199.
80 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
81 Pulled from Bugzilla 817 by Wolfgang Breyha.
83 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
86 JH/11 Add A= logging on delivery lines, and a client_set_id option on
93 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
95 This, or similar/improved, will also be change PP/11 of 4.82.
101 PP/01 Handle short writes when writing local log-files.
102 In practice, only affects FreeBSD (8 onwards).
103 Bugzilla 1053, with thanks to Dmitry Isaikin.
105 NM/01 Bugzilla 949 - Documentation tweak
107 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
110 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
112 PP/02 Implemented gsasl authenticator.
114 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
116 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
117 `pkg-config foo` for cflags/libs.
119 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
120 with rest of GSASL and with heimdal_gssapi.
122 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
123 `pkg-config foo` for cflags/libs for the TLS implementation.
125 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
126 properties get this fed in as external SSF. A number of robustness
127 and debugging improvements to the cyrus_sasl authenticator.
129 PP/08 cyrus_sasl server now expands the server_realm option.
131 PP/09 Bugzilla 1214 - Log authentication information in reject log.
132 Patch by Jeremy Harris.
134 PP/10 Added dbmjz lookup type.
136 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
138 PP/12 MAIL args handles TAB as well as SP, for better interop with
139 non-compliant senders.
140 Analysis and variant patch by Todd Lyons.
142 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
143 Bug report from Lars Müller <lars@samba.org> (via SUSE),
144 Patch from Dirk Mueller <dmueller@suse.com>
146 PP/13 tls_peerdn now print-escaped for spool files.
147 Observed some $tls_peerdn in wild which contained \n, which resulted
148 in spool file corruption.
150 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
151 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
152 or write after TLS renegotiation, which otherwise led to messages
155 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
156 as a tracking header (ie: a signed header comes before the signature).
157 Patch from Wolfgang Breyha.
159 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
160 comma-sep list; embedded commas doubled.
162 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
164 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
166 Report and patch from Dmitry Banschikov.
168 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
169 Removed SSL_clear() after SSL_new() which led to protocol negotiation
170 failures. We appear to now support TLS1.1+ with Exim.
172 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
173 lets Exim select keys and certificates based upon TLS SNI from client.
174 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
175 before an outbound SMTP session. New log_selector, +tls_sni.
177 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
178 NULL dereference. Report and patch from Alun Jones.
180 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
181 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
182 Not seeing resolver debug output on NetBSD, but suspect this is a
183 resolver implementation change.
185 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
186 Left warnings. Added "eximon gdb" invocation mode.
188 PP/21 Defaulting "accept_8bitmime" to true, not false.
190 PP/22 Added -bw for inetd wait mode support.
192 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
193 locate the relevant includes and libraries. Made this the default.
195 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
196 Bugzilla 1246, report and most of solution from Tomasz Kusy.
198 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
199 This may cause build issues on older platforms.
201 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
202 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
203 gnutls_require_mac & gnutls_require_protocols (no longer supported).
204 Added SNI support via GnuTLS too.
205 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
207 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
209 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
210 Applied second patch from Janne, implementing suggestion to default
211 multiple-strings-in-record handling to match SPF spec.
213 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
215 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
216 read-only, out of scope).
217 Patch from Wolfgang Breyha, report from Stuart Northfield.
219 PP/29 Fix three issues highlighted by clang analyser static analysis.
220 Only crash-plausible issue would require the Cambridge-specific
221 iplookup router and a misconfiguration.
222 Report from Marcin Mirosław.
224 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
226 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
227 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
228 As part of this, removing so much warning spew let me fix some minor
229 real issues in debug logging.
231 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
232 assignment on my part. Fixed.
234 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
235 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
236 Janne Snabb (who went above and beyond: thank you).
238 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
239 string otherwise requires a connection and a bunch more work and it's
240 relatively easy to get wrong. Should also expose TLS library linkage
243 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
244 64-bit ${eval} (JH/03).
246 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
247 GNU libc to support some of the 64-bit stuff, should not lead to
248 conflicts. Defined before os.h is pulled in, so if a given platform
249 needs to override this, it can.
251 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
252 protection layer was required, which is not implemented.
253 Bugzilla 1254, patch from Wolfgang Breyha.
255 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
256 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
257 tls_dhparam take prime identifiers. Also unbreak combination of
258 OpenSSL+DH_params+TLSSNI.
260 PP/39 Disable SSLv2 by default in OpenSSL support.
266 PP/01 Solaris build fix for Oracle's LDAP libraries.
267 Bugzilla 1109, patch from Stephen Usher.
269 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
271 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
274 TF/02 Fix a couple more cases where we did not log the error message
275 when unlink() failed. See also change 4.74-TF/03.
277 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
278 lock up or crash if it happened to be inside a call to libc when it
279 got a SIGUSR1 from exiwhat.
281 The SIGUSR1 handler appends the current process status to the process
282 log which is later printed by exiwhat. It used to use the general
283 purpose logging code to do this, but several functions it calls are
284 not safe for signals.
286 The new output code in the SIGUSR1 handler is specific to the process
287 log, and simple enough that it's easy to inspect for signal safety.
288 Removing some special cases also simplifies the general logging code.
289 Removing the spurious timestamps from the process log simplifies
292 TF/04 Improved ratelimit ACL condition.
294 The /noupdate option has been deprecated in favour of /readonly which
295 has clearer semantics. The /leaky, /strict, and /readonly update modes
296 are mutually exclusive. The update mode is no longer included in the
297 database key; it just determines when the database is updated. (This
298 means that when you upgrde Exim will forget old rate measurements.)
300 Exim now checks that the per_* options are used with an update mode that
301 makes sense for the current ACL. For example, when Exim is processing a
302 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
303 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
304 must specify per_mail/readonly. If you omit the update mode it defaults to
305 /leaky where that makes sense (as before) or /readonly where required.
307 The /noupdate option is now undocumented but still supported for
308 backwards compatibility. It is equivalent to /readonly except that in
309 ACLs where /readonly is required you may specify /leaky/noupdate or
310 /strict/noupdate which are treated the same as /readonly.
312 A useful new feature is the /count= option. This is a generalization
313 of the per_byte option, so that you can measure the throughput of other
314 aggregate values. For example, the per_byte option is now equivalent
315 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
317 The per_rcpt option has been generalized using the /count= mechanism
318 (though it's more complicated than the per_byte equivalence). When it is
319 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
320 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
321 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
322 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
323 Note that using per_rcpt with a non-readonly update mode in more than
324 one ACL will cause the recipients to be double-counted. (The per_mail
325 and per_byte options don't have this problem.)
327 The handling of very low rates has changed slightly. If the computed rate
328 is less than the event's count (usually one) then this event is the first
329 after a long gap. In this case the rate is set to the same as this event's
330 count, so that the first message of a spam run is counted properly.
332 The major new feature is a mechanism for counting the rate of unique
333 events. The new per_addr option counts the number of different
334 recipients that someone has sent messages to in the last time period. It
335 behaves like per_rcpt if all the recipient addresses are different, but
336 duplicate recipient addresses do not increase the measured rate. Like
337 the /count= option this is a general mechanism, so the per_addr option
338 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
339 example, measure the rate that a client uses different sender addresses
340 with the options per_mail/unique=$sender_address. There are further
341 details in the main documentation.
343 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
345 TF/06 Removed a few PCRE remnants.
347 TF/07 Automatically extract Exim's version number from tags in the git
348 repository when doing development or release builds.
350 PP/02 Raise smtp_cmd_buffer_size to 16kB.
351 Bugzilla 879. Patch from Paul Fisher.
353 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
354 Heavily based on revision 40f9a89a from Simon Arlott's tree.
357 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
359 PP/05 Variable $av_failed, true if the AV scanner deferred.
360 Bugzilla 1078. Patch from John Horne.
362 PP/06 Stop make process more reliably on build failure.
363 Bugzilla 1087. Patch from Heiko Schlittermann.
365 PP/07 Make maildir_use_size_file an _expandable_ boolean.
366 Bugzilla 1089. Patch from Heiko Schlittermann.
368 PP/08 Handle ${run} returning more data than OS pipe buffer size.
369 Bugzilla 1131. Patch from Holger Weiß.
371 PP/09 Handle IPv6 addresses with SPF.
372 Bugzilla 860. Patch from Wolfgang Breyha.
374 PP/10 GnuTLS: support TLS 1.2 & 1.1.
376 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
379 PP/11 match_* no longer expand right-hand-side by default.
380 New compile-time build option, EXPAND_LISTMATCH_RHS.
381 New expansion conditions, "inlist", "inlisti".
383 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
385 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
387 PP/14 fix log_write() format string regression from TF/03.
388 Bugzilla 1152. Patch from Dmitry Isaikin.
394 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
396 PP/02 Harmonised TLS library version reporting; only show if debugging.
397 Layout now matches that introduced for other libraries in 4.74 PP/03.
399 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
401 PP/04 New "dns_use_edns0" global option.
403 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
406 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
407 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
409 TK/01 Updated PolarSSL code to 0.14.2.
410 Bugzilla 1097. Patch from Andreas Metzler.
412 PP/07 Catch divide-by-zero in ${eval:...}.
415 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
418 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
419 format-string attack -- SECURITY: remote arbitrary code execution.
421 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
422 time unintentionally subject to list matching rules, letting the header
423 cause arbitrary Exim lookups (of items which can occur in lists, *not*
424 arbitrary string expansion). This allowed for information disclosure.
426 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
427 INT_MIN/-1 -- value coerced to INT_MAX.
433 NM/01 Workround for PCRE version dependency in version reporting
436 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
437 This fixes portability to compilers other than gcc, notably
438 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
440 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
441 makefiles for portability to HP-UX and POSIX correctness.
443 PP/01 Permit LOOKUP_foo enabling on the make command-line.
444 Also via indented variable definition in the Makefile.
445 (Debugging by Oliver Heesakkers).
447 PP/02 Restore caching of spamd results with expanded spamd_address.
448 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
450 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
451 Improves build reliability. Fix from: Frank Elsner
453 NM/02 Fix wide character breakage in the rfc2047 coding
454 Fixes bug 1064. Patch from Andrey N. Oktyabrski
456 NM/03 Allow underscore in dnslist lookups
457 Fixes bug 1026. Patch from Graeme Fowler
459 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
460 Code patches from Adam Ciarcinski of NetBSD.
462 NM/04 Fixed exiqgrep to cope with mailq missing size issue
465 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
466 is logged, to avoid truncation. Patch from John Horne.
468 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
469 Patch from Jakob Hirsch.
471 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
472 SQL string expansion failure details.
473 Patch from Andrey Oktyabrski.
475 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
476 Patch from Simon Arlott.
478 PP/09 New lookups functionality failed to compile on old gcc which rejects
479 extern declarations in function scope.
480 Patch from Oliver Fleischmann
482 PP/10 Use sig_atomic_t for flags set from signal handlers.
483 Check getgroups() return and improve debugging.
484 Fixed developed for diagnosis in bug 927 (which turned out to be
487 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
488 Patch from Mark Zealey.
490 PP/12 Bugzilla 1056: Improved spamd server selection.
491 Patch from Mark Zealey.
493 PP/13 Bugzilla 1086: Deal with maildir quota file races.
494 Based on patch from Heiko Schlittermann.
496 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
497 Patch from Uwe Doering, sign-off by Michael Haardt.
499 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
500 variable declaration deep within a block. Bug and patch from
503 PP/15 lookups-Makefile IRIX compatibilty coercion.
505 PP/16 Make DISABLE_DKIM build knob functional.
507 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
508 Patch by Simon Arlott
510 TF/03 Fix valgrind.h portability to C89 compilers that do not support
511 variable argument macros. Our copy now differs from upstream.
517 TF/01 Failure to get a lock on a hints database can have serious
518 consequences so log it to the panic log.
520 TF/02 Log LMTP confirmation messages in the same way as SMTP,
521 controlled using the smtp_confirmation log selector.
523 TF/03 Include the error message when we fail to unlink a spool file.
525 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
526 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
527 for maintaining out-of-tree patches for some time.
529 PP/01 Bugzilla 139: Documentation and portability issues.
530 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
531 Handle per-OS dynamic-module compilation flags.
533 PP/02 Let /dev/null have normal permissions.
534 The 4.73 fixes were a little too stringent and complained about the
535 permissions on /dev/null. Exempt it from some checks.
536 Reported by Andreas M. Kirchwitz.
538 PP/03 Report version information for many libraries, including
539 Exim version information for dynamically loaded libraries. Created
540 version.h, now support a version extension string for distributors
541 who patch heavily. Dynamic module ABI change.
543 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
544 privilege escalation vulnerability whereby the Exim run-time user
545 can cause root to append content of the attacker's choosing to
548 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
551 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
552 If dropping privileges for untrusted macros, we disabled normal logging
553 on the basis that it would fail; for the Exim run-time user, this is not
554 the case, and it resulted in successful deliveries going unlogged.
555 Fixed. Reported by Andreas Metzler.
561 PP/01 Date: & Message-Id: revert to normally being appended to a message,
562 only prepend for the Resent-* case. Fixes regression introduced in
563 Exim 4.70 by NM/22 for Bugzilla 607.
565 PP/02 Include check_rfc2047_length in configure.default because we're seeing
566 increasing numbers of administrators be bitten by this.
568 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
570 PP/03 Bugzilla 994: added openssl_options main configuration option.
572 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
574 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
576 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
578 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
579 without a peer certificate, leading to a segfault because of an
580 assumption that peers always have certificates. Be a little more
581 paranoid. Problem reported by Martin Tscholak.
583 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
584 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
585 NB: ClamAV planning to remove STREAM in "middle of 2010".
586 CL also introduces -bmalware, various -d+acl logging additions and
587 more caution in buffer sizes.
589 PP/09 Implemented reverse_ip expansion operator.
591 PP/10 Bugzilla 937: provide a "debug" ACL control.
593 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
595 PP/12 Bugzilla 973: Implement --version.
597 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
599 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
601 PP/15 Bugzilla 816: support multiple condition rules on Routers.
603 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
604 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
605 ignore trailing whitespace.
607 JJ/02 prevent non-panic DKIM error from being sent to paniclog
609 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
612 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
613 Notification from Dr Andrew Aitchison.
615 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
616 ExtendedDetectionInfo response format.
617 Notification from John Horne.
619 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
622 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
623 XSL and documented dependency on system catalogs, with examples of how
626 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
629 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
630 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
631 configuration file which is writeable by the Exim user or group.
633 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
634 of configuration files to cover files specified with the -C option if
635 they are going to be used with root privileges, not just the default
638 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
639 option (effectively making it always true).
641 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
642 files to be used while preserving root privileges.
644 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
645 that rogue child processes cannot use them.
647 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
648 run-time user, instead of root.
650 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
651 Exim run-time user without dropping privileges.
653 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
654 result string, instead of calling string_vformat() twice with the same
657 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
658 for other users. Others should always drop root privileges if they use
659 -C on the command line, even for a whitelisted configure file.
661 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
663 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
669 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
670 $data_path, and $header_path variables; fixed documentation bugs and
673 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
674 exipick to access non-standard spools, including the "frozen" queue
677 NM/01 Bugzilla 965: Support mysql stored procedures.
678 Patch from Alain Williams
680 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
682 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
683 Patch from Andreas Metzler
685 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
686 Patch from Kirill Miazine
688 NM/05 Bugzilla 671: Added umask to procmail example.
690 JJ/03 installed exipick 20100323.0, fixing doc bug
692 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
693 directory. Notification and patch from Dan Rosenberg.
695 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
697 TK/02 Improve log output when DKIM signing operation fails.
699 MH/01 Treat the transport option dkim_domain as a colon separated
700 list, not as a single string, and sign the message with each element,
701 omitting multiple occurences of the same signer.
703 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
704 Bugzilla 985, 986. Patch by Simon Arlott
706 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
707 Patch by Simon Arlott
709 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
710 MBX locking. Notification from Dan Rosenberg.
716 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
718 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
720 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
722 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
724 NM/04 Bugzilla 915: Flag broken perl installation during build.
730 TK/01 Added patch by Johannes Berg that expands the main option
731 "spamd_address" if it starts with a dollar sign.
733 TK/02 Write list of recipients to X-Envelope-Sender header when building
734 the mbox-format spool file for content scanning (suggested by Jakob
737 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
738 (http://www.dcc-servers.net/) support via dccifd. Activated by
739 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
741 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
742 by Mark Daniel Reidel <mr@df.eu>.
744 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
745 When building exim an external PCRE library is now needed -
746 PCRE is a system library on the majority of modern systems.
747 See entry on PCRE_LIBS in EDITME file.
749 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
750 conversation. Added nologin parameter to request.
751 Patch contributed by Kirill Miazine.
753 TF/01 Do not log submission mode rewrites if they do not change the address.
755 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
757 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
758 log files in place. Contributed by Roberto Lima.
760 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
762 TF/03 Bugzilla 615: When checking the local_parts router precondition
763 after a local_part_suffix or local_part_prefix option, Exim now
764 does not use the address's named list lookup cache, since this
765 contains cached lookups for the whole local part.
767 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
768 Robert Millan. Documentation is in experimental-spec.txt.
770 TF/04 Bugzilla 668: Fix parallel build (make -j).
772 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
774 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
775 Patch provided by Jan Srzednicki.
777 TF/05 Leading white space used to be stripped from $spam_report which
778 wrecked the formatting. Now it is preserved.
780 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
781 that they are available at delivery time.
783 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
785 TF/08 TLS error reporting now respects the incoming_interface and
786 incoming_port log selectors.
788 TF/09 Produce a more useful error message if an SMTP transport's hosts
789 setting expands to an empty string.
791 NM/06 Bugzilla 744: EXPN did not work under TLS.
792 Patch provided by Phil Pennock.
794 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
795 Patch provided by Richard Godbee.
797 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
798 acl_smtp_notquit, added index entry.
800 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
801 Patch provided by Eugene Bujak.
803 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
804 accept(). Patch provided by Maxim Dounin.
806 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
807 Patch provided by Phil Pennock.
809 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
811 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
812 Patch provided by Brad "anomie" Jorsch.
814 NM/14 Bugzilla 622: Added timeout setting to transport filter.
815 Patch provided by Dean Brooks.
817 TK/05 Add native DKIM support (does not depend on external libraries).
819 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
820 Patch provided by Graeme Fowler.
822 NM/16 Bugzilla 851: Documentation example syntax fix.
824 NM/17 Changed NOTICE file to remove references to embedded PCRE.
826 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
829 NM/19 Bugzilla 745: TLS version reporting.
830 Patch provided by Phil Pennock.
832 NM/20 Bugzilla 167: bool: condition support.
833 Patch provided by Phil Pennock.
835 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
836 clients. Patch provided by Phil Pennock.
838 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
839 Patch provided by Brad "anomie" Jorsch.
841 NM/23 Bugzilla 687: Fix misparses in eximstats.
842 Patch provided by Heiko Schlittermann.
844 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
845 Patch provided by Heiko Schlittermann.
847 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
848 plus update to original patch.
850 NM/26 Bugzilla 799: Documentation correction for ratelimit.
852 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
853 Patch provided by David Brownlee.
855 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
857 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
859 NM/30 Bugzilla 888: TLS documentation bugfixes.
861 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
863 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
864 Unlike the original bugzilla I have changed all shell scripts in src tree.
866 NM/33 Bugzilla 898: Transport filter timeout fix.
867 Patch by Todd Rinaldo.
869 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
870 Patch by Serge Demonchaux.
872 NM/35 Bugzilla 39: Base64 decode bug fixes.
873 Patch by Jakob Hirsch.
875 NM/36 Bugzilla 909: Correct connect() call in dcc code.
877 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
879 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
881 NM/39 Bugzilla 911: Fixed MakeLinks build script.
887 TK/01 Add preliminary DKIM support. Currently requires a forked version of
888 ALT-N's libdkim that I have put here:
889 http://duncanthrax.net/exim-experimental/
891 Note to Michael Haardt: I had to rename some vars in sieve.c. They
892 were called 'true' and it seems that C99 defines that as a reserved
893 keyword to be used with 'bool' variable types. That means you could
894 not include C99-style headers which use bools without triggering
895 build errors in sieve.c.
897 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
898 as mailq or other aliases. Changed the --help handling significantly
899 to do whats expected. exim_usage() emits usage/help information.
901 SC/01 Added the -bylocaldomain option to eximstats.
903 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
905 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
907 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
913 PH/01 Another patch from the Sieve maintainer.
915 PH/02 When an IPv6 address is converted to a string for single-key lookup
916 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
917 dots are used instead of colons so that keys in lsearch files need not
918 contain colons. This was done some time before quoting was made available
919 in lsearch files. However, iplsearch files do require colons in IPv6 keys
920 (notated using the quote facility) so as to distinguish them from IPv4
921 keys. This meant that lookups for IP addresses in host lists did not work
922 for iplsearch lookups.
924 This has been fixed by arranging for IPv6 addresses to be expressed with
925 colons if the lookup type is iplsearch. This is not incompatible, because
926 previously such lookups could never work.
928 The situation is now rather anomolous, since one *can* have colons in
929 ordinary lsearch keys. However, making the change in all cases is
930 incompatible and would probably break a number of configurations.
932 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
935 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
936 conversion specification without a maximum field width, thereby enabling
937 a rogue spamd server to cause a buffer overflow. While nobody in their
938 right mind would setup Exim to query an untrusted spamd server, an
939 attacker that gains access to a server running spamd could potentially
940 exploit this vulnerability to run arbitrary code as the Exim user.
942 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
943 $primary_hostname instead of what libspf2 thinks the hosts name is.
945 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
946 a directory entry by the name of the lookup key. Previously, if a
947 symlink pointed to a non-existing file or a file in a directory that
948 Exim lacked permissions to read, a lookup for a key matching that
949 symlink would fail. Now it is enough that a matching directory entry
950 exists, symlink or not. (Bugzilla 503.)
952 PH/03 The body_linecount and body_zerocount variables are now exported in the
955 PH/04 Added the $dnslist_matched variable.
957 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
958 This means they are set thereafter only if the connection becomes
961 PH/06 Added the client_condition to authenticators so that some can be skipped
962 by clients under certain conditions.
964 PH/07 The error message for a badly-placed control=no_multiline_responses left
965 "_responses" off the end of the name.
967 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
969 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
970 (without spaces) instead of just copying the configuration text.
972 PH/10 Added the /noupdate option to the ratelimit ACL condition.
974 PH/11 Added $max_received_linelength.
976 PH/12 Added +ignore_defer and +include_defer to host lists.
978 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
979 way in which PCRE > 7.0 is built.
981 PH/14 Implemented queue_only_load_latch.
983 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
984 MAIL command. The effect was to mangle the value on 64-bit systems.
986 PH/16 Another patch from the Sieve maintainer.
988 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
990 PH/18 If a system quota error occurred while trying to create the file for
991 a maildir delivery, the message "Mailbox is full" was not appended to the
992 bounce if the delivery eventually timed out. Change 4.67/27 below applied
993 only to a quota excession during the actual writing of the file.
995 PH/19 It seems that peer DN values may contain newlines (and other non-printing
996 characters?) which causes problems in log lines. The DN values are now
997 passed through string_printing() before being added to log lines.
999 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1000 and InterBase are left for another time.)
1002 PH/21 Added message_body_newlines option.
1004 PH/22 Guard against possible overflow in moan_check_errorcopy().
1006 PH/23 POSIX allows open() to be a macro; guard against that.
1008 PH/24 If the recipient of an error message contained an @ in the local part
1009 (suitably quoted, of course), incorrect values were put in $domain and
1010 $local_part during the evaluation of errors_copy.
1016 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1017 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1020 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1021 issue a MAIL command.
1023 PH/02 In an ACL statement such as
1025 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1027 if a client was not listed at all, or was listed with a value other than
1028 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1029 the condition was not true (as it should be), so access was not denied.
1030 The bug was that the ! inversion was incorrectly passed on to the second
1031 item. This has been fixed.
1033 PH/03 Added additional dnslists conditions == and =& which are different from
1034 = and & when the dns lookup returns more than one IP address.
1036 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1037 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1039 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1040 FSYNC, which compiles an option called disable_fsync that allows for
1041 bypassing fsync(). The documentation is heavily laced with warnings.
1043 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1045 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1046 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1047 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1048 including adding "make clean"; (3) Added -fPIC when compiling the test
1049 dynamically loaded module, to get rid of a warning.
1051 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1052 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1053 = 0s. The bug is otherwise harmless.
1055 PH/07 There was a bug in the dovecot authenticator such that the value of
1056 $auth1 could be overwritten, and so not correctly preserved, after a
1057 successful authentication. This usually meant that the value preserved by
1058 the server_setid option was incorrect.
1060 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1062 PH/09 Installed PCRE release 7.0.
1064 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1065 run for batched SMTP input. It is now run at the start of every message
1066 in the batch. While fixing this I discovered that the process information
1067 (output by running exiwhat) was not always getting set for -bs and -bS
1068 input. This is fixed, and it now also says "batched" for BSMTP.
1070 PH/11 Added control=no_pipelining.
1072 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1073 patch, slightly modified), and move the expansion of helo_data till after
1074 the connection is made in the smtp transport (so it can use these
1077 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1079 PH/14 Added log_selector = +pid.
1081 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1083 PH/16 Add ${if forany and ${if forall.
1085 PH/17 Added dsn_from option to vary the From: line in DSNs.
1087 PH/18 Flush SMTP output before performing a callout, unless control =
1088 no_callout_flush is set.
1090 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1091 was true (the default) a successful delivery failed to delete the retry
1092 item, thus causing premature timeout of the address. The bug is now
1095 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1097 PH/21 Long custom messages for fakedefer and fakereject are now split up
1098 into multiline reponses in the same way that messages for "deny" and
1099 other ACL rejections are.
1101 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1102 with slight modification.
1104 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1105 draft, changing the syntax and factoring some duplicate code".
1107 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1108 for deliveries of the second and subsequent messages over the same SMTP
1111 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1112 ${reduce, with only minor "tidies".
1114 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1116 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1117 expansion side effects.
1119 PH/27 When a message times out after an over-quota error from an Exim-imposed
1120 quota, the bounce message says "mailbox is full". This message was not
1121 being given when it was a system quota that was exceeded. It now should
1124 MH/03 Made $recipients available in local_scan(). local_scan() already has
1125 better access to the recipient list through recipients_list[], but
1126 $recipients can be useful in postmaster-provided expansion strings.
1128 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1129 in the case of a MAIL command with additional options following the
1130 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1131 were accidentally chopped off.
1133 PH/29 SMTP synchronization checks are implemented when a command is read -
1134 there is a check that no more input is waiting when there shouldn't be
1135 any. However, for some commands, a delay in an ACL can mean that it is
1136 some time before the response is written. In this time, more input might
1137 arrive, invalidly. So now there are extra checks after an ACL has run for
1138 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1139 pipelining has not been advertised.
1141 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1143 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1144 returned in the SMTP response when smtp_return_error_details was false.
1145 This has been fixed.
1147 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1148 instead of the C I/O that was originally supplied, because problems were
1149 reported on Solaris.
1151 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1152 Exim which did not show up earlier: it was assuming that a call to
1153 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1154 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1155 was a macro that became an assignment, so it seemed to work. This has
1156 changed to a proper function call with a void return, hence the compile
1157 error. Exim's code has been fixed.
1159 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1162 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1164 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1166 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1167 because it turns out to be too big or there is a timeout) there is no
1168 "Completed" line in the log. When some messages of this type were
1169 selected by exigrep, they were listed as "not completed". Others were
1170 picked up by some special patterns. I have improved the selection
1171 criteria to be more general.
1173 PH/38 The host_find_failed option in the manualroute router can now be set
1174 to "ignore", to completely ignore a host whose IP address cannot be
1175 found. If all hosts are ignored, the behaviour is controlled by the new
1176 host_all_ignored option.
1178 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1179 homing or because of multiple MX records with /mx) generated more than
1180 one IP address, and the following item turned out to be the local host,
1181 all the secondary addresses of the first item were incorrectly removed
1182 from the list, along with the local host and any following hosts (which
1183 is what is supposed to happen).
1185 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1186 whoever called Exim into the -H file. In the case of the daemon it was
1187 behaving confusingly. When first started, it used values for whoever
1188 started the daemon, but after a SIGHUP it used the Exim user (because it
1189 calls itself on a restart). I have changed the code so that it now always
1192 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1193 message are rejected with the same error (e.g. no authentication or bad
1194 sender address), and a DATA command is nevertheless sent (as can happen
1195 with PIPELINING or a stupid MUA), the error message that was given to the
1196 RCPT commands is included in the rejection of the DATA command. This is
1197 intended to be helpful for MUAs that show only the final error to their
1200 PH/42 Another patch from the Sieve maintainer.
1202 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1203 Eximstats - Fixed some broken HTML links and added missing column headers
1205 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1206 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1208 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1210 PH/43 Yet another patch from the Sieve maintainer.
1212 PH/44 I found a way to check for a TCP/IP connection going away before sending
1213 the response to the final '.' that terminates a message, but only in the
1214 case where the client has not sent further data following the '.'
1215 (unfortunately, this is allowed). However, in many cases there won't be
1216 any further data because there won't be any more messages to send. A call
1217 to select() can be used: if it shows that the input is "ready", there is
1218 either input waiting, or the socket has been closed. An attempt to read
1219 the next input character can distinguish the two cases. Previously, Exim
1220 would have sent an OK response which the client would never have see.
1221 This could lead to message repetition. This fix should cure that, at
1222 least in a lot of common cases.
1224 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1225 advertised in response to EHLO.
1231 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1232 fixed by 4.65/MH/01 (is this a record?) are fixed:
1234 (i) An empty string was always treated as zero by the numeric comparison
1235 operators. This behaviour has been restored.
1237 (ii) It is documented that the numeric comparison operators always treat
1238 their arguments as decimal numbers. This was broken in that numbers
1239 starting with 0 were being interpreted as octal.
1241 While fixing these problems I realized that there was another issue that
1242 hadn't been noticed. Values of message_size_limit (both the global option
1243 and the transport option) were treated as octal if they started with 0.
1244 The documentation was vague. These values are now always treated as
1245 decimal, and I will make that clear in the documentation.
1251 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1252 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1255 MH/01 Don't check that the operands of numeric comparison operators are
1256 integers when their expansion is in "skipping" mode (fixes bug
1257 introduced by 4.64-PH/07).
1259 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1260 child addresses, Exim now panics and dies. Previously, because the count
1261 is held in a short int, deliveries were likely to be lost. As such a
1262 large number of recipients for a single message is ridiculous
1263 (performance will be very, very poor), I have chosen to impose a limit
1264 rather than extend the field.
1270 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1271 leftover -K file (the existence of which was triggered by #402).
1272 While we were at it, introduced process PID as part of the -K
1273 filename. This should rule out race conditions when creating
1276 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1277 processing considerably. Previous code took too long for large mails,
1278 triggering a timeout which in turn triggers #401.
1280 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1281 in the DK code in transports.c. sendfile() is not really portable,
1282 hence the _LINUX specificness.
1284 TF/01 In the add_headers option to the mail command in an Exim filter,
1285 there was a bug that Exim would claim a syntax error in any
1286 header after the first one which had an odd number of characters
1289 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1290 callout verification, Exim cached a "reject" for the entire domain. This
1291 is correct for most verifications, but it is not correct for a recipient
1292 verification with use_sender or use_postmaster set, because in that case
1293 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1294 case of MAIL FROM:<> rejection from other early rejections (e.g.
1295 rejection of HELO). When verifying a recipient using a non-null MAIL
1296 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1297 Whatever the result of the callout, the value of the domain cache is
1298 left unchanged (for any other kind of callout, getting as far as trying
1299 RCPT means that the domain itself is ok).
1301 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1304 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1305 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1306 session, the daemon ignores ECONNECT errors and logs others; it now
1307 ignores EPIPE as well.
1309 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1310 (quoted-printable decoding).
1312 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1313 later the small subsequent patch to fix an introduced bug.
1315 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1317 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1319 PH/08 An error is now given if message_size_limit is specified negative.
1321 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1322 to be given (somewhat) arbitrary names.
1324 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1327 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1328 miscellaneous code fixes
1330 PH/10 Added the log_reject_target ACL modifier to specify where to log
1333 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1334 hostname. This is wrong, because it relates to the incoming message (and
1335 probably the interface on which it is arriving) and not to the outgoing
1336 callout (which could be using a different interface). This has been
1337 changed to use the value of the helo_data option from the smtp transport
1338 instead - this is what is used when a message is actually being sent. If
1339 there is no remote transport (possible with a router that sets up host
1340 addresses), $smtp_active_hostname is used.
1342 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1343 tweaks were necessary in order to get it to work (see also 21 below):
1344 (a) The code assumed that strncpy() returns a negative number on buffer
1345 overflow, which isn't the case. Replaced with Exim's string_format()
1347 (b) There were several signed/unsigned issues. I just did the minimum
1348 hacking in of casts. There is scope for a larger refactoring.
1349 (c) The code used strcasecmp() which is not a standard C function.
1350 Replaced with Exim's strcmpic() function.
1351 (d) The code set only $1; it now sets $auth1 as well.
1352 (e) A simple test gave the error "authentication client didn't specify
1353 service in request". It would seem that Dovecot has changed its
1354 interface. Fortunately there's a specification; I followed it and
1355 changed what the client sends and it appears to be working now.
1357 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1360 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1361 address A is aliased to B and C, where B exists and C does not. Without
1362 -v the output is "A verified" because verification stops after a
1363 successful redirection if more than one address is generated. However,
1364 with -v the child addresses are also verified. Exim was outputting "A
1365 failed to verify" and then showing the successful verification for C,
1366 with its parentage. It now outputs "B failed to verify", showing B's
1367 parentage before showing the successful verification of C.
1369 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1370 look up a TXT record in a specific list after matching in a combined
1373 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1374 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1375 they consult the DNS. I had assumed they would set it the way they
1376 wanted; and indeed my experiments on Linux seem to show that in some
1377 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1378 To be on the safe side, however, I have now made the interface to
1379 host_find_byname() similar to host_find_bydns(), with an argument
1380 containing the DNS resolver options. The host_find_byname() function now
1381 sets these options at its start, just as host_find_bydns() does. The smtp
1382 transport options dns_qualify_single and dns_search_parents are passed to
1383 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1384 of host_find_byname() use the default settings of RES_DEFNAMES
1385 (qualify_single) but not RES_DNSRCH (search_parents).
1387 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1388 spool_read_header() do less string testing, by means of a preliminary
1389 switch on the second character of optional "-foo" lines. (This is
1390 overdue, caused by the large number of possibilities that now exist.
1391 Originally there were few.) While I was there, I also converted the
1392 str(n)cmp tests so they don't re-test the leading "-" and the first
1393 character, in the hope this might squeeze out yet more improvement.
1395 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1396 flag allowing group syntax was set by the header_syntax check but not
1397 turned off, possible causing trouble later; (2) The flag was not being
1398 set at all for the header_verify test, causing "group"-style headers to
1399 be rejected. I have now set it in this case, and also caused header_
1400 verify to ignore an empty address taken from a group. While doing this, I
1401 came across some other cases where the code for allowing group syntax
1402 while scanning a header line wasn't quite right (mostly, not resetting
1403 the flag correctly in the right place). These bugs could have caused
1404 trouble for malformed header lines. I hope it is now all correct.
1406 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1407 with the "reply" argument non-NULL. The code, however (which originally
1408 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1409 but it didn't always do it. This confused somebody who was copying the
1410 code for some other use. I have removed all the tests.
1412 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1413 feature that was used to support insecure browsers during the U.S. crypto
1414 embargo. It requires special client support, and Exim is probably the
1415 only MTA that supported it -- and would never use it because real RSA is
1416 always available. This code has been removed, because it had the bad
1417 effect of slowing Exim down by computing (never used) parameters for the
1418 RSA_EXPORT functionality.
1420 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1421 authenticator to fail if there's a tab character in the incoming data
1422 (there should never be unless someone is messing about, as it's supposed
1423 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1424 if the connection is using TLS or if the remote IP is the same as the
1425 local IP, and the "valid-client-cert option" if a client certificate has
1428 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1429 authenticators. This can be used for authorization after authentication
1430 succeeds. (In the case of plaintext, it servers for both authentication
1433 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1434 if any retry times were supplied.
1436 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1437 connection, where there is no client IP address to check. In this
1438 situation, the verify now always succeeds.
1440 PH/25 Applied John Jetmore's -Mset patch.
1442 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1444 PH/27 In a string expansion for a processed (not raw) header when multiple
1445 headers of the same name were present, leading whitespace was being
1446 removed from all of them, but trailing whitespace was being removed only
1447 from the last one. Now trailing whitespace is removed from each header
1448 before concatenation. Completely empty headers in a concatenation (as
1449 before) are ignored.
1451 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1452 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1454 PH/29 [Removed. This was a change that I later backed out, and forgot to
1455 correct the ChangeLog entry (that I had efficiently created) before
1456 committing the later change.]
1458 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1459 address errors (4xx response to RCPT) over the same connection as other
1460 messages routed to the same hosts. Such deliveries are always "forced",
1461 so retry times are not inspected. This resulted in far too many retries
1462 for the affected addresses. The effect occurred only when there were more
1463 hosts than the hosts_max_try setting in the smtp transport when it had
1464 the 4xx errors. Those hosts that it had tried were not added to the list
1465 of hosts for which the message was waiting, so if all were tried, there
1466 was no problem. Two fixes have been applied:
1468 (i) If there are any address or message errors in an SMTP delivery, none
1469 of the hosts (tried or untried) are now added to the list of hosts
1470 for which the message is waiting, so the message should not be a
1471 candidate for sending over the same connection that was used for a
1472 successful delivery of some other message. This seems entirely
1473 reasonable: after all the message is NOT "waiting for some host".
1474 This is so "obvious" that I'm not sure why it wasn't done
1475 previously. Hope I haven't missed anything, but it can't do any
1476 harm, as the worst effect is to miss an optimization.
1478 (ii) If, despite (i), such a delivery is accidentally attempted, the
1479 routing retry time is respected, so at least it doesn't keep
1480 hammering the server.
1482 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1483 in ${readsocket because some servers need this prod.
1485 PH/32 Added some extra debug output when updating a wait-xxx database.
1487 PH/33 The hint "could be header name not terminated by colon", which has been
1488 given for certain expansion errors for a long time, was not being given
1489 for the ${if def:h_colon_omitted{... case.
1491 PH/34 The spec says: "With one important exception, whenever a domain list is
1492 being scanned, $domain contains the subject domain." There was at least
1493 one case where this was not true.
1495 PH/35 The error "getsockname() failed: connection reset by peer" was being
1496 written to the panic log as well as the main log, but it isn't really
1497 panic-worthy as it just means the connection died rather early on. I have
1498 removed the panic log writing for the ECONNRESET error when getsockname()
1501 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1502 runs only) independently of the message's sender address. This meant
1503 that, if the 4xx error was in fact related to the sender, a different
1504 message to the same recipient with a different sender could confuse
1505 things. In particualar, this can happen when sending to a greylisting
1506 server, but other circumstances could also provoke similar problems.
1507 I have changed the default so that the retry time for these errors is now
1508 based a combination of the sender and recipient addresses. This change
1509 can be overridden by setting address_retry_include_sender=false in the
1512 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1513 remote server are returned as part of bounce messages. This was not
1514 happening for LMTP over a pipe (the lmtp transport), but now it is the
1515 same for both kinds of LMTP.
1517 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1518 in header lines that were in fact CNAMEs. This is no longer the case.
1520 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1521 and queue runs started by the daemon processed all messages. This has
1522 been fixed so that -R and -S can now usefully be given with -q<time>.
1524 PH/40 Import PCRE release 6.7 (fixes some bugs).
1526 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1528 PH/42 Give an error if -q is specified more than once.
1530 PH/43 Renamed the variables $interface_address and $interface_port as
1531 $received_ip_address and $received_port, to make it clear that these
1532 values apply to message reception, and not to the outgoing interface when
1533 a message is delivered. (The old names remain recognized, of course.)
1535 PH/44 There was no timeout on the connect() call when using a Unix domain
1536 socket in the ${readsocket expansion. There now is.
1538 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1539 be meaningful with "accept".
1541 SC/01 Eximstats V1.43
1542 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1544 SC/02 Eximstats V1.44
1545 Use a glob alias rather than an array ref in the generated
1546 parser. This improves both readability and performance.
1548 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1549 Collect SpamAssassin and rejection statistics.
1550 Don't display local sender or destination tables unless
1551 there is data to show.
1552 Added average volumes into the top table text output.
1554 SC/04 Eximstats V1.46
1555 Collect data on the number of addresses (recipients)
1556 as well as the number of messages.
1558 SC/05 Eximstats V1.47
1559 Added 'Message too big' to the list of mail rejection
1560 reasons (thanks to Marco Gaiarin).
1562 SC/06 Eximstats V1.48
1563 Mainlog lines which have GMT offsets and are too short to
1564 have a flag are now skipped.
1566 SC/07 Eximstats V1.49 (Alain Williams)
1567 Added the -emptyok flag.
1569 SC/08 Eximstats V1.50
1570 Fixes for obtaining the IP address from reject messages.
1572 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1573 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1574 whitesspace changes from 4.64-PH/27
1576 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1579 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1580 are found, allow negative numbers in numeric criteria)
1582 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1584 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1585 to match changes made in 4.64-PH/43
1587 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1589 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1590 "domain" is actually an IP address, and force a failure. This locks out
1591 those revolvers/nameservers that support "A-for-A" lookups, in
1592 contravention of the specifications.
1594 PH/48 When a host name was looked up from an IP address, and the subsequent
1595 forward lookup of the name timed out, the host name was left in
1596 $sender_host_name, contrary to the specification.
1598 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1599 restricted to single-key lookups, Exim was not diagnosing an error if
1600 * or *@ was used with a query-style lookup.
1602 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1604 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1605 long ago, but noone interested enough thought of it. Let's just say that
1606 the "1.1" means that there are some new functions that weren't there at
1607 some point in the past.
1609 PH/51 Error processing for expansion failure of helo_data from an smtp
1610 transport during callout processing was broken.
1612 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1613 tested/used via the -bh/-bhc/-bs options.
1615 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1616 bug, fixed in subsequent PCRE releases).
1618 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1619 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1621 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1627 SC/01 Use a glob alias rather than an array ref in eximstats generated
1628 parser. This improves both readability and performance.
1630 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1631 Don't display local sender or destination tables in eximstats unless
1632 there is data to show.
1633 Added average volumes into the eximstats top table text output.
1635 SC/03 Collect data on the number of addresses (recipients) as well
1636 as the number of messages in eximstats.
1638 TF/01 Correct an error in the documentation for the redirect router. Exim
1639 does (usually) call initgroups() when daemonizing.
1641 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1642 with consistent privilege compared to when running as a daemon.
1644 TF/03 Note in the spec that $authenticated_id is not set for local
1645 submissions from trusted users.
1647 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1648 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1650 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1651 by adding some example configuration directives to the default
1652 configuration file. A little bit of work is required to uncomment the
1653 directives and define how usernames and passwords are checked, but
1654 there is now a framework to start from.
1656 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1657 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1658 without this. I don't know how relevant this is to other LDAP libraries.
1660 PH/02 Add the verb name to the "unknown ACL verb" error.
1662 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1664 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1666 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1667 directory not expanded when it should be if an expanded home directory
1668 was set for the address (which is overridden by the transport).
1670 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1673 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1674 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1675 because it is too late at that time, and has no effect.
1677 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1678 security issue with \' (bugzilla #107). I could not use the
1679 PQescapeStringConn() function, because it needs a PGconn value as one of
1682 PH/08 When testing addresses using -bt, indicate those final addresses that
1683 are duplicates that would not cause an additional delivery. At least one
1684 person was confused, thinking that -bt output corresponded to deliveries.
1685 (Suppressing duplicates isn't a good idea as you lose the information
1686 about possibly different redirections that led to the duplicates.)
1688 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1689 systems where poll() doesn't work, in particular OS X.
1691 PH/10 Added more information to debugging output for retry time not reached.
1693 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1694 operations in malware.c.
1696 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1699 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1700 syslog_duplication set false, log lines that would normally be written
1701 both the the main log and to the reject log were not written to syslog at
1704 PH/14 In the default configuration, change the use of "message" in ACL warn
1705 statements to "add_header".
1707 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1708 not followed by a command (e.g. "seen endif").
1710 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1711 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1714 PH/17 Added extra conditions to the default value of delay_warning_condition
1718 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1719 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1720 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1723 The Auto-Submitted: and various List- headers are standardised, whereas I
1724 don't think Precedence: ever was.
1726 PH/18 Refactored debugging code in route_finduser() to show more information,
1727 in particular, the error code if getpwnam() issues one.
1729 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1730 This is apparently needed in addition to the PH/07 change above to avoid
1731 any possible encoding problems.
1733 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1734 but not after initializing Perl.
1736 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1737 output them only if debugging. By default they are written stderr,
1738 apparently, which is not desirable.
1740 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1743 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1746 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1748 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1749 authenticated or an ident call has been made. Suppress the default
1750 values for $authenticated_id and $authenticated_sender (but permit -oMai
1751 and -oMas) when testing with -bh.
1753 PH/24 Re-jigged the order of the tests in the default configuration so that the
1754 tests for valid domains and recipients precede the DNS black list and CSA
1755 tests, on the grounds that those ones are more expensive.
1757 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1758 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1759 This bug exists in every version of Exim that I still have, right back to
1762 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1763 However, an attempt to turn on case-sensitivity in a regex key by
1764 including (?-i) didn't work because the subject string was already
1765 lowercased, and the effects were non-intuitive. It turns out that a
1766 one-line patch can be used to allow (?-i) to work as expected.
1772 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1773 other effects) broke the use of negated acl sub-conditions.
1775 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1778 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1779 "Deny" causes Exim to reject the incoming connection with a 554 error.
1780 Unfortunately, if there is a major crisis, such as a disk failure,
1781 tcp-wrappers gives "deny", whereas what one would like would be some
1782 kind of temporary error. A kludge has been added to help with this.
1783 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1784 554 error is used if errno is still zero or contains ENOENT (which occurs
1785 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1788 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1790 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1791 errors. Otherwise a message that provokes a temporary error (when other
1792 messages do not) can cause a whole host to time out.
1794 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1795 addresses were routed directly to files or pipes from a redirect router.
1796 File deliveries just didn't batch; pipe deliveries might have suffered
1799 PH/06 A failure to get a lock for a hints database would erroneously always say
1800 "Failed to get write lock", even when it was really a read lock.
1802 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1803 of 0600. This has been changed to use the value of the lockfile_mode
1804 option (which defaults to 0600).
1806 PH/08 Applied small patch from the Sieve maintainer.
1808 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1809 folder from quota calculations, a direct delivery into this folder messed
1810 up the contents of the maildirsize file. This was because the regex was
1811 used only to exclude .Trash (or whatever) when the size of the mailbox
1812 was calculated. There was no check that a delivery was happening into an
1813 excluded directory. This bug has been fixed by ignoring all quota
1814 processing for deliveries into excluded directories.
1816 PH/10 Added the maildirfolder_create_regex option to appendfile.
1822 PH/01 The code for finding all the local interface addresses on a FreeBSD
1823 system running IPv6 was broken. This may well have applied to all BSD
1824 systems, as well as to others that have similar system calls. The broken
1825 code found IPv4 interfaces correctly, but gave incorrect values for the
1826 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1827 that it would not match correctly against @[] and not recognize the IPv6
1830 PH/02 The ipliteral router was not recognizing addresses of the form user@
1831 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1833 PH/03 Added disable_ipv6.
1835 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1836 decoded form, because it is most often used to construct To: headers
1837 lines in autoreplies, and the decoded form may well be syntactically
1838 invalid. However, $reply_address has leading white space removed, and all
1839 newlines turned into spaces so that the autoreply transport does not
1842 PH/05 If group was specified without a user on a router, and no group or user
1843 was specified on a transport, the group from the router was ignored.
1845 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1846 for visible compile-time settings that can be used to change these
1847 numbers, for those that want even more. Backwards compatibility with old
1848 spool files has been maintained. However, going back to a previous Exim
1849 release will lost any variables that are in spool files.
1851 PH/07 Two small changes when running in the test harness: increase delay when
1852 passing a TCP/IP connection to a new process, in case the original
1853 process has to generate a bounce, and remove special handling of
1854 127.0.0.2 (sic), which is no longer necessary.
1856 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1857 be the same on different OS.
1859 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1862 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1863 whether --show-vars was specified or not
1865 JJ/02 exipick: Added support for new ACL variable spool format introduced
1868 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1869 syntactically invalid From: or Reply-to: line, and a filter used this to
1870 generate an autoreply, and therefore failed to obtain an address for the
1871 autoreply, Exim could try to deliver to a non-existent relative file
1872 name, causing unrelated and misleading errors. What now happens is that
1873 it logs this as a hard delivery error, but does not attempt to create a
1876 PH/11 The exinext utility has a -C option for testing purposes, but although
1877 the given file was scanned by exinext itself; it wasn't being passed on
1878 when Exim was called.
1880 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1881 an end-of-file indication when reading a command response.
1883 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1884 compiled. In many other places in Exim, IPv6 addresses are always
1885 recognized, so I have changed this. It also means that IPv4 domain
1886 literals of the form [IPV4:n.n.n.n] are now always recognized.
1888 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1889 used if the router is not running as root, for example, when verifying at
1890 ACL time, or when using -bh. The debugging output from this situation was
1891 non-existent - all you got was a failure to exec. I have made two
1894 (a) Failures to set uid/gid, the current directory, or a process leader
1895 in a subprocess such as that created by queryprogram now generate
1896 suitable debugging ouput when -d is set.
1898 (b) The queryprogram router detects when it is not running as root,
1899 outputs suitable debugging information if -d is set, and then runs
1900 the subprocess without attempting to change uid/gid.
1902 PH/15 Minor change to Makefile for building test_host (undocumented testing
1905 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1906 additional section of a DNS packet that returns MX or SRV records.
1907 Instead, it always explicitly searches for A/AAAA records. This avoids
1908 major problems that occur when a DNS server includes only records of one
1909 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1910 fixed another bug: if SRV records were looked up and the corresponding
1911 address records were *not* found in the additional section, the port
1912 values from the SRV records were lost.
1914 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1915 using the correct key (the original address) when searching the retry
1916 rules in order to find which one to use for generating the retry hint.
1918 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1919 adding the default one. Similarly, if it contains a Reply-To: header, the
1920 errors_reply_to option, if set, is not used.
1922 PH/19 When calculating a retry time, Exim used to measure the "time since
1923 failure" by looking at the "first failed" field in the retry record. Now
1924 it does not use this if it is later than than the arrival time of the
1925 message. Instead it uses the arrival time. This makes for better
1926 behaviour in cases where some deliveries succeed, thus re-setting the
1927 "first failed" field. An example is a quota failure for a huge message
1928 when small messages continue to be delivered. Without this change, the
1929 "time since failure" will always be short, possible causing more frequent
1930 delivery attempts for the huge message than are intended.
1931 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1933 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1934 $1, $2, $3) because the numerical variables can be reset during some
1935 expansion items (e.g. "match"), thereby losing the authentication data.
1937 PH/21 Make -bV show the size of off_t variables so that the test suite can
1938 decide whether to run tests for quotas > 2G.
1940 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1941 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1942 filecount value is greater than 2G or if a quota value is greater than 2G
1943 on a system where the size of off_t is not greater than 4, a panic error
1946 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1947 never match. The debug and -bh output now contains an explicit error
1948 message indicating a malformed IPv4 address or mask.
1950 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1951 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1952 PH/23 above applies.
1954 PH/25 Do not write to syslog when running in the test harness. The only
1955 occasion when this arises is a failure to open the main or panic logs
1956 (for which there is an explicit test).
1958 PH/26 Added the /no_tell option to "control=freeze".
1960 PH/27 If a host name lookup failed very early in a connection, for example, if
1961 the IP address matched host_lookup and the reverse lookup yielded a name
1962 that did not have a forward lookup, an error message of the form "no IP
1963 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1964 could be logged. Now it outputs the IP address instead of "NULL".
1966 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1967 allows the sender and the authenticated sender to be set when
1968 submitting a message from within Exim. Since child_open_exim() is
1969 documented for local_scan(), the new function should be too.
1971 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1972 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1973 results in an empty string is now treated as unset.
1975 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1977 PH/31 Added sender_verify_fail logging option.
1979 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1980 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1981 tidied the source and removed it altogether.
1983 PH/33 When a queue run was abandoned because the load average was too high, a
1984 log line was always written; now it is written only if the queue_run log
1985 selector is set. In addition, the log line for abandonment now contains
1986 information about the queue run such as the pid. This is always present
1987 in "start" and "stop" lines but was omitted from the "abandon" line.
1989 PH/34 Omit spaces between a header name and the colon in the error message that
1990 is given when verify = headers_syntax fails (if there are lots of them,
1991 the message gets confusing).
1993 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1994 names, as there are now some PTR records that contain slashes. This check
1995 is only to protect against broken name servers that fall over on strange
1996 characters, so the fact that it applies to all lookups doesn't matter.
1998 PH/36 Now that the new test suite is complete, we can remove some of the
1999 special code in Exim that was needed for the old test suite. For example,
2000 sorting DNS records because real resolvers return them in an arbitrary
2001 order. The new test suite's fake resolver always returns records in the
2004 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2005 bounces) except when queue_only is set, to avoid logging races between
2006 the different processes.
2008 PH/38 Panic-die if .include specifies a non-absolute path.
2010 PH/39 A tweak to the "H" retry rule from its user.
2012 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2013 a label. They prevented compilation on older perls.
2015 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2016 a warning to be raised on newish perls.
2018 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2019 on queue. Changes to match documented behaviour of showing count of
2020 messages matching specified criteria.
2022 PH/40 Changed the default ident timeout from 30s to 5s.
2024 PH/41 Added support for the use of login_cap features, on those BSD systems
2025 that have them, for controlling the resources used by pipe deliveries.
2027 PH/42 The content-scanning code uses fopen() to create files in which to put
2028 message data. Previously it was not paying any attention to the mode of
2029 the files. Exim runs with umask(0) because the rest of the code creates
2030 files with open(), and sets the required mode explicitly. Thus, these
2031 files were ending up world-writeable. This was not a big issue, because,
2032 being within the spool directory, they were not world-accessible. I have
2033 created a function called modefopen, which takes an additional mode
2034 argument. It sets umask(777), creates the file, chmods it to the required
2035 mode, then resets the umask. All the relevant calls to fopen() in the
2036 content scanning code have been changed to use this function.
2038 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2039 to 24 hours. This avoids potential overflow problems when processing G
2040 and H retry rules. I suspect nobody ever tinkers with this value.
2042 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2044 PH/45 When the plaintext authenticator is running as a client, the server's
2045 challenges are checked to ensure they are valid base64 strings. By
2046 default, the authentication attempt is cancelled if an invalid string is
2047 received. Setting client_ignore_invalid_base64 true ignores these errors.
2048 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2049 they are received. Thus, the responses can be made to depend on the
2050 challenges. If an invalid string is ignored, an empty string is placed in
2053 PH/46 Messages that are created by the autoreply transport now contains a
2054 References: header, in accordance with RFCs 2822 and 3834.
2056 PH/47 Added authenticated_sender_force to the smtp transport.
2058 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2060 PH/49 Installed latest patch from the Sieve maintainer.
2062 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2063 was also set, the appendfile transport was unnecessarily scanning a
2064 directory of message files (e.g. for maildir delivery) to find the count
2065 of files (along with the size), even though it did not need this
2066 information. It now does the scan only if it needs to find either the
2067 size of the count of files.
2069 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2071 PH/52 Two bugs concerned with error handling when the smtp transport is
2074 (i) Exim was not creating retry information for temporary errors given
2075 for individual recipients after the DATA command when the smtp transport
2076 was used in LMTP mode. This meant that they could be retried too
2077 frequently, and not timed out correctly.
2079 (ii) Exim was setting the flag that allows error details to be returned
2080 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2081 recipients that were returned after the DATA command.
2083 PH/53 This is related to PH/52, but is more general: for any failing address,
2084 when detailed error information was permitted to be returned to the
2085 sender, but the error was temporary, then after the final timeout, only
2086 "retry timeout exceeded" was returned. Now it returns the full error as
2087 well as "retry timeout exceeded".
2089 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2090 do this, and (what is worse) MTAs that accept it.
2092 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2093 will now be deprecated.
2095 PH/56 New os.c-cygwin from the Cygwin maintainer.
2097 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2098 formats (previously only available in exim formats via -bpr, -bpru,
2099 and -bpra. Now also available in native and exiqgrep formats)
2101 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2102 with very large, slow to parse queues
2104 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2106 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2108 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2109 responses to authentication challenges, though it was showing the
2110 challenges; (ii) I've removed the CR characters from the debug output for
2113 PH/58 Allow for the insertion of a newline as well as a space when a string
2114 is turned into more than one encoded-word during RFC 2047 encoding. The
2115 Sieve code now uses this.
2117 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2118 data_4xx, lost_connection, tls_required.
2120 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2121 message was being sent as an SMTP response.
2123 PH/61 Add -l and -k options to exicyclog.
2125 PH/62 When verifying, if an address was redirected to one new address, so that
2126 verification continued, and the new address failed or deferred after
2127 having set something in $address_data, the value of $address_data was not
2128 passed back to the ACL. This was different to the case when no
2129 redirection occurred. The value is now passed back in both cases.
2131 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2132 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2133 use login_cap.h, so on its own it isn't the distinguishing feature. The
2134 new name refers directly to the setclassresources() function.
2136 PH/65 Added configuration files for NetBSD3.
2138 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2140 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2141 is preferred over IPv4.
2143 PH/68 The bounce_return_message and bounce_return_body options were not being
2144 honoured for bounces generated during the reception of non-SMTP messages.
2145 In particular, this applied to messages rejected by the ACL. This bug has
2146 been fixed. However, if bounce_return_message is true and bounce_return_
2147 body is false, the headers that are returned for a non-SMTP message
2148 include only those that have been read before the error was detected.
2149 (In the case of an ACL rejection, they have all been read.)
2151 PH/69 The HTML version of the specification is now built in a directory called
2152 spec_html instead of spec.html, because the latter looks like a path with
2153 a MIME-type, and this confuses some software.
2155 PH/70 Catch two compiler warnings in sieve.c.
2157 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2158 function verify_get_ident() calls ip_connect() to connect a socket, but
2159 if the "connect()" function timed out, ip_connect() used to close the
2160 socket. However, verify_get_ident() also closes the socket later, and in
2161 between Exim writes to the log, which may get opened at this point. When
2162 the socket was closed in ip_connect(), the log could get the same file
2163 descriptor number as the socket. This naturally causes chaos. The fix is
2164 not to close the socket in ip_connect(); the socket should be closed by
2165 the function that creates it. There was only one place in the code where
2166 this was missing, in the iplookup router, which I don't think anybody now
2167 uses, but I've fixed it anyway.
2169 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2170 well as to direct DNS lookups. Otherwise the handling of names in host
2171 lists is inconsistent and therefore confusing.
2177 PH/01 Two changes to the default runtime configuration:
2179 (1) Move the checks for relay_from_hosts and authenticated clients from
2180 after to before the (commented out) DNS black list checks.
2182 (2) Add control=submission to the relay_from_hosts and authenticated
2183 clients checks, on the grounds that messages accepted by these
2184 statements are most likely to be submissions.
2186 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2188 (1) Generate an error if the third argument for the ${prvs expansion is
2191 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2194 (3) Reset the variables that are obtained from the first argument of
2195 ${prvscheck and used in the second argument before leaving the code,
2196 because their memory is reclaimed, so using them afterwards may do
2199 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2200 one (it's much easier than Tom thought :-).
2202 (5) Because of (4), we can now allow for the use of $prvscheck_result
2203 inside the third argument.
2205 PH/03 For some reason, the default setting of PATH when running a command from
2206 a pipe transport was just "/usr/bin". I have changed it to
2209 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2210 anything to be listed in the output from -bV.
2212 PH/05 When a filter generated an autoreply, the entire To: header line was
2213 quoted in the delivery log line, like this:
2215 => >A.N.Other <ano@some.domain> <original@ddress> ...
2217 This has been changed so that it extracts the operative address. There
2218 may be more than one such address. If so, they are comma-separated, like
2221 => >ano@some.domain,ona@other.domain <original@ddress> ...
2223 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2224 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2225 not being looked up in the rDNS to get a host name, Exim was showing the
2226 IP address twice in Received: lines, even though the IP addresses were
2227 identical. For example:
2229 Received: from [1.2.3.4] (helo=[1.2.3.4])
2231 However, if the real host name was known, it was omitting the HELO data
2232 if it matched the actual IP address. This has been tidied up so that it
2233 doesn't show the same IP address twice.
2235 PH/07 When both +timestamp and +memory debugging was on, the value given by
2236 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2237 called by the debug printing, thereby overwriting the timestamp buffer.
2238 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2240 PH/08 When the original message was included in an autoreply transport, it
2241 always said "this is a copy of the message, including all the headers",
2242 even if body_only or headers_only was set. It now gives an appropriate
2245 PH/09 Applied a patch from the Sieve maintainer which:
2247 o fixes some comments
2248 o adds the (disabled) notify extension core
2249 o adds some debug output for the result of if/elsif tests
2250 o points to the current vacation draft in the documentation
2251 and documents the missing references header update
2255 o fixes a bug in processing the envelope test (when testing
2256 multiple envelope elements, the last element determinted the
2259 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2260 Electronic Mail") by including:
2262 Auto-submitted: auto-generated
2264 in the messages that it generates (bounce messages and others, such as
2265 warnings). In the case of bounce messages for non-SMTP mesages, there was
2266 also a typo: it was using "Auto_submitted" (underscore instead of
2267 hyphen). Since every message generated by Exim is necessarily in response
2268 to another message, thes have all been changed to:
2270 Auto-Submitted: auto-replied
2272 in accordance with these statements in the RFC:
2274 The auto-replied keyword:
2276 - SHOULD be used on messages sent in direct response to another
2277 message by an automatic process,
2279 - MUST NOT be used on manually-generated messages,
2281 - MAY be used on Delivery Status Notifications (DSNs) and Message
2282 Disposition Notifications (MDNs),
2284 - MUST NOT be used on messages generated by automatic or periodic
2285 processes, except for messages which are automatic responses to
2288 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2289 to the default Received: header definition.
2291 PH/12 Added log selector acl_warn_skipped (default on).
2293 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2294 variables because (a) they are in the wrong storage pool and (b) even if
2295 they were copied, it wouldn't work properly because of the caching.
2297 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2298 checking when decoding. Apparently there are clients that generate
2299 overlong encoded strings. Why am I not surprised?
2301 PH/15 If the first argument of "${if match_address" was not empty, but did not
2302 contain an "@" character, Exim crashed. Now it writes a panic log message
2303 and treats the condition as false.
2305 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2307 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2308 extension "envelope-auth". The code is finished and in agreement with
2309 other implementations, but there is no documentation so far and in fact,
2310 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2311 not changing the active code.
2313 Print executed "if" and "elsif" statements when debugging is used. This
2314 helps a great deal to understand what a filter does.
2316 Document more things not specified clearly in RFC3028. I had all this
2317 sorted out, when out of a sudden new issues came to my mind. Oops."
2319 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2322 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2323 canonical form (as documented). However, after a host name lookup from
2324 the IP address, check_host() was doing a simple string comparison with
2325 addresses acquired from the DNS when checking that the found name did
2326 have the original IP as one of its addresses. Since any found IPv6
2327 addresses are likely to be in abbreviated form, the comparison could
2328 fail. Luckily, there already exists a function for doing the comparison
2329 by converting both addresses to binary, so now that is used instead of
2330 the text comparison.
2332 PH/20 There was another similar case to PH/19, when a complete host name was
2333 given in a host list; looking up its IP address could give an abbreviated
2334 form, whereas the current host's name might or might not be abbreviated.
2335 The same fix has been applied.
2341 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2342 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2345 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2346 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2348 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2350 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2351 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2352 Also, for "Auto-Submitted", treat anything other than "no" as
2353 non-personal, in accordance with RFC 3834. (Previously it treated
2354 anything starting "auto-" as non-personal.)
2356 TF/01 The control=submission/name=... option had a problem with syntax
2357 errors if the name included a slash character. The /name= option
2358 now slurps the rest of the string, so it can include any characters
2359 but it must come last in the list of options (after /sender_retain
2362 PH/05 Some modifications to the interface to the fake nameserver for the new
2370 TK/01 Added the "success_on_redirect" address verification option. See
2371 NewStuff for rationale and an example.
2373 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2375 PH/02 Patch to exigrep to allow it to work on syslog lines.
2377 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2378 fread() to skip over the body file's header line, because in Cygwin the
2379 header line is locked and is inaccessible.
2381 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2382 co-exist for some time) to make it clear that it is the Exim ID that is
2383 referenced, not the Message-ID: header line.
2385 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2386 string_format() function, because snprintf() does not exist on all
2389 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2390 ${stat: expansion item.
2392 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2393 protocol synchronization error", to keep the pedants happy.
2395 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2396 well as for IRIX systems, when gcc is being used. See the host.c source
2399 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2401 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2404 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2405 transport and to the smtp transport in LMTP mode.
2407 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2409 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2410 run a filter in a subprocess. This could lead to confusion in subsequent
2411 lookups in the parent process. There should also be a search_tidyup() at
2412 the end of the subprocess.
2414 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2415 only if the host matched helo_try_verify_hosts, which caused the
2416 verification to occur when the EHLO/HELO command was issued. The ACL just
2417 tested the remembered result. Now, if a previous verification attempt has
2418 not happened, "verify = helo" does it there and then.
2420 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2422 TK/03 Fix log output including CR from clamd.
2424 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2425 exist provoked a memory error which could cause a segfault.
2427 PH/15 Installed PCRE 6.2
2429 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2431 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2432 of the problem. Specifically, suggested +O2 rather than +O1 for the
2435 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2437 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2438 record, the port was not correctly logged when the outgoing_port log
2439 selector was set (it logged the transort's default port).
2441 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2442 fallback_hosts, and "hosts" in the smtp transport.
2444 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2445 host errors such as "Connection refused".
2447 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2448 authentication with radiusclient 0.4.9:
2450 - Error returned from rc_read_config was caught wrongly
2451 - Username/password not passed on to radius server due to wrong length.
2453 The presumption is that some radiusclient API changes for 4.51/PH/17
2454 were not taken care of correctly. The code is still untested by me (my
2455 Linux distribution still has 0.3.2 of radiusclient), but it was
2456 contributed by a Radius user.
2458 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2459 expanding the "port" option of the smtp transport.
2461 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2462 while reading a MIME header. Thanks to Tom Hughes for a patch.
2464 PH/24 Include config.h inside local_scan.h so that configuration settings are
2467 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2468 that in an ACL for RCPT (for example), you can examine exactly what was
2471 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2472 commands, but it was not correctly comparing the address with the actual
2473 client host address. Thus, it would show the EHLO address in Received:
2474 header lines when this was not necessary.
2476 PH/27 Added the % operator to ${eval:}.
2478 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2479 it should be ignoring failures (because with -C, for example, it has lost
2480 privilege). It wasn't ignoring creation failures other than "already
2483 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2486 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2487 because an input line was too long, either on its own, or by virtue of
2488 too many continuations, the temporary file was not being removed, and the
2489 return code was incorrect.
2491 PH/31 Missing "BOOL" in function definition in filtertest.c.
2493 PH/32 Applied Sieve patches from the maintainer.
2495 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2497 PH/33 Added "verify = not_blind".
2499 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2500 Local/Makefile (with some defaults set). These are used in built scripts
2501 such as exicyclog, but they have never been used in the exim_install
2502 script (though there are many overriding facilities there). I have
2503 arranged that the exim_install script now takes note of these two
2506 PH/35 Installed configuration files for Dragonfly.
2508 PH/36 When a locally submitted message by a trusted user did not contain a
2509 From: header, and the sender address was obtained from -f or from an SMTP
2510 MAIL command, and the trusted user did not use -F to supply a sender
2511 name, $originator_name was incorrectly used when constructing a From:
2512 header. Furthermore, $originator_name was used for submission mode
2513 messages from external hosts without From: headers in a similar way,
2514 which is clearly wrong.
2516 PH/37 Added control=suppress_local_fixups.
2518 PH/38 When log_selector = +received_sender was set, and the addition of the
2519 sender made the log line's construction buffer exactly full, or one byte
2520 less than full, an overflow happened when the terminating "\n" was
2523 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2524 when the result of a list match is failure because a DNS lookup failed.
2526 PH/40 RM_COMMAND is now used in the building process.
2528 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2529 the "build-* directories that it finds.
2531 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2532 address in a domain literal was a prefix of an interface address.
2534 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2535 when verifying a sender address, unless rewrite_headers is false.
2537 PH/44 Wrote a long comment about why errors_to addresses are verified as
2538 recipients, not senders.
2540 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2541 the ratelimit ACL was added.
2543 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2545 PH/46 Added extra information about PostgreSQL errors to the error string.
2547 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2548 avoiding the need to install special zones in a real server. This is
2549 backwards compatible; if it can't find the fake resolver, it drops back.
2550 Thus, both old and new test suites can be run.
2552 TF/02 Added util/ratelimit.pl
2554 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2555 clock is set back in time.
2557 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2558 Candler <B.Candler@pobox.com>.
2560 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2561 OK for addresses that are the result of redirections.
2563 PH/48 A number of further additions for the benefit of the new test suite,
2564 including a fake gethostbyname() that interfaces to the fake DNS resolver
2567 TF/06 The fix for widen_domains has also been applied to qualify_single and
2568 search_parents which are the other dnslookup options that can cause
2571 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2574 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2576 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2577 service" if it encountered an MX record with an empty target hostname.
2578 The message is now "an MX or SRV record indicated no SMTP service".
2580 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2581 if the DNS of the sending site is misconfigured. This is quite a
2582 common situation. This change restores the behaviour of treating a
2583 helo verification defer as a failure.
2585 PH/51 If self=fail was set on a router, the bounce message did not include the
2586 actual error message.
2592 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2594 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2595 command itself ended in error, the underlying message about the transport
2596 filter timeout was being overwritten with the pipe command error. Now the
2597 underlying error message should be appended to the second error message.
2599 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2601 PH/02 Reduce the amount of output that "make" produces by default. Full output
2602 can still be requested.
2604 PH/03 The warning log line about a condition test deferring for a "warn" verb
2605 was being output only once per connection, rather than after each
2606 occurrence (because it was using the same function as for successful
2607 "warn" verbs). This seems wrong, so I have changed it.
2609 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2610 it should not have, which might have caused a crash in the right
2611 circumstances, but probably never did.
2613 PH/04 Installed a modified version of Tony Finch's patch to make submission
2614 mode fix the return path as well as the Sender: header line, and to
2615 add a /name= option so that you can make the user's friendly name appear
2618 TF/03 Added the control = fakedefer ACL modifier.
2620 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2621 Mark Lowes for thorough testing.
2623 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2625 TK/03 Merged latest SRS patch from Miles Wilton.
2627 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2628 with the definition in sysexits.h (which is #included earlier).
2629 Fortunately, Exim does not actually use EX_OK. The code used to try to
2630 preserve the sysexits.h value, by assumimg that macro definitions were
2631 scanned for macro replacements. I have been disabused of this notion,
2632 so now the code just undefines EX_OK before #including unistd.h.
2634 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2635 in the smtp transport. When a block could not be written in a single
2636 write() function, the timeout was being re-applied to each part-write.
2637 This seems wrong - if the receiver was accepting one byte at a time it
2638 would take for ever. The timeout is now adjusted when this happens. It
2639 doesn't have to be particularly precise.
2641 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2642 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2644 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2645 without a domain if the check to <postmaster@domain> fails.
2647 SC/01 Eximstats: added -xls and the ability to specify output files
2648 (patch written by Frank Heydlauf).
2650 SC/02 Eximstats: use FileHandles for outputing results.
2652 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2654 SC/04 Eximstats: fixed display of large numbers with -nvr option
2656 SC/05 Eximstats: fixed merging of reports with empty tables.
2658 SC/06 Eximstats: added the -include_original_destination flag
2660 SC/07 Eximstats: removed tabs and trailing whitespace.
2662 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2664 TK/06 MBOX spool code: Add real "From " MBOX separator line
2665 so the .eml file is really in mbox format (even though
2666 most programs do not really care). Patch from Alex Miller.
2668 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2669 The latter is generated from $received_to and is only set if the
2670 message has one envelope recipient. SA can use these headers,
2671 obviously out-of-the-box. Patch from Alex Miller.
2673 PH/08 The ${def test on a variable was returning false if the variable's
2674 value was "0", contrary to what the specification has always said!
2675 The result should be true unless the variable is empty.
2677 PH/09 The syntax error of a character other than { following "${if
2678 def:variable_name" (after optional whitespace) was not being diagnosed.
2679 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2680 accidental colon was present, for example, could give incorrect results.
2682 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2683 result is used (not including appendfile, where other changes are about
2686 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2687 This involved changing a lot of size variables from int to off_t. It
2688 should work with maildirs and everything.
2690 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2691 spamd dying while we are connected to it.
2693 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2696 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2697 function for BDB 4.3.
2699 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2701 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2702 the sources for PCRE have been re-arranged and more files are now
2705 PH/15 The code I had for printing potentially long long variables in PH/11
2706 above was not the best (it lost precision). The length of off_t variables
2707 is now inspected at build time, and an appropriate printing format (%ld
2708 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2709 to be "long long int" or "long int". This is needed for the internal
2710 formatting function string_vformat().
2712 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2713 the configuration file to be ":syslog", then the script "guesses" where
2714 the logs files are, rather than using the compiled in default. In our
2715 case the guess is not the same as the compiled default, so the script
2716 suddenly stopped working when I started to use syslog. The patch checks
2717 to see if log_file_path is "". If so, it attempts to read it from exim
2718 with no configuration file to get the compiled in version, before it
2719 falls back to the previous guessing code."
2721 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2722 implementing BATV in an Exim configuration. See NewStuff for the gory
2725 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2726 Makefile that are specific to HP-UX.
2728 PH/18 If the "use_postmaster" option was set for a recipient callout together
2729 with the "random" option, the postmaster address was used as the MAIL
2730 FROM address for the random test, but not for the subsequent recipient
2731 test. It is now used for both.
2733 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2734 patch removes a few documentation additions to RFC 3028, because the
2735 latest draft now contains them. It adds the new en;ascii-case comparator
2736 and a new error check for 8bit text in MIME parts. Comparator and
2737 require names are now matched exactly. I enabled the subaddress
2738 extension, but it is not well tested yet (read: it works for me)."
2740 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2741 rework some of the code of TK/09 above to avoid the hardwired use of
2742 "%lld" and "long long". Replaced the call to snprintf() with a call to
2745 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2746 records point to non-existent hosts", "retry timeout exceeded", and
2747 "retry time not reached for any host after a long failure period".
2749 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2750 experimental DomainKeys support:
2752 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2753 (2) On an error such as an illegally used "control", the wrong name for
2754 the control was given.
2756 These problems did NOT occur unless DomainKeys support was compiled.
2758 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2760 PH/24 Added ${if match_ip condition.
2762 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2763 databases so that it will be absolutely obvious if a crash occurs in the
2764 DB library. This is a regular occurrence (often caused by mis-matched
2767 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2768 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2769 picked up on a user's system that detects such things. There doesn't seem
2770 to be a gcc warning option for this - only an attribute that has to be
2771 put on the function's prototype. It seems that in Fedora Core 4 they have
2772 set this on a number of new functions. No doubt there will be more in due
2775 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2776 specify a transport. However, if an address that was verified by such a
2777 router was the subject of a callout, Exim crashed because it tried to
2778 read the rcpt_include_affixes from the non-existent transport. Now it
2779 just assumes that the setting of that option is false. This bug was
2780 introduced by 4.51/PH/31.
2782 PH/28 Changed -d+all to exclude +memory, because that information is very
2783 rarely of interest, but it makes the output a lot bigger. People tend to
2784 do -d+all out of habit.
2786 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2787 code in os-type was giving problems when libc.so lives in lib64, like on
2790 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2791 aren't the modern standard, and it seems that some systems' include files
2792 don't always have them. Exim was already checking for some of the newer
2793 ones like T_AAAA, and defining it itself. I've added checks for all the
2794 record types that Exim uses.
2796 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2797 not automatically generating a new one, as it is supposed to. This
2798 prevented TLS from working. If the file did exist, but contained invalid
2799 data, a new version was generated, as expected. It was only the case of a
2800 non-existent file that was broken.
2802 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2803 with a change in libdomainkeys > 0.64.
2805 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2806 from DNS. If the selector record carries the flag, it now has
2807 precedence over the domain-wide flag.
2809 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2811 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2812 the use of a port name that isn't defined in /etc/services) occurred, the
2813 message was deferred as in a normal delivery, and thus remained on the
2814 spool, instead of being failed because of the mua_wrapper setting. This
2815 is now fixed, and I tidied up some of the mua_wrapper messages at the
2818 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2819 the messages in a hash of arrays rather than using individual hashes.
2820 This is a bit cleaner and results in dramatic memory savings, albeit
2821 at a slight CPU cost.
2823 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2824 as requested by Marc Sherman.
2826 SC/10 Eximstats: added histograms for user specified patterns as requested
2829 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2831 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2832 fopen() in the content-scanning modules that did not already have it.
2838 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2839 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2841 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2843 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2845 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2846 to test Sieve filters that use "vacation".
2848 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2849 that changes the way the GnuTLS parameters are stored in the cache file.
2850 The new format can be generated externally. For backward compatibility,
2851 if the data in the cache doesn't make sense, Exim assumes it has read an
2852 old-format file, and it generates new data and writes a new file. This
2853 means that you can't go back to an older release without removing the
2856 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2857 work if there are any delivery delays because "one_time" forces the
2858 parent to be marked "delivered", so its unseen clone is never tried
2859 again. For this reason, Exim now forbids the simultaneous setting of
2862 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2863 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2864 entry if you want to know the details. The fix, however, neglected to
2865 consider the case when local delivery batching is involved. The test for
2866 "previously delivered" was not happening when checking to see if an
2867 address could be batched with a previous (undelivered) one; under
2868 certain circumstances this could lead to multiple deliveries to the same
2871 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2872 in its include files, and this causes problems building Exim.
2874 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2875 header_syntax) but Exim was just ignoring anything given after a slash.
2876 In particular, this caused confusion with an attempt to use "verify =
2877 reverse_host_lookup/defer_ok". An error is now given when options are
2878 supplied for verify items that do not have them. (Maybe reverse_host_
2879 lookup should have a defer_ok option, but that's a different point.)
2881 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2882 defined by RFC 821) to 2048, because there were problems with some AUTH
2883 commands, and RFC 1869 says the size should be increased for extended
2884 SMTP commands that take arguments.
2886 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2889 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2890 "unknown" error; now it says that the functionality isn't in the binary.
2892 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2893 an address' error message when a string expansion fails (syntax or
2894 whatever). Otherwise the password may appear in the log. Following change
2895 PH/42 below, there is no longer a chance of it appearing in a bounce
2898 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2900 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2901 of its addresses was ever tried. (Bugzilla bug #2.)
2903 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2904 the result incorrectly in the debug output. (It correctly added a newline
2905 to what was transported.)
2907 TF/01 Added $received_time.
2909 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2910 commented out examples of how to interface to a virus scanner and to
2911 SpamAssassin. Also added commented examples of av_scanner and
2912 spamd_address settings.
2914 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2915 and controls are allowed in which ACLs. There were a couple of minor
2916 errors. Some of the entries in the conditions table (which is a table of
2917 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2918 negation of where the condition IS allowed.
2920 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2922 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2923 header file does not have a version number, so I've had to invent a new
2924 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2925 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2926 radiusclient), but it was contributed by a Radius user.
2928 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2929 files or directories, for OpenSSL.
2931 PH/19 When an Exim process that is running as root has to create an Exim log
2932 file, it does so in a subprocess that runs as exim:exim so as to get the
2933 ownership right at creation (otherwise, other Exim processes might see
2934 the file with the wrong ownership). There was no test for failure of this
2935 fork() call, which would lead to the process getting stuck as it waited
2936 for a non-existent subprocess. Forks do occasionally fail when resources
2937 run out. I reviewed all the other calls to fork(); they all seem to check
2940 PH/20 When checking for unexpected SMTP input at connect time (before writing
2941 the banner), Exim was not dealing correctly with a non-positive return
2942 from the read() function. If the client had disconnected by this time,
2943 the result was a log entry for a synchronization error with an empty
2944 string after "input=" when read() returned zero. If read() returned -1
2945 (an event I could not check), uninitialized data bytes were printed.
2946 There were reports of junk text (parts of files, etc) appearing after
2949 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2951 PH/22 Added support for macro redefinition, and (re)definition in between
2952 driver and ACL definitions.
2954 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2955 forgetting to use the resulting value; it was using the unexpanded value.
2957 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2958 hadn't been configured. The fix is from Juergen Kreileder, who
2959 understands it better than I do:
2961 "Here's what I see happening with three configured cyrus_sasl
2962 authenticators configured (plain, login, cram-md5):
2964 On startup auth_cyrus_sasl_init() gets called for each of these.
2965 This means three calls to sasl_listmech() without a specified mech_list.
2966 => SASL tests which mechs of all available mechs actually work
2967 => three warnings about OTP not working
2968 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2970 With the patch, sasl_listmech() also gets called three times. But now
2971 SASL's mech_list option is set to the server_mech specified in the the
2972 authenticator. Or in other words, the answer from sasl_listmech()
2973 gets limited to just the mech you're testing for (which is different
2975 => the return list contains just 'plain' or 'login', 'cram-md5' or
2976 nothing depending on the value of ob->server_mech.
2978 I've just tested the patch: Authentication still works fine,
2979 unavailable mechs specified in the exim configuration are still
2980 caught, and the auth.log warnings about OTP are gone."
2982 PH/25 When debugging is enabled, the contents of the command line are added
2983 to the debugging output, even when log_selector=+arguments is not
2986 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2987 answer is "GNU", and only if the return is "GNU/something" is the answer
2990 PH/27 $acl_verify_message is now set immediately after the failure of a
2991 verification in an ACL, and so is available in subsequent modifiers. In
2992 particular, the message can be preserved by coding like this:
2994 warn !verify = sender
2995 set acl_m0 = $acl_verify_message
2997 Previously, $acl_verify_message was set only while expanding "message"
2998 and "log_message" when a very denied access.
3000 PH/28 Modified OS/os.c-Linux with
3002 -#ifndef OS_LOAD_AVERAGE
3003 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3005 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3006 nomenclature these days.)
3008 PH/29 Installed patch from the Sieve maintainer that adds the options
3009 sieve_useraddress and sieve_subaddress to the redirect router.
3011 PH/30 In these circumstances:
3012 . Two addresses routed to the same list of hosts;
3013 . First host does not offer TLS;
3014 . First host accepts first address;
3015 . First host gives temporary error to second address;
3016 . Second host offers TLS and a TLS session is established;
3017 . Second host accepts second address.
3018 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3019 and peerdn, if requested) that were in fact used only for the second
3022 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3023 attention to any local part prefix or suffix that was matched by the
3024 router that accepted the address. It now behaves in the same way as it
3025 does for delivery: the affixes are removed from the local part unless
3026 rcpt_include_affixes is set on the transport.
3028 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3029 timeout during the DATA phase of an incoming message.
3031 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3032 applied a patch sanctioned by the Sieve maintainer.
3034 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3035 the uid or gid is negative. A case of a negative gid caused this to be
3036 noticed. The fix allows for either to be negative.
3038 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3039 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3042 PH/36 The change PH/12 above was broken. Fixed it.
3044 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3045 the grounds that routing the same address twice would always produce the
3046 same answer. This might have been true once, but it is certainly no
3047 longer true now. Routing a child address may depend on the previous
3048 routing that produced that child. Some complicated redirection strategies
3049 went wrong when messages had multiple recipients, and made Exim's
3050 behaviour dependent on the order in which the addresses were given.
3052 I have moved the duplicate checking until after the routing is complete.
3053 Exim scans the addresses that are assigned to local and remote
3054 transports, and removes any duplicates. This means that more work will be
3055 done, as duplicates will always all be routed, but duplicates are
3056 presumably rare, so I don't expect this is of any significance.
3058 For deliveries to pipes, files, and autoreplies, the duplicate checking
3059 still happens during the routing process, since they are not going to be
3062 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3063 It corrects a timeout issue with spamd. This is Ian's comment: "The
3064 background is that sometimes spamd either never reads data from a
3065 connection it has accepted, or it never writes response data. The exiscan
3066 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3067 blindly assumes that writes won't block so it may never time out."
3069 PH/39 Allow G after quota size as well as K and M.
3071 PH/40 The value set for $authenticated_id in an authenticator may not contain
3072 binary zeroes or newlines because the value is written to log lines and
3073 to spool files. There was no check on this. Now the value is run through
3074 the string_printing() function so that such characters are converted to
3075 printable escape sequences.
3077 PH/41 $message_linecount is a new variable that contains the total number of
3078 lines in the message. Compare $body_linecount, which is the count for the
3081 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3082 bounce and delay warning messages, except in certain special cases, which
3085 (a) An SMTP error message from a remote host;
3086 (b) A message specified in a :fail: redirection;
3087 (c) A message specified in a "fail" command in a system filter;
3088 (d) A message specified in a FAIL return from the queryprogram router;
3089 (e) A message specified by the cannot_route_message router option.
3091 In these cases only, Exim does include the error details in bounce and
3092 warning messages. There are also a few cases where bland messages such
3093 as "unrouteable address" or "local delivery error" are given.
3095 PH/43 $value is now also set for the "else" part of a ${run expansion.
3097 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3098 being worked on, but at least Exim now implements the latest version to
3101 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3102 process to complete was treated as a delivery failure, a timeout while
3103 writing the message to the pipe was logged, but erroneously treated as a
3104 successful delivery. Such timeouts include transport filter timeouts. For
3105 consistency with the overall process timeout, these timeouts are now
3106 treated as errors, giving rise to delivery failures by default. However,
3107 there is now a new Boolean option for the pipe transport called
3108 timeout_defer, which, if set TRUE, converts the failures into defers for
3109 both kinds of timeout. A transport filter timeout is now identified in
3112 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3113 systems where "make" and "gmake" are different, calling "gmake" at top
3114 level broke things. I've arranged for the value of $(MAKE) to be passed
3115 from the Makefile to this script so that it can call the same version of
3119 A note about Exim versions 4.44 and 4.50
3120 ----------------------------------------
3122 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3123 changes of various kinds. As a consequence, a big documentation update was
3124 needed. This delayed the release for rather longer than seemed good, especially
3125 in the light of a couple of (minor) security issues. Therefore, the changes
3126 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3127 release. So 4.44 and 4.50 are in effect two different branches that both start
3130 I have left the 4.50 change log unchanged; it contains all the changes since
3131 4.43. The change log for 4.44 is below; many of its items are identical to
3132 those for 4.50. This seems to be the most sensible way to preserve the
3133 historical information.
3139 1. Minor wording change to the doc/README.SIEVE file.
3141 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3142 computation of the current number of files was incorrect.
3144 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3145 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3148 4. Give more explanation in the error message when the command for a transport
3149 filter fails to execute.
3151 5. There are several places where Exim runs a non-Exim command in a
3152 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3153 was being done only for the command run by the queryprogram router. It is
3154 now done for all such subprocesses. The other cases are: ${run, transport
3155 filters, and the commands run by the lmtp and pipe transports.
3157 6. Added CONFIGURE_GROUP build-time option.
3159 7. Some older OS have a limit of 256 on the maximum number of file
3160 descriptors. Exim was using setrlimit() to set 1000 as a large value
3161 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3162 systems. I've change it so that if it can't get 1000, it tries for 256.
3164 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3165 was an oversight, and furthermore, ever since the addition of extra
3166 controls (e.g. 4.43/32), the checks on when to allow different forms of
3167 "control" were broken. There should now be diagnostics for all cases when a
3168 control that does not make sense is encountered.
3170 9. Added the /retain_sender option to "control=submission".
3172 10. $recipients is now available in the predata ACL (oversight).
3174 11. Tidy the search cache before the fork to do a delivery from a message
3175 received from the command line. Otherwise the child will trigger a lookup
3176 failure and thereby defer the delivery if it tries to use (for example) a
3177 cached ldap connection that the parent has called unbind on.
3179 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3180 of $address_data from the recipient verification was clobbered by the
3181 sender verification.
3183 13. The value of address_data from a sender verification is now available in
3184 $sender_address_data in subsequent conditions in the ACL statement.
3186 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3188 15. Added a new option "connect=<time>" to callout options, to set a different
3191 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3192 was its contents. (It was OK if the option was not defined at all.)
3194 17. A "Completed" log line is now written for messages that are removed from
3195 the spool by the -Mrm option.
3197 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3198 information about exactly what failed.
3200 19. Added -dd to debug only the daemon process.
3202 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3203 handles timeouts, both on the server side and network timeouts. Renamed the
3204 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3206 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3207 It is now set to "smtps".
3209 22. $host_address is now set to the target address during the checking of
3210 ignore_target_hosts.
3212 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3213 being passed; this would have caused $sender_host_name to have been used if
3214 matching the list had actually called for a host name (not very likely,
3215 since this list is usually IP addresses). A host name is now passed as
3218 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3219 code that specifies a non-restarting handler (typically sigaction() in
3220 modern systems) in an attempt to fix a rare and obscure crash bug.
3222 25. Narrowed the window for a race in the daemon that could cause it to ignore
3223 SIGCHLD signals. This is not a major problem, because they are used only to
3224 wake it up if nothing else does.
3226 26. A malformed maildirsize file could cause Exim to calculate negative values
3227 for the mailbox size or file count. Odd effects could occur as a result.
3228 The maildirsize information is now recalculated if the size or filecount
3231 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3232 support for a long time. Removed HAVE_SYS_VFS_H.
3234 28. Installed the latest version of exipick from John Jetmore.
3236 29. In an address list, if the pattern was not a regular expression, an empty
3237 subject address (from a bounce message) matched only if the pattern was an
3238 empty string. Non-empty patterns were not even tested. This was the wrong
3239 because it is perfectly reasonable to use an empty address as part of a
3240 database query. An empty address is now tested by patterns that are
3241 lookups. However, all the other forms of pattern expect the subject to
3242 contain a local part and a domain, and therefore, for them, an empty
3243 address still always fails if the pattern is not itself empty.
3245 30. Exim went into a mad DNS loop when attempting to do a callout where the
3246 host was specified on an smtp transport, and looking it up yielded more
3247 than one IP address.
3249 31. Re-factored the code for checking spool and log partition space into a
3250 function that finds that data and another that does the check. The former
3251 is then used to implement four new variables: $spool_space, $log_space,
3252 $spool_inodes, and $log_inodes.
3254 32. The RFC2047 encoding function was originally intended for short strings
3255 such as real names; it was not keeping to the 75-character limit for
3256 encoded words that the RFC imposes. It now respects the limit, and
3257 generates multiple encoded words if necessary. To be on the safe side, I
3258 have increased the buffer size for the ${rfc2047: expansion operator from
3261 33. It is now permitted to omit both strings after an "if" condition; if the
3262 condition is true, the result is "true". As before, when the second string
3263 is omitted, a false condition yields an empty string. This makes it less
3264 cumbersome to write custom ACL and router conditions.
3266 34. Failure to deliver a bounce message always caused it to be frozen, even if
3267 there was an errors_to setting on the router. The errors_to setting is now
3270 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3271 canonical form (fully expanded) before being placed in
3272 $sender_host_address.
3274 36. The table in the code that translates DNS record types into text (T_A to
3275 "A" for instance) was missing entries for NS and CNAME. It is just possible
3276 that this could have caused confusion if both these types were looked up
3277 for the same domain, because the text type is used as part of Exim's
3278 per-process caching. But the chance of anyone hitting this buglet seems
3281 37. The dnsdb lookup has been extended in a number of ways.
3283 (1) There is a new type, "zns", which walks up the domain tree until it
3284 finds some nameserver records. It should be used with care.
3286 (2) There is a new type, "mxh", which is like "mx" except that it returns
3287 just the host names, not the priorities.
3289 (3) It is now possible to give a list of domains (or IP addresses) to be
3290 looked up. The behaviour when one of the lookups defers can be
3291 controlled by a keyword.
3293 (4) It is now possible to specify the separator character for use when
3294 multiple records are returned.
3296 38. The dnslists ACL condition has been extended: it is now possible to supply
3297 a list of IP addresses and/or domains to be looked up in a particular DNS
3300 39. Added log_selector=+queue_time_overall.
3302 40. When running the queue in the test harness, wait just a tad after forking a
3303 delivery process, to get repeatability of debugging output.
3305 41. Include certificate and key file names in error message when GnuTLS fails
3306 to set them up, because the GnuTLS error message doesn't include the name
3307 of the failing file when there is a problem reading it.
3309 42. Allow both -bf and -bF in the same test run.
3311 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3313 44. The "Exiscan patch" is now merged into the mainline Exim source.
3315 45. Sometimes the final signoff response after QUIT could fail to get
3316 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3317 < 0 before doing a fflush(). This bug looks as though it goes back to the
3318 introduction of TLS in release 3.20, but "sometimes" must have been rare
3319 because the tests only now provoked it.
3321 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3322 (this can affect the format of dates).
3324 47. exim_tidydb, when checking for the continued existence of a message for
3325 which it has found a message-specific retry record, was not finding
3326 messages that were in split spool directories. Consequently, it was
3327 deleting retry records that should have stayed in existence.
3329 48. Steve fixed some bugs in eximstats.
3331 49. The SPA authentication driver was not abandoning authentication and moving
3332 on to the next authenticator when an expansion was forced to fail,
3333 contradicting the general specification for all authenticators. Instead it
3334 was generating a temporary error. It now behaves as specified.
3336 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3337 (the order specifies the preference for clients). The order is now AES256,
3338 AES128, 3DES, ARCFOUR128.
3340 51. Small patch to Sieve code - explicitly set From: when generating an
3343 52. Exim crashed if a remote delivery caused a very long error message to be
3344 recorded - for instance if somebody sent an entire SpamAssassin report back
3345 as a large number of 550 error lines. This bug was coincidentally fixed by
3346 increasing the size of one of Exim's internal buffers (big_buffer) that
3347 happened as part of the Exiscan merge. However, to be on the safe side, I
3348 have made the code more robust (and fixed the comments that describe what
3351 53. Now that there can be additional text after "Completed" in log lines (if
3352 the queue_time_overall log selector is set), a one-byte patch to exigrep
3353 was needed to allow it to recognize "Completed" as not the last thing in
3356 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3357 patch that reportedly fixes this has been added. I am not expert enough to
3358 create a test for it. This is what the patch creator wrote:
3360 "I found a little strange behaviour of ldap code when working with
3361 Windows 2003 AD Domain, where users was placed in more than one
3362 Organization Units. When I tried to give exim partial DN, the exit code
3363 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3364 But simultaneously result of request was absolutely normal ldap result,
3365 so I produce this patch..."
3367 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3368 REFERENCE, so I have modified the code to exclude the patch when that macro
3371 55. Some experimental protocols are using DNS PTR records for new purposes. The
3372 keys for these records are domain names, not reversed IP addresses. The
3373 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3374 leaves it alone. Component reversal etc. now happens only for IP addresses.
3377 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3379 57. Double the size of the debug message buffer (to 2048) so that more of very
3380 long debug lines gets shown.
3382 58. The exicyclog utility now does better if the number of log files to keep
3383 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3385 59. Two changes related to the smtp_active_hostname option:
3387 (1) $smtp_active_hostname is now available as a variable.
3388 (2) The default for smtp_banner uses $smtp_active_hostname instead
3389 of $primary_hostname.
3391 60. The host_aton() function is supposed to be passed a string that is known
3392 to be a valid IP address. However, in the case of IPv6 addresses, it was
3393 not checking this. This is a hostage to fortune. Exim now panics and dies
3394 if the condition is not met. A case was found where this could be provoked
3395 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3396 components; fortuitously, this particular loophole had already been fixed
3397 by change 4.50/55 above.
3399 If there are any other similar loopholes, the new check in host_aton()
3400 itself should stop them being exploited. The report I received stated that
3401 data on the command line could provoke the exploit when Exim was running as
3402 exim, but did not say which command line option was involved. All I could
3403 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3404 running as the user.
3407 61. There was a buffer overflow vulnerability in the SPA authentication code
3408 (which came originally from the Samba project). I have added a test to the
3409 spa_base64_to_bits() function which I hope fixes it.
3412 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3413 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3415 63. The daemon start-up calls getloadavg() while still root for those OS that
3416 need the first call to be done as root, but it missed one case: when
3417 deliver_queue_load_max is set with deliver_drop_privilege. This is
3418 necessary for the benefit of the queue runner, because there is no re-exec
3419 when deliver_drop_privilege is set.
3421 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3422 This has been fixed.
3424 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3425 was in use, was not putting the data itself into the right store pool;
3426 consequently, it could be overwritten for a subsequent message in the same
3427 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3430 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3432 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3433 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3434 the function were treating the return as a boolean value, which happened to
3435 work because 0=false and not-0=true, but is not correct code.
3437 68. The host_aton() function was not handling scoped IPv6 addresses (those
3438 with, for example, "%eth0" on the end) correctly.
3440 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3441 negated items (that is, ~something) in unsigned ints. Some compilers
3442 apparently mutter when there is no cast.
3444 70. If an address verification called from an ACL failed, and did not produce a
3445 user-specific message (i.e. there was only a "system" message), nothing was
3446 put in $acl_verify_message. In this situation, it now puts the system
3449 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3450 session; change 4.31/43 added the unwanted input to the log line - except
3451 that it did not do this in the start of session case. It now does.
3453 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3454 This is wrong and can cause the other end to generate a synchronization
3455 error if it is another Exim or anything else that does the synchronization
3456 check. A QUIT command is no longer sent after a timeout.
3458 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3459 during host lookups.
3461 74. The defer_ok option of callout verification was not working if it was used
3462 when verifying addresses in header lines, that is, for this case:
3464 verify = header_sender/callout=defer_ok
3466 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3467 those file descriptors could be used for SMTP connections. If anything
3468 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3469 could be sent to the SMTP client, causing chaos. The daemon now opens
3470 stdin, stdout, and stderr to /dev/null when it puts itself into the
3473 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3474 log by default. The user can override this with suitable Perl magic.
3476 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3477 the log message when discard triggers, was not working for the DATA ACL or
3478 for the non-SMTP ACL.
3480 78. Error message wording change in sieve.c.
3482 79. If smtp_accept_max_per_host was set, the number of connections could be
3483 restricted to fewer than expected, because the daemon was trying to set up
3484 a new connection before checking whether the processes handling previous
3485 connections had finished. The check for completed processes is now done
3486 earlier. On busy systems, this bug wouldn't be noticed because something
3487 else would have woken the daemon, and it would have reaped the completed
3490 80. If a message was submitted locally by a user whose login name contained one
3491 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3492 It caused a spool format error. I have fixed the spool reading code. A
3493 related problem was that the "from" clause in the Received: line became
3494 illegal because of the space(s). It is now covered by ${quote_local_part.
3496 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3499 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3501 83. Patch from Sieve maintainer to fix unterminated string problem in
3502 "vacation" handling.
3504 84. Some minor changes to the Linux configuration files to help with other
3505 OS variants using glibc.
3507 85. One more patch for Sieve to update vacation handling to latest spec.
3510 ----------------------------------------------------
3511 See the note above about the 4.44 and 4.50 releases.
3512 ----------------------------------------------------
3518 1. Change 4.43/35 introduced a bug that caused file counts to be
3519 incorrectly computed when quota_filecount was set in an appendfile
3522 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3523 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3526 3. Give more explanation in the error message when the command for a transport
3527 filter fails to execute.
3529 4. There are several places where Exim runs a non-Exim command in a
3530 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3531 was being done only for the command run by the queryprogram router. It is
3532 now done for all such subprocesses. The other cases are: ${run, transport
3533 filters, and the commands run by the lmtp and pipe transports.
3535 5. Some older OS have a limit of 256 on the maximum number of file
3536 descriptors. Exim was using setrlimit() to set 1000 as a large value
3537 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3538 systems. I've change it so that if it can't get 1000, it tries for 256.
3540 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3541 was an oversight, and furthermore, ever since the addition of extra
3542 controls (e.g. 4.43/32), the checks on when to allow different forms of
3543 "control" were broken. There should now be diagnostics for all cases when a
3544 control that does not make sense is encountered.
3546 7. $recipients is now available in the predata ACL (oversight).
3548 8. Tidy the search cache before the fork to do a delivery from a message
3549 received from the command line. Otherwise the child will trigger a lookup
3550 failure and thereby defer the delivery if it tries to use (for example) a
3551 cached ldap connection that the parent has called unbind on.
3553 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3554 of $address_data from the recipient verification was clobbered by the
3555 sender verification.
3557 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3558 was its contents. (It was OK if the option was not defined at all.)
3560 11. A "Completed" log line is now written for messages that are removed from
3561 the spool by the -Mrm option.
3563 12. $host_address is now set to the target address during the checking of
3564 ignore_target_hosts.
3566 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3567 being passed; this would have caused $sender_host_name to have been used if
3568 matching the list had actually called for a host name (not very likely,
3569 since this list is usually IP addresses). A host name is now passed as
3572 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3573 code that specifies a non-restarting handler (typically sigaction() in
3574 modern systems) in an attempt to fix a rare and obscure crash bug.
3576 15. Narrowed the window for a race in the daemon that could cause it to ignore
3577 SIGCHLD signals. This is not a major problem, because they are used only to
3578 wake it up if nothing else does.
3580 16. A malformed maildirsize file could cause Exim to calculate negative values
3581 for the mailbox size or file count. Odd effects could occur as a result.
3582 The maildirsize information is now recalculated if the size or filecount
3585 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3586 support for a long time. Removed HAVE_SYS_VFS_H.
3588 18. Updated exipick to current release from John Jetmore.
3590 19. Allow an empty sender to be matched against a lookup in an address list.
3591 Previously the only cases considered were a regular expression, or an
3594 20. Exim went into a mad DNS lookup loop when doing a callout where the
3595 host was specified on the transport, if the DNS lookup yielded more than
3598 21. The RFC2047 encoding function was originally intended for short strings
3599 such as real names; it was not keeping to the 75-character limit for
3600 encoded words that the RFC imposes. It now respects the limit, and
3601 generates multiple encoded words if necessary. To be on the safe side, I
3602 have increased the buffer size for the ${rfc2047: expansion operator from
3605 22. Failure to deliver a bounce message always caused it to be frozen, even if
3606 there was an errors_to setting on the router. The errors_to setting is now
3609 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3610 canonical form (fully expanded) before being placed in
3611 $sender_host_address.
3613 24. Updated eximstats to version 1.33
3615 25. Include certificate and key file names in error message when GnuTLS fails
3616 to set them up, because the GnuTLS error message doesn't include the name
3617 of the failing file when there is a problem reading it.
3619 26. Expand error message when OpenSSL has problems setting up cert/key files.
3622 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3623 (this can affect the format of dates).
3625 28. exim_tidydb, when checking for the continued existence of a message for
3626 which it has found a message-specific retry record, was not finding
3627 messages that were in split spool directories. Consequently, it was
3628 deleting retry records that should have stayed in existence.
3630 29. eximstats updated to version 1.35
3631 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3632 1.35 - bugfix such that pie charts by volume are generated correctly
3634 30. The SPA authentication driver was not abandoning authentication and moving
3635 on to the next authenticator when an expansion was forced to fail,
3636 contradicting the general specification for all authenticators. Instead it
3637 was generating a temporary error. It now behaves as specified.
3639 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3640 (the order specifies the preference for clients). The order is now AES256,
3641 AES128, 3DES, ARCFOUR128.
3643 31. Small patch to Sieve code - explicitly set From: when generating an
3646 32. Exim crashed if a remote delivery caused a very long error message to be
3647 recorded - for instance if somebody sent an entire SpamAssassin report back
3648 as a large number of 550 error lines. This bug was coincidentally fixed by
3649 increasing the size of one of Exim's internal buffers (big_buffer) that
3650 happened as part of the Exiscan merge. However, to be on the safe side, I
3651 have made the code more robust (and fixed the comments that describe what
3654 33. Some experimental protocols are using DNS PTR records for new purposes. The
3655 keys for these records are domain names, not reversed IP addresses. The
3656 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3657 leaves it alone. Component reversal etc. now happens only for IP addresses.
3660 34. The host_aton() function is supposed to be passed a string that is known
3661 to be a valid IP address. However, in the case of IPv6 addresses, it was
3662 not checking this. This is a hostage to fortune. Exim now panics and dies
3663 if the condition is not met. A case was found where this could be provoked
3664 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3665 components; fortuitously, this particular loophole had already been fixed
3666 by change 4.50/55 or 4.44/33 above.
3668 If there are any other similar loopholes, the new check in host_aton()
3669 itself should stop them being exploited. The report I received stated that
3670 data on the command line could provoke the exploit when Exim was running as
3671 exim, but did not say which command line option was involved. All I could
3672 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3673 running as the user.
3676 35. There was a buffer overflow vulnerability in the SPA authentication code
3677 (which came originally from the Samba project). I have added a test to the
3678 spa_base64_to_bits() function which I hope fixes it.
3681 36. The daemon start-up calls getloadavg() while still root for those OS that
3682 need the first call to be done as root, but it missed one case: when
3683 deliver_queue_load_max is set with deliver_drop_privilege. This is
3684 necessary for the benefit of the queue runner, because there is no re-exec
3685 when deliver_drop_privilege is set.
3687 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3688 was in use, was not putting the data itself into the right store pool;
3689 consequently, it could be overwritten for a subsequent message in the same
3690 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3693 38. Sometimes the final signoff response after QUIT could fail to get
3694 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3695 < 0 before doing a fflush(). This bug looks as though it goes back to the
3696 introduction of TLS in release 3.20, but "sometimes" must have been rare
3697 because the tests only now provoked it.
3703 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3704 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3705 Now it returns an integer. A number of calls were still expecting a T/F
3706 return. Fortuitously, in all cases, the tests worked in OK situations,
3707 which is the norm. However, things would have gone wrong on any write
3708 failures on the smtp file descriptor. This function is used when sending
3709 messages over SMTP and also when doing verify callouts.
3711 2. When Exim is called to do synchronous delivery of a locally submitted
3712 message (the -odf or -odi options), it no longer closes stderr before doing
3715 3. Implemented the mua_wrapper option.
3717 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3719 5. Implemented the functions header_remove(), header_testname(),
3720 header_add_at_position(), and receive_remove_recipient(), and exported them
3723 6. If an ACL "warn" statement specified the addition of headers, Exim already
3724 inserted X-ACL-Warn: at the start if there was no header name. However, it
3725 was not making this test for the second and subsequent header lines if
3726 there were newlines in the string. This meant that an invalid header could
3727 be inserted if Exim was badly configured.
3729 7. Allow an ACL "warn" statement to add header lines at the start or after all
3730 the Received: headers, as well as at the end.
3732 8. Added the rcpt_4xx retry error code.
3734 9. Added postmaster_mailfrom=xxx to callout verification option.
3736 10. Added mailfrom=xxxx to the callout verification option, for verify=
3739 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3740 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3742 12. Inserted some casts to stop certain compilers warning when using pointer
3743 differences as field lengths or precisions in printf-type calls (mostly
3744 affecting debugging statements).
3746 13. Added optional readline() support for -be (dynamically loaded).
3748 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3749 same clock tick as a message's arrival, so that its received time was the
3750 same as the "first fail" time on the retry record, and that message
3751 remained on the queue past the ultimate address timeout, every queue runner
3752 would try a delivery (because it was past the ultimate address timeout) but
3753 after another failure, the ultimate address timeout, which should have then
3754 bounced the address, did not kick in. This was a "< instead of <=" error;
3755 in most cases the first failure would have been in the next clock tick
3756 after the received time, and all would be well.
3758 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3759 being recognized when the domain list was tested by the match_domain
3760 condition in an expansion string.
3762 16. Added the ${str2b64: operator.
3764 17. Exim was always calling setrlimit() to set a large limit for the number of
3765 processes, without checking whether the existing limit was already
3766 adequate. (It did check for the limit on file descriptors.) Furthermore,
3767 errors from getrlimit() and setrlimit() were being ignored. Now they are
3768 logged to the main and panic logs, but Exim does carry on, to try to do its
3769 job under whatever limits there are.
3771 18. Imported PCRE 5.0.
3773 19. Trivial typo in log message " temporarily refused connection" (the leading
3776 20. If the log selector return_path_on_delivery was set and an address was
3777 redirected to /dev/null, the delivery process crashed because it assumed
3778 that a return path would always be set for a "successful" delivery. In this
3779 case, the whole delivery is bypassed as an optimization, and therefore no
3782 21. Internal re-arrangement: the function for sending a challenge and reading
3783 a response while authentication was assuming a zero-terminated challenge
3784 string. It's now changed to take a pointer and a length, to allow for
3785 binary data in such strings.
3787 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3789 23. Exim was not respecting finduser_retries when seeking the login of the
3790 uid under which it was called; it was always trying 10 times. (The default
3791 setting of finduser_retries is zero.) Also, it was sleeping after the final
3792 failure, which is pointless.
3794 24. Implemented tls_on_connect_ports.
3796 25. Implemented acl_smtp_predata.
3798 26. If the domain in control=submission is set empty, Exim assumes that the
3799 authenticated id is a complete email address when it generates From: or
3800 Sender: header lines.
3802 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3803 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3804 chown and chgrp in /bin and hostname in /usr/bin.
3806 28. Exim was keeping the "process log" file open after each use, just as it
3807 does for the main log. This opens the possibility of it remaining open for
3808 long periods when the USR1 signal hits a daemon. Occasional processlog
3809 errors were reported, that could have been caused by this. Anyway, it seems
3810 much more sensible not to leave this file open at all, so that is what now
3813 29. The long-running daemon process does not normally write to the log once it
3814 has entered its main loop, and it closes the log before doing so. This is
3815 so that log files can straightforwardly be renamed and moved. However,
3816 there are a couple of unusual error situations where the daemon does write
3817 log entries, and I had neglected to close the log afterwards.
3819 30. The text of an SMTP error response that was received during a remote
3820 delivery was being truncated at 512 bytes. This is too short for some of
3821 the long messages that one sometimes sees. I've increased the limit to
3824 31. It is now possible to make retry rules that apply only when a message has a
3825 specific sender, in particular, an empty sender.
3827 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3828 it possible to be selective about when SMTP synchronization is enforced.
3830 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3832 32. Implemented hosts_connection_nolog.
3834 33. Added an ACL for QUIT.
3836 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3839 35. Added mailbox_size and mailbox_filecount to appendfile.
3841 36. Added control = no_multiline_responses to ACLs.
3843 37. There was a bug in the logic of the code that waits for the clock to tick
3844 in the case where the clock went backwards by a substantial amount such
3845 that the microsecond fraction of "now" was more than the microsecond
3846 fraction of "then" (but the whole seconds number was less).
3848 38. Added support for the libradius Radius client library this is found on
3849 FreeBSD (previously only the radiusclient library was supported).
3855 1. When certain lookups returned multiple values in the form name=value, the
3856 quoting of the values was not always being done properly. Specifically:
3857 (a) If the value started with a double quote, but contained no whitespace,
3859 (b) If the value contained whitespace other than a space character (i.e.
3860 tabs or newlines or carriage returns) it was not quoted.
3861 This fix has been applied to the mysql and pgsql lookups by writing a
3862 separate quoting function and calling it from the lookup code. The fix
3863 should probably also be applied to nisplus, ibase and oracle lookups, but
3864 since I cannot test any of those, I have not disturbed their existing code.
3866 2. A hit in the callout cache for a specific address caused a log line with no
3867 reason for rejecting RCPT. Now it says "Previous (cached) callout
3868 verification failure".
3870 3. There was an off-by-one bug in the queryprogram router. An over-long
3871 return line was truncated at 256 instead of 255 characters, thereby
3872 overflowing its buffer with the terminating zero. As well as fixing this, I
3873 have increased the buffer size to 1024 (and made a note to document this).
3875 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3876 when Exim is waiting for an SMTP response from a remote server, Exim
3877 restarts its select() call on the socket, thereby resetting its timeout.
3878 This is not a problem when such interrupts are rare. Somebody set up a cron
3879 job to run exiwhat every 2 minutes, which is less than the normal select()
3880 timeout (5 or 10 minutes). This meant that the select() timeout never
3881 kicked in because it was always reset. I have fixed this by comparing the
3882 time when an interrupt arrives with the time at the start of the first call
3883 to select(). If more time than the timeout has elapsed, the interrupt is
3884 treated as a timeout.
3886 5. Some internal re-factoring in preparation for the addition of Sieve
3887 extensions (by MH). In particular, the "personal" test is moved to a
3888 separate function, and given an option for scanning Cc: and Bcc: (which is
3889 not set for Exim filters).
3891 6. When Exim created an email address using the login of the caller as the
3892 local part (e.g. when creating a From: or Sender: header line), it was not
3893 quoting the local part when it contained special characters such as @.
3895 7. Installed new OpenBSD configuration files.
3897 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3898 try to make them clearer.
3900 9. Callout options, other than the timeout value, were being ignored when
3901 verifying sender addresses in header lines. For example, when using
3903 verify = header_sender/callout=no_cache
3905 the cache was (incorrectly) being used.
3907 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3908 only the Cygwin environment.
3910 11. When return_path_on_delivery was set as a log selector, if different remote
3911 addresses in the same message used different return paths and parallel
3912 remote delivery occurred, the wrong values would sometimes be logged.
3913 (Whenever a remote delivery process finished, the return path value from
3914 the most recently started remote delivery process was logged.)
3916 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3917 lines when AUTH and/or TLS are in use. This is the "received protocol"
3918 field. Exim used to use "asmtp" for authenticated SMTP, without any
3919 indication (in the protocol name) for TLS use. Now it follows the RFC and
3920 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3921 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3922 names appear in log lines as well as in Received: header lines.
3924 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3925 extensions, and comparison tests, and to fix some bugs.
3927 14. Changes to the "personal" filter test:
3929 (1) The test was buggy in that it was just doing the equivalent of
3930 "contains" tests on header lines. For example, if a user's address was
3931 anne@some.where, the "personal" test would incorrectly be true for
3933 To: susanne@some.where
3935 This test is now done by extracting each address from the header in turn,
3936 and checking the entire address. Other tests that are part of "personal"
3937 are now done using regular expressions (for example, to check local parts
3938 of addresses in From: header lines).
3940 (2) The list of non-personal local parts in From: addresses has been
3941 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3942 taken from the Sieve specification recommendations.
3944 (3) If the message contains any header line starting with "List-" it is
3945 treated as non-personal.
3947 (4) The test for "circular" in the Subject: header line has been removed
3948 because it now seems ill-conceived.
3950 15. Minor typos in src/EDITME comments corrected.
3952 16. Installed latest exipick from John Jetmore.
3954 17. If headers_add on a router specified a text string that was too long for
3955 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3956 of string_sprintf() is now avoided.
3958 18. $message_body_size was not set (it was always zero) when running the DATA
3959 ACL and the local_scan() function.
3961 19. For the "mail" command in an Exim filter, no default was being set for
3962 the once_repeat time, causing a random time value to be used if "once" was
3963 specified. (If the value happened to be <= 0, no repeat happened.) The
3964 default is now 0s, meaning "never repeat". The "vacation" command was OK
3965 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3966 (I found it when inspecting the code).
3968 20. There is now an overall timeout for performing a callout verification. It
3969 defaults to 4 times the callout timeout, which applies to individual SMTP
3970 commands during the callout. The overall timeout applies when there is more
3971 than one host that can be tried. The timeout is checked before trying the
3972 next host. This prevents very long delays if there are a large number of
3973 hosts and all are timing out (e.g. when the network connections are timing
3974 out). The value of the overall timeout can be changed by specifying an
3975 additional sub-option for "callout", called "maxwait". For example:
3977 verify = sender/callout=5s,maxwait=20s
3979 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3980 to the end before writing, but this should make it even safer).
3982 22. Exim was forgetting that it had advertised PIPELINING for the second and
3983 subsequent messages on an SMTP connection. It was also not resetting its
3984 memory on STARTTLS and an internal HELO.
3986 23. When Exim logs an SMTP synchronization error within a session, it now
3987 records whether PIPELINING has been advertised or not.
3989 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3990 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3991 rather than long int.
3993 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3995 26. Added the never_mail option to autoreply.
4001 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4002 crash if the getsockname() call failed; this can happen if a connection is
4003 closed very soon after it is established. The problem was simply in the
4004 order in which certain operations were done, causing Exim to try to write
4005 to the SMTP stream before it had set up the file descriptor. The bug has
4006 been fixed by making things happen in the correct order.
4012 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4013 before the connection was closed, thus losing the rejection response.
4015 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4016 some early Solaris releases, but causes trouble in current releases where
4017 socklen_t is defined.
4019 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4022 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4025 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4026 that releasing the top of it at the end releases what was used for sub-
4027 expansions (unless the block got too big). However, discard this block if
4028 the first thing is a variable or header, so that we can use its block when
4029 it is dynamic (useful for very large $message_headers, for example).
4031 6. Lookups now cache *every* query, not just the most recent. A new, separate
4032 store pool is used for this. It can be recovered when all lookup caches are
4033 flushed. Lookups now release memory at the end of their result strings.
4034 This has involved some general refactoring of the lookup sources.
4036 7. Some code has been added to the store_xxx() functions to reduce the amount
4037 of flapping under certain conditions.
4039 8. log_incoming_interface used to affect only the <= reception log lines. Now
4040 it causes the local interface and port to be added to several more SMTP log
4041 lines, for example "SMTP connection from", and rejection lines.
4043 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4045 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4047 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4048 could be overwritten at the end of the current message (or the start of a
4049 new message if it was set in a HELO ACL). The value is now preserved for
4050 the duration of the SMTP connection.
4052 12. If a transport had a headers_rewrite setting, and a matching header line
4053 contained an unqualified address, that address was qualified, even if it
4054 did not match any rewriting rules. The underlying bug was that the values
4055 of the flags that permit the existence of unqualified sender and recipient
4056 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4057 non-local messages, and by -bnq for local messages) were not being
4058 preserved with the message after it was received.
4060 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4061 "next input=" as part of the text comprising the host identity instead of
4062 the correct text. The code was using the same buffer for two different
4063 strings. However, depending on which order the printing function evaluated
4064 its arguments, the bug did not always show up. Under Linux, for example, my
4065 test suite worked just fine.
4067 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4068 doesn't work with some older versions of Perl. It has been changed to "my",
4069 which in any case is probably the better facility to use.
4071 15. A really picky compiler found some instances of statements for creating
4072 error messages that either had too many or two few arguments for the format
4075 16. The size of the buffer for calls to the DNS resolver has been increased
4076 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4077 for addresses that have a lot of PTR records. This alleviates a problem; it
4078 does not fully solve it.
4080 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4081 buffer now truncates the list and logs the incident, which is the same
4082 action as happens when Exim is looking up a host name and its aliases.
4083 Previously in this situation something unpredictable would happen;
4084 sometimes it was "internal error: store_reset failed".
4086 18. If a server dropped the connection unexpectedly when an Exim client was
4087 using GnuTLS and trying to read a response, the client delivery process
4088 crashed while trying to generate an error log message.
4090 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4091 string, for example:
4093 warn message = H1: something\nH2: something
4095 the text was added as a single header line from Exim's point of view
4096 though it ended up OK in the delivered message. However, searching for the
4097 second and subsequent header lines using $h_h2: did not work. This has been
4098 fixed. Similarly, if a system filter added multiple headers in this way,
4099 the routers could not see them.
4101 20. Expanded the error message when iplsearch is called with an invalid key to
4102 suggest using net-iplsearch in a host list.
4104 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4105 ACLs are no longer actually imposed (and a message to that effect is
4108 22. If a "gecos" field in a passwd entry contained escaped characters, in
4109 particular, if it contained a \" sequence, Exim got it wrong when building
4110 a From: or a Sender: header from that name. A second bug also caused
4111 incorrect handling when an unquoted " was present following a character
4112 that needed quoting.
4114 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4115 was not being matched caselessly.
4117 24. Arranged for all hyphens in the exim.8 source to be escaped with
4120 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4121 properly. Recipient callout cache records were still being keyed to include
4122 the sender, even when use_sender was set false. This led to far more
4123 callouts that were necessary. The sender is no longer included in the key
4124 when use_sender is false.
4126 26. Added "control = submission" modifier to ACLs.
4128 27. Added the ${base62d: operator to decode base 62 numbers.
4130 28. dnsdb lookups can now access SRV records.
4132 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4133 the configuration file.
4135 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4136 (-v) mode. This makes the output for a verbose queue run more intelligible.
4138 31. Added a use_postmaster feature to recipient callouts.
4140 32. Added the $body_zerocount variable, containing the number of binary zero
4141 bytes in the message body.
4143 33. The time of last modification of the "new" subdirectory is now used as the
4144 "mailbox time last read" when there is a quota error for a maildir
4147 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4149 35. Added +ignore_unknown as a special item in host lists.
4151 36. Code for decoding IPv6 addresses in host lists is now included, even if
4152 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4153 address was recognized as an IP address, but was then not correctly decoded
4154 into binary, causing unexpected and incorrect effects when compared with
4161 1. Very minor rewording of debugging text in manualroute to say "list of
4162 hosts" instead of "hostlist".
4164 2. If verify=header_syntax was set, and a header line with an unqualified
4165 address (no domain) and a large number of spaces between the end of the
4166 name and the colon was received, the reception process suffered a buffer
4167 overflow, and (when I tested it) crashed. This was caused by some obsolete
4168 code that should have been removed. The fix is to remove it!
4170 3. When running in the test harness, delay a bit after writing a bounce
4171 message to get a bit more predictability in the log output.
4173 4. Added a call to search_tidyup() just before forking a reception process. In
4174 theory, someone could use a lookup in the expansion of smtp_accept_max_
4175 per_host which, without the tidyup, could leave open a database connection.
4177 5. Added the variables $recipient_data and $sender_data which get set from a
4178 lookup success in an ACL "recipients" or "senders" condition, or a router
4179 "senders" option, similar to $domain_data and $local_part_data.
4181 6. Moved the writing of debug_print from before to after the "senders" test
4184 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4185 problems for message scanning, either using a data ACL, or using
4186 local_scan() because the Received: header was not generated till after they
4187 were called (in order to set the time as the time of reception completion).
4188 I have revised the way this works. The header is now generated after the
4189 body is received, but before the ACL or local_scan() are called. After they
4190 are run, the timestamp in the header is updated.
4196 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4197 before starting a queue runner without re-exec. This happened only when
4198 deliver_drop_privilege was set or when the Exim user was set to root. The
4199 effect of the bug was that timeouts during subsequent deliveries caused
4200 crashes instead of being properly handled. The handler is now left at its
4201 default (and expected) setting.
4203 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4204 message, again when deliver_drop_privilege is set or Exim is run as root.
4205 The bug described in (1) was not present in this case, but the tidying up
4206 of the other signals was missing. I have made the two cases consistent.
4208 3. The ignore_target_hosts setting on a manualroute router was being ignored
4209 for hosts that were looked up using the /MX notation.
4211 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4214 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4215 operated on the sender address. After changing the $sender_address to <>
4216 for the sender address verify, Exim was re-instated it as the original
4217 (before rewriting) address, but remembering that it had rewritten it, so it
4218 wasn't rewriting it again. This bug also had the effect of breaking the
4219 sender address verification caching when the sender address was rewritten.
4221 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4222 This has been changed so that if the ip literal address matches
4223 ignore_target_hosts, the router declines.
4225 7. Added expansion conditions match_domain, match_address, and match_local_
4226 part (NOT match_host).
4228 8. The placeholder for the Received: header didn't have a length field set.
4230 9. Added code to Exim itself and to exim_lock to test for a specific race
4231 condition that could lead to file corruption when using MBX delivery. The
4232 issue is with the lockfile that is created in /tmp. If this file is removed
4233 after a process has opened it but before that process has acquired a lock,
4234 there is the potential for a second process to recreate the file and also
4235 acquire a lock. This could lead to two Exim processes writing to the file
4236 at the same time. The added code performs the same test as UW imapd; it
4237 checks after acquiring the lock that its file descriptor still refers to
4238 the same named file.
4240 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4241 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4244 11. Added the smtp_active_hostname option. If used, this will typically be made
4245 to depend on the incoming interface address. Because $interface_address is
4246 not set up until the daemon has forked a reception process, error responses
4247 that can happen earlier (such as "too many connections") no longer contain
4250 12. If an expansion in a condition on a "warn" statement fails because a lookup
4251 defers, the "warn" statement is abandoned, and the next ACL statement is
4252 processed. Previously this caused the whole ACL to be aborted.
4254 13. Added the iplsearch lookup type.
4256 14. Added ident_timeout as a log selector.
4258 15. Added tls_certificate_verified as a log selector.
4260 16. Added a global option tls_require_ciphers (compare the smtp transport
4261 option of the same name). This controls incoming TLS connections.
4263 17. I finally figured out how to make tls_require_ciphers do a similar thing
4264 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4265 before starting the TLS session.
4267 18. Tabs are now shown as \t in -bP output.
4269 19. If the log selector return_path_on_delivery was set, Exim crashed when
4270 bouncing a message because it had too many Received: header lines.
4272 20. If two routers both had headers_remove settings, and the first one included
4273 a superfluous trailing colon, the final name in the first list and the
4274 first name in the second list were incorrectly joined into one item (with a
4275 colon in the middle).
4281 1. Added -C and -D options to the exinext utility, mainly to make it easier
4282 to include in the automated testing, but these could be helpful when
4283 multiple configurations are in use.
4285 2. The exinext utility was not formatting the output nicely when there was
4286 an alternate port involved in the retry record key, nor when there was a
4287 message id as well (for retries that were specific to a specific message
4288 and a specific host). It was also confused by IPv6 addresses, because of
4289 the additional colons they contain. I have fixed the IPv4 problem, and
4290 patched it up to do a reasonable job for IPv6.
4292 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4293 delivery, the log line now contains "pipelined" if PIPELINING was used.
4295 4. An SMTP transport process used to panic and die if the bind() call to set
4296 an explicit outgoing interface failed. This has been changed; it is now
4297 treated in the same way as a connect() failure.
4299 5. A reference to $sender_host_name in the part of a conditional expansion
4300 that was being skipped was still causing a DNS lookup. This no longer
4303 6. The def: expansion condition was not recognizing references to header lines
4304 that used bh_ and bheader_.
4306 7. Added the _cache feature to named lists.
4308 8. The code for checking quota_filecount in the appendfile transport was
4309 allowing one more file than it should have been.
4311 9. For compatibility with Sendmail, the command line option
4319 and sets the incoming protocol and host name (for trusted callers). The
4320 host name and its colon can be omitted when only the protocol is to be set.
4321 Note the Exim already has two private options, -pd and -ps, that refer to
4322 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4323 "s", but I don't think that's a major issue.
4325 10. A number of refactoring changes to the code, none of which should affect
4328 (a) The number of logging options was getting close to filling up the
4329 32-bit word that was used as a bit map. I have split them into two classes:
4330 those that are passed in the argument to log_write(), and those that are
4331 only ever tested independently outside of that function. These are now in
4332 separate 32-bit words, so there is plenty of room for expansion again.
4333 There is no change in the user interface or the logging behaviour.
4335 (b) When building, for example, log lines, the code previously used a
4336 macro that called string_cat() twice, in order to add two strings. This is
4337 not really sufficiently general. Furthermore, there was one instance where
4338 it was actually wrong because one of the argument was used twice, and in
4339 one call a function was used. (As it happened, calling the function twice
4340 did not affect the overall behaviour.) The macro has been replaced by a
4341 function that can join an arbitrary number of extra strings onto a growing
4344 (c) The code for expansion conditions now uses a table and a binary chop
4345 instead of a serial search (which was left over from when there were very
4346 few conditions). Also, it now recognizes conditions like "pam" even when
4347 the relevant support is not compiled in: a suitably worded error message is
4348 given if an attempt is made to use such a condition.
4350 11. Added ${time_interval:xxxxx}.
4352 12. A bug was causing one of the ddress fields not to be passed back correctly
4353 from remote delivery subprocesses. The field in question was not being
4354 subsequently used, so this caused to problems in practice.
4356 13. Added new log selectors queue_time and deliver_time.
4358 14. Might have fixed a bug in maildirsizefile handling that threw up
4359 "unexpected character" debug warnings, and recalculated the data
4360 unnecessarily. In any case, I expanded the warning message to give more
4363 15. Added the message "Restricted characters in address" to the statements in
4364 the default ACL that block characters like @ and % in local parts.
4366 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4367 Three changes have been made:
4369 (a) There was a serious bug; a negative response to MAIL caused the whole
4370 recipient domain to be cached as invalid, thereby blocking all messages
4371 to all local parts at the same domain, from all senders. This bug has
4372 been fixed. The domain is no longer cached after a negative response to
4373 MAIL if the sender used is not empty.
4375 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4378 (c) A new callout option, "use_sender" has been added for people who want
4379 the modified behaviour.
4385 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4388 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4389 indeed breaks things for older releases.
4391 3. Added additional logging to the case where there is a problem reading data
4392 from a filter that is running in a subprocess using a pipe, in order to
4393 try to track down a specific problem.
4395 4. Testing facility fudge: when running in the test harness and attempting
4396 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4397 getting "No route to host". Convert this to a timeout.
4399 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4402 6. Some OS don't have socklen_t but use size_t instead. This affects the
4403 fifth argument of getsockopt() amongst other things. This is now
4404 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4405 can be set for individual OS. I have set it for SunOS5, OSF1, and
4406 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4407 some earlier ones do not.
4409 7. Change 4.30/15 was not doing the test caselessly.
4411 8. The standard form for an IPv6 address literal was being rejected by address
4412 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4413 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4414 this, as well as the form without the "IPv6" on the front (but only when
4415 address literals are enabled, of course).
4417 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4419 10. Exim crashed if a message with an empty sender address specified by -f
4420 encountered a router with an errors_to setting. This could be provoked only
4421 by a command such as
4425 where an empty string was supplied; "<>" did not hit this bug.
4427 11. Installed PCRE release 4.5.
4429 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4430 remained set. It is now erased.
4432 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4433 times from message ids (which are base 36 rather than the normal 62).
4435 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4436 were being counted as actual protocol errors, and logged if the log
4437 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4438 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4439 and DATA following a set of rejected RCPTs do not count as protocol errors.
4440 In other words, Exim assumes they were pipelined, though this may not
4441 actually be the case. Of course, in all cases the client gets an
4442 appropriate error code.
4444 15. If a lookup fails in an ACL condition, a message about the failure may
4445 be available; it is used if testing the ACL cannot continue, because most
4446 such messages specify what the cause of the deferral is. However, some
4447 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4448 that caused an old message to be retained and used if a later statement
4449 caused a defer, replacing the real cause of the deferral.
4451 16. If an IP address had so many PTR records that the DNS lookup buffer
4452 was not large enough to hold them, Exim could crash while trying to process
4453 the truncated data. It now detects and logs this case.
4455 17. Further to 4.21/58, another change has been made: if (and only if) the
4456 first line of a message (the first header line) ends with CRLF, a bare LF
4457 in a subsequent header line has a space inserted after it, so as not to
4458 terminate the header.
4460 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4461 unnecessarily, used atoi() instead of strtol(), and didn't check the
4462 termination when getting file sizes from file names by regex.
4464 19. Completely re-implemented the support for maildirsize files, in the light
4465 of a number of problems with the previous contributed implementation
4466 (4.30/29). In particular:
4468 . If the quota is zero, the maildirsize file is maintained, but no quota is
4471 . If the maildir directory does not exist, it is created before any attempt
4472 to write a maildirsize file.
4474 . The quota value in the file is just a cache; if the quota is changed in
4475 the transport, the new value overrides.
4477 . A regular expression is available for excluding directories from the
4480 20. The autoreply transport checks the characters in options that define the
4481 message's headers; it allows continued headers, but it was checking with
4482 isspace() after an embedded newline instead of explicitly looking for a
4485 21. If all the "regular" hosts to which an address was routed had passed their
4486 expiry times, and had not reached their retry times, the address was
4487 bounced, even if fallback hosts were defined. Now Exim should go on to try
4490 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4491 equivalent code in the SMTP transport. Some hosts send humungous responses
4492 to HELO/EHLO, more than 1024 it seems.
4494 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4495 gives compiler warnings in some environments. I've now done it "properly",
4498 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4499 (because of problems with the built-in one) was declared to return uschar *
4500 instead of char *, causing compiler failure.
4502 25. Fixed a file descriptor leak when processing alias/forward files.
4504 26. Fixed a minor format string issue in dbfn.c.
4506 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4508 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4509 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4512 29. When a sender address is verified, it is cached, to save repeating the test
4513 when there is more than one recipient in a message. However, when the
4514 verification involves a callout, it is possible for different callout
4515 options to be set for different recipients. It is too complicated to keep
4516 track of this in the cache, so now Exim always runs a verification when a
4517 callout is required, relying on the callout cache for the optimization.
4518 The overhead is duplication of the address routing, but this should not be
4521 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4522 to be verified with callout=postmaster, and the main callout worked but the
4523 postmaster check failed, the verification correctly failed. However, if a
4524 subsequent RCPT command asked for sender verification *without* the
4525 postmaster check, incorrect caching caused this verification also to fail,
4528 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4529 it was not caching the DNS options (qualify_single, search_parents) that
4530 were used when the lookup failed. A subsequent lookup with different
4531 options therefore always gave the same answer, though there were cases
4532 where it should not have. (Example: a "domains = !$mx_any" option on a
4533 dnslookup router: the "domains" option is always processed without any
4534 widening, but the router might have qualify_single set.) Now Exim uses the
4535 cached value only when the same options are set.
4537 32. Added John Jetmore's "exipick" utility to the distribution.
4539 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4540 than a timeout (e.g. a certificate is required, and is not provided), an
4541 Exim server now closes the connection immediately. Previously it waited for
4542 the client to close - but if the client is SSL, it seems that they each
4543 wait for each other, leading to a delay before one of them times out.
4545 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4546 maintained 0.8.x compatibility because I don't think many are using it, and
4547 it is clearly obsolete.
4549 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4552 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4553 client certificate was expired. A simple patch fixes this, though I don't
4554 understand the full logic of why the verify callback is called multiple
4557 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4558 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4559 which causes problems with some clients (such as the Certicom SSL Plus
4560 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4561 disables the coutermeasure allowing Eudora to connect."
4563 38. Exim was not checking that a write() to a log file succeeded. This could
4564 lead to Bad Things if a log got too big, in particular if it hit a file
4565 size limit. Exim now panics and dies if it cannot write to a log file, just
4566 as it does if it cannot open a log file.
4568 39. Modified OS/Makefile-Linux so that it now contains
4570 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4572 The two -D definitions ensure that Exim is compiled with large file
4573 support, which makes it possible to handle log files that are bigger than
4576 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4577 instance) a domain was checked against a named list that involved a lookup,
4578 causing $domain_data to be set, then another domain was checked against the
4579 same list, then the first domain was re-checked, the value of $domain_data
4580 after the final check could be wrong. In particular, if the second check
4581 failed, it could be set empty. This bug probably also applied to
4584 41. The strip_trailing_dot option was not being applied to the address given
4585 with the -f command-line option.
4587 42. The code for reading a message's header from the spool was incrementing
4588 $received_count, but never initializing it. This meant that the value was
4589 incorrect (doubled) while delivering a message in the same process in which
4590 it was received. In the most common configuration of Exim, this never
4591 happens - a fresh exec is done - but it can happen when
4592 deliver_drop_privilege is set.
4594 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4595 it now includes up to 150 characters of the unexpected data in the log
4598 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4599 and building data strings. The size of both of these buffers was 10 000
4600 bytes - far larger than anybody would *ever* want, thought I. Needless to
4601 say, somebody hit the limit. I have increased the maximum line length to
4602 20 000 and the maximum data length of concatenated lines to 100 000. I have
4603 also fixed two bugs, because there was no checking on these buffers. Tsk,
4604 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4605 buffer is too small.
4607 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4608 lsearch lookups. Now it does.
4610 46. When parsing a route_list item in a manualroute router, a fixed-length
4611 buffer was used for the list of hosts. I made this 1024 bytes long,
4612 thinking that nobody would ever have a list of hosts that long. Wrong.
4613 Somebody had a whole pile of complicated expansion conditions, and the
4614 string was silently truncated, leading to an expansion error. It turns out
4615 that it is easier to change to an unlimited length (owing to other changes
4616 that have happened since this code was originally written) than to build
4617 structure for giving a limitation error. The length of the item that
4618 expands into the list of hosts is now unlimited.
4620 47. The lsearch lookup could not handle data where the length of text line was
4621 more than 4095 characters. Such lines were truncated, leading to shortened
4622 data being returned. It should now handle lines of any length.
4624 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4625 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4628 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4629 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4630 no longer happens in comment lines. A list of replacements is now placed
4631 at the head of all of the source files, except those whose only change is
4632 to replace PERL_COMMAND in the very first #! line.
4634 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4635 messages on the queue, with a bottom-up merge sort, using code contributed
4636 by Michael Haardt. This should make operations like -bp somewhat faster on
4637 large queues. It won't affect queue runners, except when queue_run_in_order
4640 51. Installed eximstats 1.31 in the distribution.
4642 52. Added support for SRV lookups to the dnslookup router.
4644 53. If an ACL referred to $message_body or $message_body_end, the value was not
4645 reset for any messages that followed in the same SMTP session.
4647 54. The store-handling optimization for building very long strings was not
4648 differentiating between the different store pools. I don't think this
4649 actually made any difference in practice, but I've tidied it.
4651 55. While running the routers to verify a sender address, $sender_address
4652 was still set to the sender address. This is wrong, because when routing to
4653 send a bounce to the sender, it would be empty. Therefore, I have changed
4654 it so that, while verifying a sender address, $sender_address is set to <>.
4655 (There is no change to what happens when verifying a recipient address.)
4657 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4658 target A or AAAA records (if not already returned) without resetting the
4659 qualify_single or search_parents options of the DNS resolver. These are
4660 inappropriate in this case because the targets of MX and SRV records must
4661 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4662 target that, when qualified, matched something in the local domain. These
4663 two options are now turned off when doing these lookups.
4665 57. It seems that at least some releases of Reiserfs (which does not have the
4666 concept of a fixed number of inodes) returns zero and not -1 for the
4667 number of available inodes. This interacted badly with check_spool_inodes,
4668 which assumed that -1 was the "no such thing" setting. What I have done is
4669 to check that the total number of inodes is greater than zero before doing
4670 the test of how many are available.
4672 58. When a "warn" ACL statement has a log_message modifier, the message is
4673 remembered, and not repeated. This is to avoid a lot of repetition when a
4674 message has many recipients that cause the same warning to be written.
4675 Howewer, Exim was preserving the list of already written lines for an
4676 entire SMTP session, which doesn't seem right. The memory is now reset if a
4677 new message is started.
4679 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4680 debugging output unless log_rewrite was also set.
4682 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4683 of (int)(handle) when we know that handle contains (void *)(-1).
4685 61. The Exim daemon panic-logs an error return when it closes the incoming
4686 connection. However "connection reset by peer" seems to be common, and
4687 isn't really an error worthy of noting specially, so that particular error
4690 62. When Exim is trying to find all the local interfaces, it used to panic and
4691 die if the ioctl to get the interface flags failed. However, it seems that
4692 on at least one OS (Solaris 9) it is possible to have an interface that is
4693 included in the list of interfaces, but for which you get a failure error
4694 for this call. This happens when the interface is not "plumbed" into a
4695 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4696 failure of the "get flags" call assumes that the interface is down.
4698 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4699 makes life easier for people who are doing arithmetic on fields extracted
4700 from dates, where you often get leading zeros that should not be
4701 interpreted as octal.
4703 64. Added qualify_domain to the redirect router, to override the global
4706 65. If a pathologically long header line contained very many addresses (the
4707 report of this problem mentioned 10 000) and each of them was rewritten,
4708 Exim could use up a very large amount of memory. (It kept on making new
4709 copies of the header line as it rewrote, and never released the old ones.)
4710 At the expense of a bit more processing, the header rewriting function has
4711 been changed so that it no longer eats memory in this way.
4713 66. The generation of the Received: header has been moved from the time that a
4714 message starts to be received, to the time that it finishes. The timestamp
4715 in the Received: header should now be very close to that of the <= log
4716 line. There are two side-effects of this change:
4718 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4719 logged header lines no longer include the local Received: line, because
4720 it has not yet been created. The same applies to a copy of the message
4721 that is returned to a non-SMTP sender when a message is rejected.
4723 (b) When a filter file is tested using -bf, no additional Received: header
4724 is added to the test message. After some thought, I decided that this
4727 This change does not affect the value of $received_for. It is still set
4728 after address rewriting, but before local_scan() is called.
4730 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4732 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4733 gave an unhelpful panic error message, and a defer error. I have managed to
4734 change this behaviour so that it now rejects any supplied certificate,
4735 which seems right, as the list of acceptable certificates is empty.
4737 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4738 gave an unhelpful defer error. I have not managed to make this reject any
4739 supplied certificates, but the error message it gives is "no certificate
4740 supplied", which is not helpful.
4742 70. exigrep's output now also includes lines that are not associated with any
4743 message, but which match the given pattern. Implemented by a patch from
4744 Martin Sluka, which also tidied up the Perl a bit.
4746 71. Recipient callout verification, like sender verification, was using <> in
4747 the MAIL FROM command. This isn't really the right thing, since the actual
4748 sender may affect whether the remote host accepts the recipient or not. I
4749 have changed it to use the actual sender in the callout; this means that
4750 the cache record is now keyed on a recipient/sender pair, not just the
4751 recipient address. There doesn't seem to be a real danger of callout loops,
4752 since a callout by the remote host to check the sender would use <>.
4753 [SEE ABOVE: changed after hitting problems.]
4755 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4756 temporary errors. However, in the case of such a code being given after
4757 the end of a data transmission (i.e. after ".") Exim was failing to write
4758 a retry record for the message. (Yes, there was some broken host that was
4759 actually sending 8xx at this point.)
4761 73. An unknown lookup type in a host list could cause Exim to panic-die when
4762 the list was checked. (An example that provoked this was putting <; in the
4763 middle of a list instead of at the start.) If this happened during a DATA
4764 ACL check, a -D file could be left lying around. This kind of configuration
4765 error no longer causes Exim to die; instead it causes a defer errror. The
4766 incident is still logged to the main and panic logs.
4768 74. Buglet left over from Exim 3 conversion. The message "too many messages
4769 in one connection" was written to the rejectlog but not the mainlog, except
4770 when address rewriting (yes!) was being logged.
4772 75. Added write_rejectlog option.
4774 76. When a system filter was run not as root (that is, when system_filter_user
4775 was set), the values of the $n variables were not being returned to the
4776 main process; thus, they were not subsequently available in the $sn
4779 77. Added +return_path_on_delivery log selector.
4781 78. A connection timeout was being treated differently from recipients deferred
4782 when testing hosts_max_try with a message that was older than the host's
4783 retry timeout. (The host should not be counted, thus allowing all hosts to
4784 be tried at least once before bouncing.) This may have been the cause of an
4785 occasionally reported bug whereby a message would remain on the queue
4786 longer than the retry timeout, but would be bounced if a delivery was
4787 forced. I say "may" because I never totally pinned down the problem;
4788 setting up timeout/retry tests is difficult. See also the next item.
4790 79. The ultimate address timeout was not being applied to errors that involved
4791 a combination of host plus message (for example, a timeout on a MAIL
4792 command). When an address resolved to a number of possible hosts, and they
4793 were not all tried for each delivery (e.g. because of hosts_max_try), a
4794 message could remain on the queue longer than the retry timeout.
4796 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4799 81. Fixed an obscure SMTP outgoing bug which required at least the following
4800 conditions: (a) there was another message waiting for the same server;
4801 (b) the server returned 5xx to all RCPT commands in the first message so
4802 that the message was not completed; (c) the server dropped the connection
4803 or gave a negative response to the RSET that Exim sends to abort the
4804 transaction. The observed case was a dropped connection after DATA that had
4805 been sent in pipelining mode. That is, the server had advertised PIPELINING
4806 but was not implementing it correctly. The effect of the bug was incorrect
4807 behaviour, such as trying another host, and this could lead to a crash.
4813 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4814 and daemon.c were passed as pointers to ints; they should have been
4815 pointers to socklen_t variables (which are typically unsigned ints).
4817 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4820 3. Fixed a really odd bug that affected only the testing scheme; patching a
4821 certain fixed string in the binary changed the value of another string that
4822 happened to be identical to the end of the original first string.
4824 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4825 name", it returns that address as the IP address. On some operating
4826 systems (e.g. Solaris), it also passes back the IP address string as the
4827 "host name". However, on others (e.g. Linux), it passes back an empty
4828 string. Exim wasn't checking for this, and was changing the host name to an
4829 empty string, assuming it had been canonicized.
4831 5. Although rare, it is permitted to have more than one PTR record for a given
4832 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4833 all the names associated with an address, because they do in Solaris.
4834 However, it seems that they do not in Linux for data that comes from the
4835 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4836 I found this out when I moved to a new Linux workstation and tried to run
4837 the Exim test suite.
4839 To get round this problem I have changed the code so that it now does its
4840 own call to the DNS to look up PTR records when searching for a host name.
4841 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4842 addresses that are only in /etc/hosts are still found.
4844 This behaviour is, however, controlled by an option called host_lookup_
4845 order, which defaults to "bydns:byaddr". If people want to use the other
4846 order, or indeed, just use one or the other means of lookup, they can
4847 specify it in this variable.
4849 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4850 some operating systems, this comes back from gethostbyaddr() as an empty
4851 string, and this is what Exim used to test for. However, it seems that in
4852 other systems, "." is yielded. Exim now tests for this case too.
4854 7. The values of check_spool_space and check_log_space are now held internally
4855 as a number of kilobytes instead of an absolute number of bytes. If a
4856 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4857 kilobyte. This means that much larger values can be stored.
4859 8. Exim monitor: an attempt to get the action menu when not actually pointing
4860 at a message produces an empty menu entitled "No message selected". This
4861 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4862 no entries in it ("Shell widget menu has zero width and/or height"). So I
4863 have added a single, blank menu entry in this case.
4865 9. Added ${quote_local_part.
4867 10. MIME decoding is now applied to the contents of Subject: header lines when
4870 11. Now that a reference to $sender_host_address automatically causes a reverse
4871 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4872 host lookup before query-style lookups in lists that might use this
4873 variable. This has therefore been abolished, and the "net-" prefix is no
4874 longer necessary for query-style lookups.
4876 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4877 have been a typo for LFLAGS, so it has been changed.
4879 13. The install script calls Exim with "-C /dev/null" in order to find the
4880 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4881 to be output. Howeve, since Exim outputs its version number before the
4882 error, it didn't break the script. It just looked ugly. I fixed this by
4883 always allowing "-C /dev/null" if the caller is root.
4885 14. Ignore overlarge ACL variable number when reading spool file - insurance
4886 against a later release with more variables having written the file.
4888 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4889 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4890 this, as well as the form without the "IPv6" on the front.
4892 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4893 OS/Makefile-Darwin file.
4895 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4896 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4898 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4899 variable. However, some OS use ints for inodes. I've added cast to long int
4900 to get rid of the compiler warning.
4902 19. I had forgotten to lock out "/../" in configuration file names when
4903 ALT_CONFIG_PREFIX was set.
4905 20. Routers used for verification do not need to specify transports. However,
4906 if such a router generated a host list, and callout was configured, Exim
4907 crashed, because it could not find a port number from the (non-existent)
4908 transport. It now assumes port 25 in this circumstance.
4910 21. Added the -t option to exigrep.
4912 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4913 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4914 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4915 them is set, LOOKUP_LSEARCH is forced.
4917 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4919 24. Added sender and host information to the "rejected by local_scan()" log
4920 line; previously there was no indication of these.
4922 25. Added .include_if_exists.
4924 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4925 Linux. It turns out that not all file systems support this. Apparently some
4926 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4927 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4928 not supported on the file descriptor, is now ignored when Exim is trying to
4929 sync a directory. This applies only to Linux.
4931 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4933 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4934 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4937 29. Added support for maildirsize files from supplied patch (modified a bit).
4939 30. The use of :fail: followed by an empty string could lead Exim to respond to
4940 sender verification failures with (e.g.):
4942 550 Verification failed for <xxx>
4943 550 Sender verify failed
4945 where the first response line was missing the '-' that indicates it is not
4946 the final line of the response.
4948 31. The loop for finding the name of the user that called Exim had a hardwired
4949 limit of 10; it now uses the value of finduser_retries, which is used for
4950 all other user lookups.
4952 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4955 33. Exim was neglecting to zero errno before one call of strtol() when
4956 expanding a string and expecting an integer value. On some systems this
4957 resulted in spurious "integer overflow" errors. Also, it was casting the
4958 result into an int without checking.
4960 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4961 did not work. The code looks as if it has *never* worked, though it appears
4962 to have been documented since at least releast 1.62. I have made it work.
4964 35. The "timeout_DNS" error in retry rules, also documented since at least
4965 1.62, also never worked. As it isn't clear exactly what this means, and
4966 clearly it isn't a major issue, I have abolished the feature by treating it
4967 as "timeout", and writing a warning to the main and panic logs.
4969 36. The display of retry rules for -brt wasn't always showing the error code
4972 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4973 timeout_connect_A, timeout_connect_MX.
4975 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4976 to the empty sender.
4978 39. The daemon was not analysing the content of -oX till after it had closed
4979 stderr and disconnected from the controlling terminal. This meant that any
4980 syntax errors were only noted on the panic log, and the return code from
4981 the command was 0. By re-arranging the code a little, I've made the
4982 decoding happen first, so such errors now appear on stderr, and the return
4983 code is 1. However, the actual setting up of the sockets still happens in
4984 the disconnected process, so errors there are still only recorded on the
4987 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4988 connections (as happens on some IP stacks) was logged at start up time as
4989 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4990 it from "IPv6 and IPv4", which means that two separate sockets are being
4993 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4994 says whether AF_INET or AF_INET6 was passed as an argument.
4996 42. Exiwhat output was messed up when time zones were included in log
4999 43. Exiwhat now gives more information about the daemon's listening ports,
5000 and whether -tls-on-connect was used.
5002 44. The "port" option of the smtp transport is now expanded.
5004 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5005 silently ignored. Now an error message is written to the main and panic
5008 46. There's a new ACL modifier called "logwrite" which writes to a log file
5009 as soon as it is encountered.
5011 47. Added $local_user_uid and $local_user_gid at routing time.
5013 48. Exim crashed when trying to verify a sender address that was being
5016 49. Exim was recognizing only a space character after ".include". It now also
5017 recognizes a tab character.
5019 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5020 extracting the relevant information from the specification. The man page no
5021 longer contains scrambled data for the -d option, and I've added a section
5022 at the front about calling Exim under different names.
5024 51. Added "extra_headers" argument to the "mail" command in filter files.
5026 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5029 53. Installed eximstats 1.29.
5031 54. Added transport_filter_timeout as a generic transport option.
5033 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5034 Cc: header lines. This was required by RFC 822, but it not required by RFC
5037 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5038 incoming messages that did not have them. Now it does so only if the
5039 message originates locally, that is, if there is no associated remote host
5040 address. When Resent- header lines are present, this applies to the Resent-
5041 lines rather than the non-Resent- lines.
5043 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5044 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5046 58. Messages for configuration errors now include the name of the main
5047 configuration file - useful now that there may be more than one file in a
5048 list (.included file names were always shown).
5050 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5051 for those rare installations that do not start the daemon as root or run it
5052 setuid root. I've cut out the call to initgroups() if the daemon is not
5055 60. The Exim user and group can now be bound into the binary as text strings
5056 that are looked up at the start of Exim's processing.
5058 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5060 62. Added $mailstore_basename variable.
5062 63. Installed patch to sieve.c from Michael Haardt.
5064 64. When Exim failed to open the panic log after failing to open the main log,
5065 the original message it was trying to log was written to stderr and debug
5066 output, but if they were not available (the usual case in production), it
5067 was lost. Now it is written to syslog before the two lines that record the
5068 failures to open the logs.
5070 65. Users' Exim filters run in subprocesses under the user's uid. It is
5071 possible for a "deliver" command or an alias in a "personal" command to
5072 provoke an address rewrite. If logging of address rewriting is configured,
5073 this fails because the process is not running as root or exim. There may be
5074 a better way of dealing with this, but for the moment (because 4.30 needs
5075 to be released), I have disabled address rewrite logging when running a
5076 filter in a non-root, non-exim process.
5082 1. The buildconfig auxiliary program wasn't quoting the value set for
5083 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5084 not defined. This bug was masked in 4.22 by the effect that was fixed in
5087 2. Some messages that were rejected after a message id was allocated were
5088 shown as "incomplete" by exigrep. It no longer does this for messages that
5089 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5091 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5092 have allow_domain_literals set, the ID did not get logged in the <= line.
5093 Domain literals are now always recognized in Message-ID: header lines.
5095 4. The first argument for a ${extract expansion item is the key name or field
5096 number. Leading and trailing spaces in this item were not being ignored,
5097 causing some misleading effects.
5099 5. When deliver_drop_privilege was set, single queue runner processes started
5100 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5101 same command in the process it spins off) were not dropping privilege.
5103 6. When the daemon running as "exim" started a queue runner, it always
5104 re-executed Exim in the spun-off process. This is a waste of effort when
5105 deliver_drop_privilege is set. The new process now just calls the
5106 queue-runner function directly.
5112 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5115 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5116 ignored. Though the use of -oP was forcing the writing of a pid file, it
5117 was always written to the default place.
5119 3. If the message "no IP address found for host xxxx" is generated during
5120 incoming verification, it is now followed by identification of the incoming
5121 connection (so you can more easily find what provoked it).
5123 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5125 5. Added some features to "harden" Exim a bit more against certain attacks:
5127 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5128 be put in Local/Makefile. This is like the never_users runtime option,
5129 but it cannot be overridden. The default setting is "root".
5131 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5132 prefix string with which any file named in a -C command line option
5135 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5136 is retained for -C and -D only if the caller of Exim is root. Without
5137 it, the exim user may also use -C and -D and retain privilege.
5139 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5140 command line option is disabled.
5142 6. Macro names set by the -D option must start with an upper case letter, just
5143 like macro names defined in the configuration file.
5145 7. Added "dereference=" facility to LDAP.
5147 8. Two instances of the typo "uknown" in the source files are fixed.
5149 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5150 the Configure-Makefile script screwed up while processing it.
5152 10. Incorporated PCRE 4.4.
5154 11. The SMTP synchronization check was not operating right at the start of an
5155 SMTP session. For example, it could not catch a HELO sent before the client
5156 waited for the greeting. There is now a check for outstanding input at the
5157 point when the greeting is written. Because of the duplex, asynchronous
5158 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5159 way, but not yet received, when the check is performed.
5161 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5162 on TCP/IP sockets, because this apparently causes some broken clients to
5165 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5166 unchanged) from the Cygwin maintainer.
5168 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5169 was supported instead of testing for mbx. Effectively a typo.
5171 15. The spa authenticator server code was not checking that the input it
5172 received was valid base64.
5174 16. The debug output line for the "set" modifier in ACLs was not showing the
5175 name of the variable that was being set.
5177 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5179 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5180 lookup on its own. Something else was needed to trigger the lookup. For
5181 example, a match in host_lookup or the need for a host name in a host list.
5182 Now, if $sender_host_name is referenced and the host name has not yet been
5183 looked up, a lookup is performed. If the lookup fails, the variable remains
5184 empty, and $host_lookup_failed is set to "1".
5186 19. Added "eqi" as a case-independent comparison operator.
5188 20. The saslauthd authentication condition could segfault if neither service
5189 nor realm was specified.
5191 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5192 error message that was logged was misleading, and incoming SMTP
5193 connections were dropped. The message is now more accurate, and temporary
5194 errors are given to SMTP connections.
5196 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5197 (except RSET) with a 421 error, until QUIT is received. However, it was
5198 failing to send a response to QUIT.
5200 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5201 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5202 ACL is now run after the helo_try_verify_hosts code.
5204 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5205 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5206 case-independent, so other case variants are also recognized). Apparently
5207 some systems use these upper case variants.
5209 25. If more than two messages were waiting for the same host, and a transport
5210 filter was specified for the transport, Exim sent two messages over the
5211 same TCP/IP connection, and then failed with "socket operation on non-
5212 socket" when it tried to send the third.
5214 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5216 27. The extern definition of crypt16() in expand.c was not being excluded when
5217 the OS had its own crypt16() function.
5219 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5220 as a preferred synonym for return_size_limit, both as an option and as an
5223 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5225 30. Changed the default configuration ACL to relax the local part checking rule
5226 for addresses that are not in any local domains. For these addresses,
5227 slashes and pipe symbols are allowed within local parts, but the sequence
5228 /../ is explicitly forbidden.
5230 31. SPA server authentication was not clearing the challenge buffer before
5233 32. log_message in a "warn" ACL statement was writing to the reject log as
5234 well as to the main log, which contradicts the documentation and doesn't
5235 seem right (because no rejection is happening). So I have stopped it.
5237 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5238 However, I am unable to do any testing of this.
5240 34. Fixed an infelicity in the appendfile transport. When checking directories
5241 for a mailbox, to see if any needed to be created, it was accidentally
5242 using path names with one or more superfluous leading slashes; tracing
5243 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5245 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5246 contents are added to the log line that is written for every discarded
5247 recipient. (Previously a log_message setting was ignored.)
5249 36. The ${quote: operator now quotes the string if it is empty.
5251 37. The install script runs exim in order to find its version number. If for
5252 some reason other than non-existence or emptiness, which it checks, it
5253 could not run './exim', it was installing it with an empty version number,
5254 i.e. as "exim-". This error state is now caught, and the installation is
5257 38. An argument was missing from the function that creates an error message
5258 when Exim fails to connect to the socket for saslauthd authentication.
5259 This could cause Exim to crash, or give a corrupted message.
5261 39. Added isip, isip4, and isip6 to ${if conditions.
5263 40. The ACL variables $acl_xx are now saved with the message, and can be
5264 accessed later in routers, transports, and filters.
5266 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5267 strings in the file are not string-expanded.
5269 42. If a MAIL command specified a SIZE value that was too large to fit into an
5270 int variable, the check against message_size_limit failed. Such values are
5271 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5272 day this will have to be increased, but I don't think I want to be around
5273 when emails are that large.
5280 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5281 iconv() is not standard in FreeBSD.
5283 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5284 IPv6 enabled. The observed symptom was a segmentation fault on return from
5285 the function os_common_find_running_interfaces() in src/os.c.
5287 3. In the check_special_case() function in daemon.c I had used "errno" as an
5288 argument name, which causes warnings on some systems. This was basically a
5289 typo, since it was named "eno" in the comments!
5291 4. The code that waits for the clock to tick (at a resolution of some fraction
5292 of a second) so as to ensure message-id uniqueness was always waiting for
5293 at least one whole tick, when it could have waited for less. [This is
5294 almost certainly not relevant at current processor speeds, where it is
5295 unlikely to ever wait at all. But we try to future-proof.]
5297 5. The function that sleeps for a time interval that includes fractions of a
5298 second contained a race. It did not block SIGALRM between setting the
5299 timer, and suspending (a couple of lines later). If the interval was short
5300 and the sigsuspend() was delayed until after it had expired, the suspension
5301 never ended. On busy systems this could lead to processes getting stuck for
5304 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5305 process, before it forks any delivery processes. The open lookup caching
5306 mechanism meant that the open file or database connection was passed into
5307 the delivery process. The problem was that delivery processes always tidy
5308 up cached lookup data. This could cause a problem for the next delivery
5309 process started by the queue runner, because the external queue runner
5310 process does not know about the closure. So the next delivery process
5311 still has data in the lookup cache. In the case of a file lookup, there was
5312 no problem because closing a file descriptor in a subprocess doesn't affect
5313 the parent. However, if the lookup was caching a connection to a database,
5314 the connection was closed, and the second delivery process was likely to
5315 see errors such as "PGSQL: query failed: server closed the connection
5316 unexpectedly". The problem has been fixed by closing all cached lookups
5317 in a queue runner before running a delivery process.
5319 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5320 seem to have the "const" qualifier which it has on other OS. I've
5323 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5324 *of the same type* (client or server) with the same public name that an
5325 error should be diagnosed.
5327 9. When Exim looked up a host name for an IP address, but failed to find the
5328 original IP address when looking up the host name (a safety check), it
5329 output the message "<ip address> does not match any IP for NULL", which was
5330 confusing, to say the least. The bug was that the host name should have
5331 appeared instead of "NULL".
5333 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5334 user that is built into the binary, and the -C or -D options is used, root
5335 privilege is dropped before the configuration file is read. In addition,
5336 logging is switched to stderr instead of the normal log files. If the
5337 configuration then re-defines the Exim user, the unprivileged environment
5338 is probably not what is expected, so Exim logs a panic warning message (but
5341 However, if deliver_drop_privilege is set, the unprivileged state may well
5342 be exactly what is intended, so the warning has been cut out in that case,
5343 and Exim is allowed to try to write to its normal log files.
5349 1. smtp_return_error_details was not giving details for temporary sender
5350 or receiver verification errors.
5352 2. Diagnose a configuration error if two authenticators have the same public
5355 3. Exim used not to create the message log file for a message until the first
5356 delivery attempt. This could be confusing when incoming messages were held
5357 for policy or load reasons. The message log file is now created at the time
5358 the message is received, and an initial "Received" line is written to it.
5360 4. The automatically generated man page for command line options had a minor
5361 bug that caused no ill effects; however, a more serious problem was that
5362 the procedure for building the man page automatically didn't always
5363 operate. Consequently, release 4.20 contains an out-of-date version. This
5364 shouldn't happen again.
5366 5. When building Exim with embedded Perl support, the script that builds the
5367 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5368 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5370 6. The freeze_tell option was not being used for messages that were frozen on
5371 arrival, either by an ACL or by local_scan().
5373 7. Added the smtp_incomplete_transaction log selector.
5375 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5376 was accepting AUTH without a new EHLO.
5378 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5379 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5382 10. Logging of TCP/IP connections (when configured) now happens in the main
5383 daemon process instead of the child process, so that the TCP/IP connection
5384 count is more accurate (but it can never be perfect).
5386 11. The use of "drop" in a nested ACL was not being handled correctly in the
5387 outer ACL. Now, if condition failure induced by the nested "drop" causes
5388 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5389 or "require"), the connection is dropped.
5391 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5392 that yield "discard" can now be used with an "accept" or a "discard" verb,
5393 but an error is generated for any others (because I can't see a useful way
5394 to define what should happen).
5396 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5397 are now processed in the same way as lines in the Exim configuration file.
5398 In particular, continuation lines are supported.
5400 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5402 15. Added -ti meaning -t -i.
5404 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5405 domains, and warn by logging if others are found.
5407 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5408 returned from GIFCONF when Exim is trying to find the list of interfaces on
5409 a host. The code in os.c has been modified to copy each ifreq to an aligned
5410 structure in all cases.
5412 Also, in some cases, the returned ifreq's were being copied to a 'struct
5413 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5414 means the last couple of bytes of an IPv6 address could be chopped if the
5415 ifreq contained only a normal sockaddr (14 bytes storage).
5417 18. Named domain lists were not supported in the hosts_treat_as_local option.
5418 An entry such as +xxxx was not recognized, and was treated as a literal
5421 19. Ensure that header lines added by a DATA ACL are included in the reject log
5422 if the ACL subsequently rejects the message.
5424 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5425 MD5 (which is deprecated).
5427 21. When testing a filter file using -bf, Exim was writing a message when it
5428 took the sender from a "From " line in the message, but it was not doing so
5429 when it took $return_path from a Return-Path: header line. It now does.
5431 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5432 with a valid header line field name (a series of printing characters
5433 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5435 23. Changed "disc" in the source to "disk" to conform to the documentation and
5436 the book and for uniformity.
5438 24. Ignore Sendmail's -Ooption=value command line item.
5440 25. When execve() failed while trying to run a command in a pipe transport,
5441 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5442 could be confused with a return value of 69 from the command itself. This
5443 has been changed to 127, the value the shell returns if it is asked to run
5444 a non-existent command. The wording for the related log line suggests a
5445 non-existent command as the problem.
5447 26. If received_header_text expands to an empty string, do not add a Received:
5448 header line to the message. (Well, it adds a token one on the spool, but
5449 marks it "old" so that it doesn't get used or transmitted.)
5451 27. Installed eximstats 1.28 (addition of -nt option).
5453 28. There was no check for failure on the call to getsockname() in the daemon
5454 code. This can fail if there is a shortage of resources on the system, with
5455 ENOMEM, for example. A temporary error is now given on failure.
5457 29. Contrary to the C standard, it seems that in some environments, the
5458 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5459 program. Exim now does this explicitly; it affects the formatting of
5460 timestamps using strftime().
5462 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5463 complaints. I've now initialized all the variables, to avoid this.
5465 32. Header lines added by a system filter were not being "seen" during
5466 transport-time rewrites.
5468 33. The info_callback() function passed to OpenSSL is set up with type void
5469 (*)(SSL *, int, int), as described somewhere. However, when calling the
5470 function (actually a macro) that sets it up, the type void(*)() is
5471 expected. I've put in a cast to prevent warnings from picky compilers.
5473 34. If a DNS black list lookup found a CNAME record, but there were no A
5474 records associated with the domain it pointed at, Exim crashed.
5476 35. If a DNS black list lookup returned more than one A record, Exim ignored
5477 all but the first. It now scans all returned addresses if a particular IP
5478 value is being sought. In this situation, the contents of the
5479 $dnslist_value variable are a list of all the addresses, separated by a
5482 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5483 to accept a host name and all its aliases if the forward lookup for any of
5484 them yielded the IP address of the incoming connection. Now it accepts only
5485 those names whose forward lookup yields the correct IP address. Any other
5486 names are discarded. This closes a loophole whereby a rogue DNS
5487 administrator could create reverse DNS records to break through a
5488 wildcarded host restriction in an ACL.
5490 37. If a user filter or a system filter that ran in a subprocess used any of
5491 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5492 the wrong values were passed to the pipe command ($thisaddress had the
5493 value of $0, $0 had the value of $1, etc). This bug was introduced by
5494 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5496 38. Improved the line breaking for long SMTP error messages from ACLs.
5497 Previously, if there was no break point between 40 and 75 characters, Exim
5498 left the rest of the message alone. Two changes have been made: (a) I've
5499 reduced the minimum length to 35 characters; (b) if it can't find a break
5500 point between 35 and 75 characters, it looks ahead and uses the first one
5501 that it finds. This may give the occasional overlong line, but at least the
5502 remaining text gets split now.
5504 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5505 file descriptors might be low, and that setting 1000 would always raise it.
5506 It turns out that in some environments, the limit is already over 1000 and
5507 that lowering it causes trouble. So now Exim takes care not to decrease it.
5509 40. When delivering a message, the value of $return_path is set to $sender_
5510 address at the start of routing (routers may change the value). By an
5511 oversight, this default was not being set up when an address was tested by
5512 -bt or -bv, which affected the outcome if any router or filter referred to
5515 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5516 log only when "message" or "log_message" are set. However, if one of the
5517 conditions was an address verification, or a call to a nested ACL, the
5518 messages generated by the underlying test were being passed through. This
5519 no longer happens. The underlying message is available in $acl_verify_
5520 message for both "message" and "log_message" expansions, so it can be
5521 passed through if needed.
5523 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5524 new expansion $bh_ to give the encoded byte string without charset
5525 translation. Translation happens only if iconv() is available; HAVE_ICONV
5526 indicates this at build time. HEADERS_CHARSET gives the charset to
5527 translate to; headers_charset can change it in the configuration, and
5528 "headers charset" can change it in an individual filter file.
5530 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5531 that creates RFC 2047 encoded "words" labels them as that charset instead
5532 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5533 expansion operator; (ii) when Exim creates a From: line for a local
5534 message; (iii) when a header line is rewritten to include a "phrase" part.
5536 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5537 buggy, causing it to skip the first lines of messages whose message ID
5538 ended in 'D'. This would not have bitten before Exim release 4.14, because
5539 message IDs were unlikely to end in 'D' before then. The effect was to have
5540 incorrect size information for certain domains.
5542 45. #include "config.h" was missing at the start of the crypt16.c module. This
5543 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5546 46. If there was a timeout during a "random" callout check, Exim treated it as
5547 a failure of the random address, and carried on sending RSET and the real
5548 address. If the delay was just some slowness somewhere, the response to the
5549 original RCPT would be taken as a response to RSET and so on, causing
5550 mayhem of various kinds.
5552 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5553 when I implemented it. It didn't allow for the fact that some option values
5554 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5555 the right test for positive values.
5557 48. Domain names in DNS records are case-independent. Exim always looks them up
5558 in lower case. Some resolvers return domain names in exactly the case they
5559 appear in the zone file, that is, they may contain uppercase letters. Not
5560 all resolvers do this - some return always lower case. Exim was treating a
5561 change of case by a resolver as a change of domain, similar to a widening
5562 of a domain abbreviation. This triggered its re-routing code and so it was
5563 trying to route what was effectively the same domain again. This normally
5564 caused routing to fail (because the router wouldn't handle the domain
5565 twice). Now Exim checks for this case specially, and just changes the
5566 casing of the domain that it ultimately uses when it transmits the message
5569 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5572 50. If a filter generated a file delivery with a non-absolute name (possible if
5573 no home directory exists for the router), the forbid_file option was not
5576 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5577 the existing equality matching.
5579 52. Exim was using ints instead of ino_t variables in some places where it was
5580 dealing with inode numbers.
5582 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5583 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5584 called TMPDIR, and if it finds it is different, it changes its value.
5586 54. The smtp_printf() function is now made available to local_scan() so
5587 additional output lines can be written before returning. There is also an
5588 smtp_fflush() function to enable the detection of a dropped connection.
5589 The variables smtp_input and smtp_batched_input are exported to
5592 55. Changed the default runtime configuration: the message "Unknown user"
5593 has been removed from the ACL, and instead placed on the localuser router,
5594 using the cannot_route_message feature. This means that any verification
5595 failures that generate their own messages won't get overridden. Similarly,
5596 the "Unrouteable address" message that was in the ACL for unverifiable
5597 relay addresses has also been removed.
5599 56. Added hosts_avoid_esmtp to the smtp transport.
5601 57. The exicyclog script was not checking for the esoteric option
5602 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5603 will work only if exicyclog is run under the appropriate euid.
5605 58. Following a discussion on the list, the rules by which Exim recognises line
5606 endings on incoming messages have been changed. The -dropcr and drop_cr
5607 options are now no-ops, retained only for backwards compatibility. The
5608 following line terminators are recognized: LF CRLF CR. However, special
5609 processing applies to CR:
5611 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5612 nor a local message in the state where . is a terminator.
5614 (ii) If a bare CR is encountered in a header line, an extra space is added
5615 after the line terminator so as not to end the header. The reasoning
5616 behind this is that bare CRs in header lines are most likely either
5617 to be mistakes, or people trying to play silly games.
5619 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5620 was being incorrectly given as 18 bytes larger than it should have been.
5621 This is a VOB (very old bug).
5623 60. This may never have affected anything current, but just in case it has:
5624 When the local host is found other than at the start of a list of hosts,
5625 the local host, those with the same MX, and any that follow, are discarded.
5626 When the list in question was part of a longer list of hosts, the following
5627 hosts (not currently being processed) were also being discarded. This no
5628 longer happens. I'm not sure if this situation could ever has previously
5631 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5634 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5635 header. This is something that is recommended in a new Internet Draft, and
5636 is something that is documented as being done by Sendmail. There are two
5637 possible values. For messages generated by the autoreply transport, Exim
5640 Auto-Submitted: auto-replied
5642 whereas for all other generated messages (e.g. bounces) it adds
5644 Auto-Submitted: auto-generated
5646 63. The "personal" condition in filters now includes a test for the
5647 Auto-Submitted: header. If it contains the string "auto-" the message it
5648 not considered personal.
5650 64. Added rcpt_include_affixes as a generic transport option.
5652 65. Added queue_only_override (default true).
5654 66. Added the syslog_duplication option.
5656 67. If what should have been the first header line of a message consisted of
5657 a space followed by a colon, Exim was mis-interpreting it as a header line.
5658 It isn't of course - it is syntactically invalid and should therefore be
5659 treated as the start of the message body. The misbehaviour could have
5660 caused a number of strange effects, including loss of data in subsequent
5661 header lines, and spool format errors.
5663 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5664 client host had authenticated. This control can now be exercised by an ACL
5665 for more flexibility.
5667 69. By default, callouts do not happen when testing with -bh. There is now a
5668 variant, -bhc, which does actually run the callout code, including
5669 consulting and updating the callout cache.
5671 70. Added support for saslauthd authentication, courtesy of Alexander
5674 71. If statvfs() failed on the spool or log directories while checking their
5675 size for availability, Exim confusingly gave the error "space shortage".
5676 Furthermore, in debugging mode it crashed with a floating point exception.
5677 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5678 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5679 serious problem, Exim now writes to the main and panic logs when this
5680 happens, with details of the failure. It then refuses to accept the
5681 incoming message, giving the message "spool directory problem" or "log
5682 directory problem" with a 421 code for SMTP messages.
5684 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5685 0, 1, and 2 exist, because some OS complain for execs without them (see
5686 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5687 descriptors. However, the code omitted to check that the open succeeded,
5688 causing mysterious errors if for some reason the permissions on /dev/null
5689 got screwed. Now Exim writes a message to the main and panic logs, and
5690 bombs out if it can't open /dev/null.
5692 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5693 interact so that it is all more flexible. It is supposed to remain
5694 backwards compatible. Also added extra_local_interfaces.
5696 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5697 to bomb out with an assertion failure - to the client this appears as a
5698 connection drop. This problem occurs in the part of the code that was taken
5699 from the Samba project. Fortunately, the assertion is in a very simple
5700 function, so I have fixed this by reproducing the function inline in the
5701 one place where it is called, and arranging for authentication to fail
5702 instead of killing the process with assert().
5704 75. The SPA client code was not working when the server requested OEM rather
5705 than Unicode encoding.
5707 76. Added code to make require_files with a specific uid setting more usable in
5708 the case where statting the file as root fails - usually a non-root-mounted
5709 NFS file system. When this happens and the failure is EACCES, Exim now
5710 forks a subprocess and does the per-uid checking as the relevant uid.
5712 77. Added process_log_path.
5714 78. If log_file_path was not explicitly set, a setting of check_log_space or
5715 check_log_inodes was ignored.
5717 79. If a space check for the spool or log partitions fails, the incident is now
5718 logged. Of course, in the latter case the data may get lost...
5720 80. Added the %p formatting code to string_format() so that it can be used to
5721 print addresses in debug_print(). Adjusted all the address printing in the
5722 debugging in store.c to use %p rather than %d.
5724 81. There was a concern that a line of code in smtp_in.c could overflow a
5725 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5726 initially expressed, the concern was not well-founded, because trailing
5727 spaces are removed early. However, if the trailing spaces were followed by
5728 a NULL, they did not get removed, so the overflow was possible. Two fixes
5731 (a) I re-wrote the offending code in a cleaner fashion.
5732 (b) If an incoming SMTP command contains a NULL character, it is rejected
5735 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5736 runs initgroups(), so that if the Exim user is in any additional groups,
5737 they will be used during message reception.
5743 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob