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 TL/06 Add Experimental DMARC support using libopendmarc libraries.
49 TL/07 Fix an out of order global option causing a segfault. Reported to dev
50 mailing list by by Dmitry Isaikin.
52 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
54 JH/02 Support "G" suffix to numbers in ${if comparisons.
56 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
58 NM/01 Bugzilla 1197 - Spec typo
59 Bugzilla 1196 - Spec examples corrections
61 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
63 PP/09 Add gnutls_enable_pkcs11 option.
65 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
66 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
68 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
69 "acl {{name}{arg}...}", and optional args on acl condition
72 JH/05 Permit multiple router/transport headers_add/remove lines.
74 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
76 JH/07 Avoid using a waiting database for a single-message-only transport.
77 Performance patch from Paul Fisher. Bugzilla 1262.
79 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
82 JH/09 Add $headers_added variable, with content from use of ACL modifier
83 add_header (but not yet added to the message). Bugzilla 199.
85 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
86 Pulled from Bugzilla 817 by Wolfgang Breyha.
88 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
91 JH/11 Add A= logging on delivery lines, and a client_set_id option on
94 JH/12 Add optional authenticated_sender logging to A= and a log_selector
97 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
99 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
100 advertise SMTP AUTH mechanism to us, instead of a generic
101 protocol violation error. Also, make Exim more robust to bad
102 data from the Dovecot auth socket.
104 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
106 When a queue runner is handling a message, Exim first routes the
107 recipient addresses, during which it prunes them based on the retry
108 hints database. After that it attempts to deliver the message to
109 any remaining recipients. It then updates the hints database using
112 So if a recipient address works intermittently, it can get repeatedly
113 deferred at routing time. The retry hints record remains fresh so the
114 address never reaches the final cutoff time.
116 This is a fairly common occurrence when a user is bumping up against
117 their storage quota. Exim had some logic in its local delivery code
118 to deal with this. However it did not apply to per-recipient defers
119 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
121 This change adds a proper retry rule check during routing so that the
122 final cutoff time is checked against the message's age. We only do
123 this check if there is an address retry record and there is not a
124 domain retry record; this implies that previous attempts to handle
125 the address had the retry_use_local_parts option turned on. We use
126 this as an approximation for the destination being like a local
127 delivery, as in LMTP.
129 I suspect this new check makes the old local delivery cutoff check
130 redundant, but I have not verified this so I left the code in place.
132 TF/02 Correct gecos expansion when From: is a prefix of the username.
134 Test 0254 submits a message to Exim with the header
138 When I ran the test suite under the user fanf2, Exim expanded
139 the header to contain my full name, whereas it should have added
140 a Resent-Sender: header. It erroneously treats any prefix of the
141 username as equal to the username.
143 This change corrects that bug.
145 GF/01 DCC debug and logging tidyup
146 Error conditions log to paniclog rather than rejectlog.
147 Debug lines prefixed by "DCC: " to remove any ambiguity.
149 TF/03 Avoid unnecessary rebuilds of lookup-related code.
151 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
152 Bug spotted by Jeremy Harris; was flawed since initial commit.
153 Would have resulted in OCSP responses post-SNI triggering an Exim
154 NULL dereference and crash.
156 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
158 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
159 Bug detection, analysis and fix by Samuel Thibault.
160 Bugzilla 1331, Debian bug #698092.
162 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
164 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
165 Server implementation by Todd Lyons, client by JH.
166 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
167 config variable "prdr_enable" controls whether the server
168 advertises the facility. If the client requests PRDR a new
169 acl_data_smtp_prdr ACL is called once for each recipient, after
170 the body content is received and before the acl_smtp_data ACL.
171 The client is controlled by bolth of: a hosts_try_prdr option
172 on the smtp transport, and the server advertisement.
173 Default client logging of deliveries and rejections involving
174 PRDR are flagged with the string "PRDR".
176 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
177 fclose(). Diagnosis by Todd Lyons.
179 PP/17 Update configure.default to handle IPv6 localhost better.
180 Patch by Alain Williams (plus minor tweaks).
183 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
184 This is now consistent with GnuTLS, and is now documented: the
185 previous undocumented portable approach to treating the option as
186 unset was to force an expansion failure. That still works, and
187 an empty string is now equivalent.
189 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
190 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
191 not performing validation itself.
193 PP/20 Added force_command boolean option to pipe transport.
194 Patch from Nick Koston, of cPanel Inc.
196 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
199 PP/21 Fix eximon continuous updating with timestamped log-files.
200 Broken in a format-string cleanup in 4.80, missed when I repaired the
201 other false fix of the same issue.
202 Report and fix from Heiko Schlichting.
205 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
206 Report from Prashanth Katuri.
208 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
209 It's SecureTransport, so affects any MacOS clients which use the
210 system-integrated TLS libraries, including email clients.
216 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
218 This, or similar/improved, will also be change PP/11 of 4.82.
224 PP/01 Handle short writes when writing local log-files.
225 In practice, only affects FreeBSD (8 onwards).
226 Bugzilla 1053, with thanks to Dmitry Isaikin.
228 NM/01 Bugzilla 949 - Documentation tweak
230 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
233 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
235 PP/02 Implemented gsasl authenticator.
237 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
239 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
240 `pkg-config foo` for cflags/libs.
242 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
243 with rest of GSASL and with heimdal_gssapi.
245 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
246 `pkg-config foo` for cflags/libs for the TLS implementation.
248 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
249 properties get this fed in as external SSF. A number of robustness
250 and debugging improvements to the cyrus_sasl authenticator.
252 PP/08 cyrus_sasl server now expands the server_realm option.
254 PP/09 Bugzilla 1214 - Log authentication information in reject log.
255 Patch by Jeremy Harris.
257 PP/10 Added dbmjz lookup type.
259 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
261 PP/12 MAIL args handles TAB as well as SP, for better interop with
262 non-compliant senders.
263 Analysis and variant patch by Todd Lyons.
265 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
266 Bug report from Lars Müller <lars@samba.org> (via SUSE),
267 Patch from Dirk Mueller <dmueller@suse.com>
269 PP/13 tls_peerdn now print-escaped for spool files.
270 Observed some $tls_peerdn in wild which contained \n, which resulted
271 in spool file corruption.
273 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
274 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
275 or write after TLS renegotiation, which otherwise led to messages
278 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
279 as a tracking header (ie: a signed header comes before the signature).
280 Patch from Wolfgang Breyha.
282 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
283 comma-sep list; embedded commas doubled.
285 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
287 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
289 Report and patch from Dmitry Banschikov.
291 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
292 Removed SSL_clear() after SSL_new() which led to protocol negotiation
293 failures. We appear to now support TLS1.1+ with Exim.
295 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
296 lets Exim select keys and certificates based upon TLS SNI from client.
297 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
298 before an outbound SMTP session. New log_selector, +tls_sni.
300 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
301 NULL dereference. Report and patch from Alun Jones.
303 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
304 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
305 Not seeing resolver debug output on NetBSD, but suspect this is a
306 resolver implementation change.
308 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
309 Left warnings. Added "eximon gdb" invocation mode.
311 PP/21 Defaulting "accept_8bitmime" to true, not false.
313 PP/22 Added -bw for inetd wait mode support.
315 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
316 locate the relevant includes and libraries. Made this the default.
318 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
319 Bugzilla 1246, report and most of solution from Tomasz Kusy.
321 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
322 This may cause build issues on older platforms.
324 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
325 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
326 gnutls_require_mac & gnutls_require_protocols (no longer supported).
327 Added SNI support via GnuTLS too.
328 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
330 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
332 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
333 Applied second patch from Janne, implementing suggestion to default
334 multiple-strings-in-record handling to match SPF spec.
336 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
338 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
339 read-only, out of scope).
340 Patch from Wolfgang Breyha, report from Stuart Northfield.
342 PP/29 Fix three issues highlighted by clang analyser static analysis.
343 Only crash-plausible issue would require the Cambridge-specific
344 iplookup router and a misconfiguration.
345 Report from Marcin Mirosław.
347 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
349 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
350 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
351 As part of this, removing so much warning spew let me fix some minor
352 real issues in debug logging.
354 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
355 assignment on my part. Fixed.
357 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
358 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
359 Janne Snabb (who went above and beyond: thank you).
361 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
362 string otherwise requires a connection and a bunch more work and it's
363 relatively easy to get wrong. Should also expose TLS library linkage
366 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
367 64-bit ${eval} (JH/03).
369 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
370 GNU libc to support some of the 64-bit stuff, should not lead to
371 conflicts. Defined before os.h is pulled in, so if a given platform
372 needs to override this, it can.
374 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
375 protection layer was required, which is not implemented.
376 Bugzilla 1254, patch from Wolfgang Breyha.
378 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
379 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
380 tls_dhparam take prime identifiers. Also unbreak combination of
381 OpenSSL+DH_params+TLSSNI.
383 PP/39 Disable SSLv2 by default in OpenSSL support.
389 PP/01 Solaris build fix for Oracle's LDAP libraries.
390 Bugzilla 1109, patch from Stephen Usher.
392 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
394 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
397 TF/02 Fix a couple more cases where we did not log the error message
398 when unlink() failed. See also change 4.74-TF/03.
400 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
401 lock up or crash if it happened to be inside a call to libc when it
402 got a SIGUSR1 from exiwhat.
404 The SIGUSR1 handler appends the current process status to the process
405 log which is later printed by exiwhat. It used to use the general
406 purpose logging code to do this, but several functions it calls are
407 not safe for signals.
409 The new output code in the SIGUSR1 handler is specific to the process
410 log, and simple enough that it's easy to inspect for signal safety.
411 Removing some special cases also simplifies the general logging code.
412 Removing the spurious timestamps from the process log simplifies
415 TF/04 Improved ratelimit ACL condition.
417 The /noupdate option has been deprecated in favour of /readonly which
418 has clearer semantics. The /leaky, /strict, and /readonly update modes
419 are mutually exclusive. The update mode is no longer included in the
420 database key; it just determines when the database is updated. (This
421 means that when you upgrde Exim will forget old rate measurements.)
423 Exim now checks that the per_* options are used with an update mode that
424 makes sense for the current ACL. For example, when Exim is processing a
425 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
426 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
427 must specify per_mail/readonly. If you omit the update mode it defaults to
428 /leaky where that makes sense (as before) or /readonly where required.
430 The /noupdate option is now undocumented but still supported for
431 backwards compatibility. It is equivalent to /readonly except that in
432 ACLs where /readonly is required you may specify /leaky/noupdate or
433 /strict/noupdate which are treated the same as /readonly.
435 A useful new feature is the /count= option. This is a generalization
436 of the per_byte option, so that you can measure the throughput of other
437 aggregate values. For example, the per_byte option is now equivalent
438 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
440 The per_rcpt option has been generalized using the /count= mechanism
441 (though it's more complicated than the per_byte equivalence). When it is
442 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
443 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
444 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
445 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
446 Note that using per_rcpt with a non-readonly update mode in more than
447 one ACL will cause the recipients to be double-counted. (The per_mail
448 and per_byte options don't have this problem.)
450 The handling of very low rates has changed slightly. If the computed rate
451 is less than the event's count (usually one) then this event is the first
452 after a long gap. In this case the rate is set to the same as this event's
453 count, so that the first message of a spam run is counted properly.
455 The major new feature is a mechanism for counting the rate of unique
456 events. The new per_addr option counts the number of different
457 recipients that someone has sent messages to in the last time period. It
458 behaves like per_rcpt if all the recipient addresses are different, but
459 duplicate recipient addresses do not increase the measured rate. Like
460 the /count= option this is a general mechanism, so the per_addr option
461 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
462 example, measure the rate that a client uses different sender addresses
463 with the options per_mail/unique=$sender_address. There are further
464 details in the main documentation.
466 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
468 TF/06 Removed a few PCRE remnants.
470 TF/07 Automatically extract Exim's version number from tags in the git
471 repository when doing development or release builds.
473 PP/02 Raise smtp_cmd_buffer_size to 16kB.
474 Bugzilla 879. Patch from Paul Fisher.
476 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
477 Heavily based on revision 40f9a89a from Simon Arlott's tree.
480 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
482 PP/05 Variable $av_failed, true if the AV scanner deferred.
483 Bugzilla 1078. Patch from John Horne.
485 PP/06 Stop make process more reliably on build failure.
486 Bugzilla 1087. Patch from Heiko Schlittermann.
488 PP/07 Make maildir_use_size_file an _expandable_ boolean.
489 Bugzilla 1089. Patch from Heiko Schlittermann.
491 PP/08 Handle ${run} returning more data than OS pipe buffer size.
492 Bugzilla 1131. Patch from Holger Weiß.
494 PP/09 Handle IPv6 addresses with SPF.
495 Bugzilla 860. Patch from Wolfgang Breyha.
497 PP/10 GnuTLS: support TLS 1.2 & 1.1.
499 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
502 PP/11 match_* no longer expand right-hand-side by default.
503 New compile-time build option, EXPAND_LISTMATCH_RHS.
504 New expansion conditions, "inlist", "inlisti".
506 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
508 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
510 PP/14 fix log_write() format string regression from TF/03.
511 Bugzilla 1152. Patch from Dmitry Isaikin.
517 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
519 PP/02 Harmonised TLS library version reporting; only show if debugging.
520 Layout now matches that introduced for other libraries in 4.74 PP/03.
522 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
524 PP/04 New "dns_use_edns0" global option.
526 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
529 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
530 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
532 TK/01 Updated PolarSSL code to 0.14.2.
533 Bugzilla 1097. Patch from Andreas Metzler.
535 PP/07 Catch divide-by-zero in ${eval:...}.
538 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
541 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
542 format-string attack -- SECURITY: remote arbitrary code execution.
544 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
545 time unintentionally subject to list matching rules, letting the header
546 cause arbitrary Exim lookups (of items which can occur in lists, *not*
547 arbitrary string expansion). This allowed for information disclosure.
549 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
550 INT_MIN/-1 -- value coerced to INT_MAX.
556 NM/01 Workround for PCRE version dependency in version reporting
559 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
560 This fixes portability to compilers other than gcc, notably
561 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
563 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
564 makefiles for portability to HP-UX and POSIX correctness.
566 PP/01 Permit LOOKUP_foo enabling on the make command-line.
567 Also via indented variable definition in the Makefile.
568 (Debugging by Oliver Heesakkers).
570 PP/02 Restore caching of spamd results with expanded spamd_address.
571 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
573 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
574 Improves build reliability. Fix from: Frank Elsner
576 NM/02 Fix wide character breakage in the rfc2047 coding
577 Fixes bug 1064. Patch from Andrey N. Oktyabrski
579 NM/03 Allow underscore in dnslist lookups
580 Fixes bug 1026. Patch from Graeme Fowler
582 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
583 Code patches from Adam Ciarcinski of NetBSD.
585 NM/04 Fixed exiqgrep to cope with mailq missing size issue
588 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
589 is logged, to avoid truncation. Patch from John Horne.
591 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
592 Patch from Jakob Hirsch.
594 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
595 SQL string expansion failure details.
596 Patch from Andrey Oktyabrski.
598 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
599 Patch from Simon Arlott.
601 PP/09 New lookups functionality failed to compile on old gcc which rejects
602 extern declarations in function scope.
603 Patch from Oliver Fleischmann
605 PP/10 Use sig_atomic_t for flags set from signal handlers.
606 Check getgroups() return and improve debugging.
607 Fixed developed for diagnosis in bug 927 (which turned out to be
610 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
611 Patch from Mark Zealey.
613 PP/12 Bugzilla 1056: Improved spamd server selection.
614 Patch from Mark Zealey.
616 PP/13 Bugzilla 1086: Deal with maildir quota file races.
617 Based on patch from Heiko Schlittermann.
619 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
620 Patch from Uwe Doering, sign-off by Michael Haardt.
622 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
623 variable declaration deep within a block. Bug and patch from
626 PP/15 lookups-Makefile IRIX compatibilty coercion.
628 PP/16 Make DISABLE_DKIM build knob functional.
630 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
631 Patch by Simon Arlott
633 TF/03 Fix valgrind.h portability to C89 compilers that do not support
634 variable argument macros. Our copy now differs from upstream.
640 TF/01 Failure to get a lock on a hints database can have serious
641 consequences so log it to the panic log.
643 TF/02 Log LMTP confirmation messages in the same way as SMTP,
644 controlled using the smtp_confirmation log selector.
646 TF/03 Include the error message when we fail to unlink a spool file.
648 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
649 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
650 for maintaining out-of-tree patches for some time.
652 PP/01 Bugzilla 139: Documentation and portability issues.
653 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
654 Handle per-OS dynamic-module compilation flags.
656 PP/02 Let /dev/null have normal permissions.
657 The 4.73 fixes were a little too stringent and complained about the
658 permissions on /dev/null. Exempt it from some checks.
659 Reported by Andreas M. Kirchwitz.
661 PP/03 Report version information for many libraries, including
662 Exim version information for dynamically loaded libraries. Created
663 version.h, now support a version extension string for distributors
664 who patch heavily. Dynamic module ABI change.
666 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
667 privilege escalation vulnerability whereby the Exim run-time user
668 can cause root to append content of the attacker's choosing to
671 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
674 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
675 If dropping privileges for untrusted macros, we disabled normal logging
676 on the basis that it would fail; for the Exim run-time user, this is not
677 the case, and it resulted in successful deliveries going unlogged.
678 Fixed. Reported by Andreas Metzler.
684 PP/01 Date: & Message-Id: revert to normally being appended to a message,
685 only prepend for the Resent-* case. Fixes regression introduced in
686 Exim 4.70 by NM/22 for Bugzilla 607.
688 PP/02 Include check_rfc2047_length in configure.default because we're seeing
689 increasing numbers of administrators be bitten by this.
691 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
693 PP/03 Bugzilla 994: added openssl_options main configuration option.
695 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
697 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
699 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
701 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
702 without a peer certificate, leading to a segfault because of an
703 assumption that peers always have certificates. Be a little more
704 paranoid. Problem reported by Martin Tscholak.
706 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
707 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
708 NB: ClamAV planning to remove STREAM in "middle of 2010".
709 CL also introduces -bmalware, various -d+acl logging additions and
710 more caution in buffer sizes.
712 PP/09 Implemented reverse_ip expansion operator.
714 PP/10 Bugzilla 937: provide a "debug" ACL control.
716 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
718 PP/12 Bugzilla 973: Implement --version.
720 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
722 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
724 PP/15 Bugzilla 816: support multiple condition rules on Routers.
726 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
727 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
728 ignore trailing whitespace.
730 JJ/02 prevent non-panic DKIM error from being sent to paniclog
732 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
735 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
736 Notification from Dr Andrew Aitchison.
738 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
739 ExtendedDetectionInfo response format.
740 Notification from John Horne.
742 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
745 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
746 XSL and documented dependency on system catalogs, with examples of how
749 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
752 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
753 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
754 configuration file which is writeable by the Exim user or group.
756 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
757 of configuration files to cover files specified with the -C option if
758 they are going to be used with root privileges, not just the default
761 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
762 option (effectively making it always true).
764 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
765 files to be used while preserving root privileges.
767 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
768 that rogue child processes cannot use them.
770 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
771 run-time user, instead of root.
773 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
774 Exim run-time user without dropping privileges.
776 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
777 result string, instead of calling string_vformat() twice with the same
780 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
781 for other users. Others should always drop root privileges if they use
782 -C on the command line, even for a whitelisted configure file.
784 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
786 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
792 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
793 $data_path, and $header_path variables; fixed documentation bugs and
796 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
797 exipick to access non-standard spools, including the "frozen" queue
800 NM/01 Bugzilla 965: Support mysql stored procedures.
801 Patch from Alain Williams
803 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
805 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
806 Patch from Andreas Metzler
808 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
809 Patch from Kirill Miazine
811 NM/05 Bugzilla 671: Added umask to procmail example.
813 JJ/03 installed exipick 20100323.0, fixing doc bug
815 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
816 directory. Notification and patch from Dan Rosenberg.
818 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
820 TK/02 Improve log output when DKIM signing operation fails.
822 MH/01 Treat the transport option dkim_domain as a colon separated
823 list, not as a single string, and sign the message with each element,
824 omitting multiple occurences of the same signer.
826 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
827 Bugzilla 985, 986. Patch by Simon Arlott
829 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
830 Patch by Simon Arlott
832 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
833 MBX locking. Notification from Dan Rosenberg.
839 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
841 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
843 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
845 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
847 NM/04 Bugzilla 915: Flag broken perl installation during build.
853 TK/01 Added patch by Johannes Berg that expands the main option
854 "spamd_address" if it starts with a dollar sign.
856 TK/02 Write list of recipients to X-Envelope-Sender header when building
857 the mbox-format spool file for content scanning (suggested by Jakob
860 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
861 (http://www.dcc-servers.net/) support via dccifd. Activated by
862 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
864 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
865 by Mark Daniel Reidel <mr@df.eu>.
867 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
868 When building exim an external PCRE library is now needed -
869 PCRE is a system library on the majority of modern systems.
870 See entry on PCRE_LIBS in EDITME file.
872 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
873 conversation. Added nologin parameter to request.
874 Patch contributed by Kirill Miazine.
876 TF/01 Do not log submission mode rewrites if they do not change the address.
878 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
880 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
881 log files in place. Contributed by Roberto Lima.
883 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
885 TF/03 Bugzilla 615: When checking the local_parts router precondition
886 after a local_part_suffix or local_part_prefix option, Exim now
887 does not use the address's named list lookup cache, since this
888 contains cached lookups for the whole local part.
890 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
891 Robert Millan. Documentation is in experimental-spec.txt.
893 TF/04 Bugzilla 668: Fix parallel build (make -j).
895 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
897 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
898 Patch provided by Jan Srzednicki.
900 TF/05 Leading white space used to be stripped from $spam_report which
901 wrecked the formatting. Now it is preserved.
903 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
904 that they are available at delivery time.
906 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
908 TF/08 TLS error reporting now respects the incoming_interface and
909 incoming_port log selectors.
911 TF/09 Produce a more useful error message if an SMTP transport's hosts
912 setting expands to an empty string.
914 NM/06 Bugzilla 744: EXPN did not work under TLS.
915 Patch provided by Phil Pennock.
917 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
918 Patch provided by Richard Godbee.
920 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
921 acl_smtp_notquit, added index entry.
923 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
924 Patch provided by Eugene Bujak.
926 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
927 accept(). Patch provided by Maxim Dounin.
929 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
930 Patch provided by Phil Pennock.
932 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
934 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
935 Patch provided by Brad "anomie" Jorsch.
937 NM/14 Bugzilla 622: Added timeout setting to transport filter.
938 Patch provided by Dean Brooks.
940 TK/05 Add native DKIM support (does not depend on external libraries).
942 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
943 Patch provided by Graeme Fowler.
945 NM/16 Bugzilla 851: Documentation example syntax fix.
947 NM/17 Changed NOTICE file to remove references to embedded PCRE.
949 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
952 NM/19 Bugzilla 745: TLS version reporting.
953 Patch provided by Phil Pennock.
955 NM/20 Bugzilla 167: bool: condition support.
956 Patch provided by Phil Pennock.
958 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
959 clients. Patch provided by Phil Pennock.
961 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
962 Patch provided by Brad "anomie" Jorsch.
964 NM/23 Bugzilla 687: Fix misparses in eximstats.
965 Patch provided by Heiko Schlittermann.
967 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
968 Patch provided by Heiko Schlittermann.
970 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
971 plus update to original patch.
973 NM/26 Bugzilla 799: Documentation correction for ratelimit.
975 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
976 Patch provided by David Brownlee.
978 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
980 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
982 NM/30 Bugzilla 888: TLS documentation bugfixes.
984 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
986 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
987 Unlike the original bugzilla I have changed all shell scripts in src tree.
989 NM/33 Bugzilla 898: Transport filter timeout fix.
990 Patch by Todd Rinaldo.
992 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
993 Patch by Serge Demonchaux.
995 NM/35 Bugzilla 39: Base64 decode bug fixes.
996 Patch by Jakob Hirsch.
998 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1000 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1002 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1004 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1010 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1011 ALT-N's libdkim that I have put here:
1012 http://duncanthrax.net/exim-experimental/
1014 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1015 were called 'true' and it seems that C99 defines that as a reserved
1016 keyword to be used with 'bool' variable types. That means you could
1017 not include C99-style headers which use bools without triggering
1018 build errors in sieve.c.
1020 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1021 as mailq or other aliases. Changed the --help handling significantly
1022 to do whats expected. exim_usage() emits usage/help information.
1024 SC/01 Added the -bylocaldomain option to eximstats.
1026 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1028 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1030 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1036 PH/01 Another patch from the Sieve maintainer.
1038 PH/02 When an IPv6 address is converted to a string for single-key lookup
1039 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1040 dots are used instead of colons so that keys in lsearch files need not
1041 contain colons. This was done some time before quoting was made available
1042 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1043 (notated using the quote facility) so as to distinguish them from IPv4
1044 keys. This meant that lookups for IP addresses in host lists did not work
1045 for iplsearch lookups.
1047 This has been fixed by arranging for IPv6 addresses to be expressed with
1048 colons if the lookup type is iplsearch. This is not incompatible, because
1049 previously such lookups could never work.
1051 The situation is now rather anomolous, since one *can* have colons in
1052 ordinary lsearch keys. However, making the change in all cases is
1053 incompatible and would probably break a number of configurations.
1055 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1058 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1059 conversion specification without a maximum field width, thereby enabling
1060 a rogue spamd server to cause a buffer overflow. While nobody in their
1061 right mind would setup Exim to query an untrusted spamd server, an
1062 attacker that gains access to a server running spamd could potentially
1063 exploit this vulnerability to run arbitrary code as the Exim user.
1065 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1066 $primary_hostname instead of what libspf2 thinks the hosts name is.
1068 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1069 a directory entry by the name of the lookup key. Previously, if a
1070 symlink pointed to a non-existing file or a file in a directory that
1071 Exim lacked permissions to read, a lookup for a key matching that
1072 symlink would fail. Now it is enough that a matching directory entry
1073 exists, symlink or not. (Bugzilla 503.)
1075 PH/03 The body_linecount and body_zerocount variables are now exported in the
1078 PH/04 Added the $dnslist_matched variable.
1080 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1081 This means they are set thereafter only if the connection becomes
1084 PH/06 Added the client_condition to authenticators so that some can be skipped
1085 by clients under certain conditions.
1087 PH/07 The error message for a badly-placed control=no_multiline_responses left
1088 "_responses" off the end of the name.
1090 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1092 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1093 (without spaces) instead of just copying the configuration text.
1095 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1097 PH/11 Added $max_received_linelength.
1099 PH/12 Added +ignore_defer and +include_defer to host lists.
1101 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1102 way in which PCRE > 7.0 is built.
1104 PH/14 Implemented queue_only_load_latch.
1106 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1107 MAIL command. The effect was to mangle the value on 64-bit systems.
1109 PH/16 Another patch from the Sieve maintainer.
1111 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1113 PH/18 If a system quota error occurred while trying to create the file for
1114 a maildir delivery, the message "Mailbox is full" was not appended to the
1115 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1116 only to a quota excession during the actual writing of the file.
1118 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1119 characters?) which causes problems in log lines. The DN values are now
1120 passed through string_printing() before being added to log lines.
1122 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1123 and InterBase are left for another time.)
1125 PH/21 Added message_body_newlines option.
1127 PH/22 Guard against possible overflow in moan_check_errorcopy().
1129 PH/23 POSIX allows open() to be a macro; guard against that.
1131 PH/24 If the recipient of an error message contained an @ in the local part
1132 (suitably quoted, of course), incorrect values were put in $domain and
1133 $local_part during the evaluation of errors_copy.
1139 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1140 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1143 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1144 issue a MAIL command.
1146 PH/02 In an ACL statement such as
1148 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1150 if a client was not listed at all, or was listed with a value other than
1151 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1152 the condition was not true (as it should be), so access was not denied.
1153 The bug was that the ! inversion was incorrectly passed on to the second
1154 item. This has been fixed.
1156 PH/03 Added additional dnslists conditions == and =& which are different from
1157 = and & when the dns lookup returns more than one IP address.
1159 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1160 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1162 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1163 FSYNC, which compiles an option called disable_fsync that allows for
1164 bypassing fsync(). The documentation is heavily laced with warnings.
1166 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1168 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1169 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1170 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1171 including adding "make clean"; (3) Added -fPIC when compiling the test
1172 dynamically loaded module, to get rid of a warning.
1174 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1175 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1176 = 0s. The bug is otherwise harmless.
1178 PH/07 There was a bug in the dovecot authenticator such that the value of
1179 $auth1 could be overwritten, and so not correctly preserved, after a
1180 successful authentication. This usually meant that the value preserved by
1181 the server_setid option was incorrect.
1183 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1185 PH/09 Installed PCRE release 7.0.
1187 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1188 run for batched SMTP input. It is now run at the start of every message
1189 in the batch. While fixing this I discovered that the process information
1190 (output by running exiwhat) was not always getting set for -bs and -bS
1191 input. This is fixed, and it now also says "batched" for BSMTP.
1193 PH/11 Added control=no_pipelining.
1195 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1196 patch, slightly modified), and move the expansion of helo_data till after
1197 the connection is made in the smtp transport (so it can use these
1200 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1202 PH/14 Added log_selector = +pid.
1204 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1206 PH/16 Add ${if forany and ${if forall.
1208 PH/17 Added dsn_from option to vary the From: line in DSNs.
1210 PH/18 Flush SMTP output before performing a callout, unless control =
1211 no_callout_flush is set.
1213 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1214 was true (the default) a successful delivery failed to delete the retry
1215 item, thus causing premature timeout of the address. The bug is now
1218 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1220 PH/21 Long custom messages for fakedefer and fakereject are now split up
1221 into multiline reponses in the same way that messages for "deny" and
1222 other ACL rejections are.
1224 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1225 with slight modification.
1227 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1228 draft, changing the syntax and factoring some duplicate code".
1230 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1231 for deliveries of the second and subsequent messages over the same SMTP
1234 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1235 ${reduce, with only minor "tidies".
1237 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1239 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1240 expansion side effects.
1242 PH/27 When a message times out after an over-quota error from an Exim-imposed
1243 quota, the bounce message says "mailbox is full". This message was not
1244 being given when it was a system quota that was exceeded. It now should
1247 MH/03 Made $recipients available in local_scan(). local_scan() already has
1248 better access to the recipient list through recipients_list[], but
1249 $recipients can be useful in postmaster-provided expansion strings.
1251 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1252 in the case of a MAIL command with additional options following the
1253 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1254 were accidentally chopped off.
1256 PH/29 SMTP synchronization checks are implemented when a command is read -
1257 there is a check that no more input is waiting when there shouldn't be
1258 any. However, for some commands, a delay in an ACL can mean that it is
1259 some time before the response is written. In this time, more input might
1260 arrive, invalidly. So now there are extra checks after an ACL has run for
1261 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1262 pipelining has not been advertised.
1264 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1266 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1267 returned in the SMTP response when smtp_return_error_details was false.
1268 This has been fixed.
1270 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1271 instead of the C I/O that was originally supplied, because problems were
1272 reported on Solaris.
1274 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1275 Exim which did not show up earlier: it was assuming that a call to
1276 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1277 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1278 was a macro that became an assignment, so it seemed to work. This has
1279 changed to a proper function call with a void return, hence the compile
1280 error. Exim's code has been fixed.
1282 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1285 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1287 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1289 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1290 because it turns out to be too big or there is a timeout) there is no
1291 "Completed" line in the log. When some messages of this type were
1292 selected by exigrep, they were listed as "not completed". Others were
1293 picked up by some special patterns. I have improved the selection
1294 criteria to be more general.
1296 PH/38 The host_find_failed option in the manualroute router can now be set
1297 to "ignore", to completely ignore a host whose IP address cannot be
1298 found. If all hosts are ignored, the behaviour is controlled by the new
1299 host_all_ignored option.
1301 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1302 homing or because of multiple MX records with /mx) generated more than
1303 one IP address, and the following item turned out to be the local host,
1304 all the secondary addresses of the first item were incorrectly removed
1305 from the list, along with the local host and any following hosts (which
1306 is what is supposed to happen).
1308 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1309 whoever called Exim into the -H file. In the case of the daemon it was
1310 behaving confusingly. When first started, it used values for whoever
1311 started the daemon, but after a SIGHUP it used the Exim user (because it
1312 calls itself on a restart). I have changed the code so that it now always
1315 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1316 message are rejected with the same error (e.g. no authentication or bad
1317 sender address), and a DATA command is nevertheless sent (as can happen
1318 with PIPELINING or a stupid MUA), the error message that was given to the
1319 RCPT commands is included in the rejection of the DATA command. This is
1320 intended to be helpful for MUAs that show only the final error to their
1323 PH/42 Another patch from the Sieve maintainer.
1325 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1326 Eximstats - Fixed some broken HTML links and added missing column headers
1328 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1329 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1331 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1333 PH/43 Yet another patch from the Sieve maintainer.
1335 PH/44 I found a way to check for a TCP/IP connection going away before sending
1336 the response to the final '.' that terminates a message, but only in the
1337 case where the client has not sent further data following the '.'
1338 (unfortunately, this is allowed). However, in many cases there won't be
1339 any further data because there won't be any more messages to send. A call
1340 to select() can be used: if it shows that the input is "ready", there is
1341 either input waiting, or the socket has been closed. An attempt to read
1342 the next input character can distinguish the two cases. Previously, Exim
1343 would have sent an OK response which the client would never have see.
1344 This could lead to message repetition. This fix should cure that, at
1345 least in a lot of common cases.
1347 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1348 advertised in response to EHLO.
1354 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1355 fixed by 4.65/MH/01 (is this a record?) are fixed:
1357 (i) An empty string was always treated as zero by the numeric comparison
1358 operators. This behaviour has been restored.
1360 (ii) It is documented that the numeric comparison operators always treat
1361 their arguments as decimal numbers. This was broken in that numbers
1362 starting with 0 were being interpreted as octal.
1364 While fixing these problems I realized that there was another issue that
1365 hadn't been noticed. Values of message_size_limit (both the global option
1366 and the transport option) were treated as octal if they started with 0.
1367 The documentation was vague. These values are now always treated as
1368 decimal, and I will make that clear in the documentation.
1374 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1375 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1378 MH/01 Don't check that the operands of numeric comparison operators are
1379 integers when their expansion is in "skipping" mode (fixes bug
1380 introduced by 4.64-PH/07).
1382 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1383 child addresses, Exim now panics and dies. Previously, because the count
1384 is held in a short int, deliveries were likely to be lost. As such a
1385 large number of recipients for a single message is ridiculous
1386 (performance will be very, very poor), I have chosen to impose a limit
1387 rather than extend the field.
1393 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1394 leftover -K file (the existence of which was triggered by #402).
1395 While we were at it, introduced process PID as part of the -K
1396 filename. This should rule out race conditions when creating
1399 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1400 processing considerably. Previous code took too long for large mails,
1401 triggering a timeout which in turn triggers #401.
1403 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1404 in the DK code in transports.c. sendfile() is not really portable,
1405 hence the _LINUX specificness.
1407 TF/01 In the add_headers option to the mail command in an Exim filter,
1408 there was a bug that Exim would claim a syntax error in any
1409 header after the first one which had an odd number of characters
1412 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1413 callout verification, Exim cached a "reject" for the entire domain. This
1414 is correct for most verifications, but it is not correct for a recipient
1415 verification with use_sender or use_postmaster set, because in that case
1416 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1417 case of MAIL FROM:<> rejection from other early rejections (e.g.
1418 rejection of HELO). When verifying a recipient using a non-null MAIL
1419 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1420 Whatever the result of the callout, the value of the domain cache is
1421 left unchanged (for any other kind of callout, getting as far as trying
1422 RCPT means that the domain itself is ok).
1424 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1427 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1428 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1429 session, the daemon ignores ECONNECT errors and logs others; it now
1430 ignores EPIPE as well.
1432 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1433 (quoted-printable decoding).
1435 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1436 later the small subsequent patch to fix an introduced bug.
1438 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1440 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1442 PH/08 An error is now given if message_size_limit is specified negative.
1444 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1445 to be given (somewhat) arbitrary names.
1447 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1450 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1451 miscellaneous code fixes
1453 PH/10 Added the log_reject_target ACL modifier to specify where to log
1456 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1457 hostname. This is wrong, because it relates to the incoming message (and
1458 probably the interface on which it is arriving) and not to the outgoing
1459 callout (which could be using a different interface). This has been
1460 changed to use the value of the helo_data option from the smtp transport
1461 instead - this is what is used when a message is actually being sent. If
1462 there is no remote transport (possible with a router that sets up host
1463 addresses), $smtp_active_hostname is used.
1465 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1466 tweaks were necessary in order to get it to work (see also 21 below):
1467 (a) The code assumed that strncpy() returns a negative number on buffer
1468 overflow, which isn't the case. Replaced with Exim's string_format()
1470 (b) There were several signed/unsigned issues. I just did the minimum
1471 hacking in of casts. There is scope for a larger refactoring.
1472 (c) The code used strcasecmp() which is not a standard C function.
1473 Replaced with Exim's strcmpic() function.
1474 (d) The code set only $1; it now sets $auth1 as well.
1475 (e) A simple test gave the error "authentication client didn't specify
1476 service in request". It would seem that Dovecot has changed its
1477 interface. Fortunately there's a specification; I followed it and
1478 changed what the client sends and it appears to be working now.
1480 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1483 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1484 address A is aliased to B and C, where B exists and C does not. Without
1485 -v the output is "A verified" because verification stops after a
1486 successful redirection if more than one address is generated. However,
1487 with -v the child addresses are also verified. Exim was outputting "A
1488 failed to verify" and then showing the successful verification for C,
1489 with its parentage. It now outputs "B failed to verify", showing B's
1490 parentage before showing the successful verification of C.
1492 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1493 look up a TXT record in a specific list after matching in a combined
1496 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1497 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1498 they consult the DNS. I had assumed they would set it the way they
1499 wanted; and indeed my experiments on Linux seem to show that in some
1500 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1501 To be on the safe side, however, I have now made the interface to
1502 host_find_byname() similar to host_find_bydns(), with an argument
1503 containing the DNS resolver options. The host_find_byname() function now
1504 sets these options at its start, just as host_find_bydns() does. The smtp
1505 transport options dns_qualify_single and dns_search_parents are passed to
1506 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1507 of host_find_byname() use the default settings of RES_DEFNAMES
1508 (qualify_single) but not RES_DNSRCH (search_parents).
1510 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1511 spool_read_header() do less string testing, by means of a preliminary
1512 switch on the second character of optional "-foo" lines. (This is
1513 overdue, caused by the large number of possibilities that now exist.
1514 Originally there were few.) While I was there, I also converted the
1515 str(n)cmp tests so they don't re-test the leading "-" and the first
1516 character, in the hope this might squeeze out yet more improvement.
1518 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1519 flag allowing group syntax was set by the header_syntax check but not
1520 turned off, possible causing trouble later; (2) The flag was not being
1521 set at all for the header_verify test, causing "group"-style headers to
1522 be rejected. I have now set it in this case, and also caused header_
1523 verify to ignore an empty address taken from a group. While doing this, I
1524 came across some other cases where the code for allowing group syntax
1525 while scanning a header line wasn't quite right (mostly, not resetting
1526 the flag correctly in the right place). These bugs could have caused
1527 trouble for malformed header lines. I hope it is now all correct.
1529 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1530 with the "reply" argument non-NULL. The code, however (which originally
1531 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1532 but it didn't always do it. This confused somebody who was copying the
1533 code for some other use. I have removed all the tests.
1535 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1536 feature that was used to support insecure browsers during the U.S. crypto
1537 embargo. It requires special client support, and Exim is probably the
1538 only MTA that supported it -- and would never use it because real RSA is
1539 always available. This code has been removed, because it had the bad
1540 effect of slowing Exim down by computing (never used) parameters for the
1541 RSA_EXPORT functionality.
1543 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1544 authenticator to fail if there's a tab character in the incoming data
1545 (there should never be unless someone is messing about, as it's supposed
1546 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1547 if the connection is using TLS or if the remote IP is the same as the
1548 local IP, and the "valid-client-cert option" if a client certificate has
1551 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1552 authenticators. This can be used for authorization after authentication
1553 succeeds. (In the case of plaintext, it servers for both authentication
1556 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1557 if any retry times were supplied.
1559 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1560 connection, where there is no client IP address to check. In this
1561 situation, the verify now always succeeds.
1563 PH/25 Applied John Jetmore's -Mset patch.
1565 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1567 PH/27 In a string expansion for a processed (not raw) header when multiple
1568 headers of the same name were present, leading whitespace was being
1569 removed from all of them, but trailing whitespace was being removed only
1570 from the last one. Now trailing whitespace is removed from each header
1571 before concatenation. Completely empty headers in a concatenation (as
1572 before) are ignored.
1574 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1575 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1577 PH/29 [Removed. This was a change that I later backed out, and forgot to
1578 correct the ChangeLog entry (that I had efficiently created) before
1579 committing the later change.]
1581 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1582 address errors (4xx response to RCPT) over the same connection as other
1583 messages routed to the same hosts. Such deliveries are always "forced",
1584 so retry times are not inspected. This resulted in far too many retries
1585 for the affected addresses. The effect occurred only when there were more
1586 hosts than the hosts_max_try setting in the smtp transport when it had
1587 the 4xx errors. Those hosts that it had tried were not added to the list
1588 of hosts for which the message was waiting, so if all were tried, there
1589 was no problem. Two fixes have been applied:
1591 (i) If there are any address or message errors in an SMTP delivery, none
1592 of the hosts (tried or untried) are now added to the list of hosts
1593 for which the message is waiting, so the message should not be a
1594 candidate for sending over the same connection that was used for a
1595 successful delivery of some other message. This seems entirely
1596 reasonable: after all the message is NOT "waiting for some host".
1597 This is so "obvious" that I'm not sure why it wasn't done
1598 previously. Hope I haven't missed anything, but it can't do any
1599 harm, as the worst effect is to miss an optimization.
1601 (ii) If, despite (i), such a delivery is accidentally attempted, the
1602 routing retry time is respected, so at least it doesn't keep
1603 hammering the server.
1605 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1606 in ${readsocket because some servers need this prod.
1608 PH/32 Added some extra debug output when updating a wait-xxx database.
1610 PH/33 The hint "could be header name not terminated by colon", which has been
1611 given for certain expansion errors for a long time, was not being given
1612 for the ${if def:h_colon_omitted{... case.
1614 PH/34 The spec says: "With one important exception, whenever a domain list is
1615 being scanned, $domain contains the subject domain." There was at least
1616 one case where this was not true.
1618 PH/35 The error "getsockname() failed: connection reset by peer" was being
1619 written to the panic log as well as the main log, but it isn't really
1620 panic-worthy as it just means the connection died rather early on. I have
1621 removed the panic log writing for the ECONNRESET error when getsockname()
1624 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1625 runs only) independently of the message's sender address. This meant
1626 that, if the 4xx error was in fact related to the sender, a different
1627 message to the same recipient with a different sender could confuse
1628 things. In particualar, this can happen when sending to a greylisting
1629 server, but other circumstances could also provoke similar problems.
1630 I have changed the default so that the retry time for these errors is now
1631 based a combination of the sender and recipient addresses. This change
1632 can be overridden by setting address_retry_include_sender=false in the
1635 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1636 remote server are returned as part of bounce messages. This was not
1637 happening for LMTP over a pipe (the lmtp transport), but now it is the
1638 same for both kinds of LMTP.
1640 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1641 in header lines that were in fact CNAMEs. This is no longer the case.
1643 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1644 and queue runs started by the daemon processed all messages. This has
1645 been fixed so that -R and -S can now usefully be given with -q<time>.
1647 PH/40 Import PCRE release 6.7 (fixes some bugs).
1649 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1651 PH/42 Give an error if -q is specified more than once.
1653 PH/43 Renamed the variables $interface_address and $interface_port as
1654 $received_ip_address and $received_port, to make it clear that these
1655 values apply to message reception, and not to the outgoing interface when
1656 a message is delivered. (The old names remain recognized, of course.)
1658 PH/44 There was no timeout on the connect() call when using a Unix domain
1659 socket in the ${readsocket expansion. There now is.
1661 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1662 be meaningful with "accept".
1664 SC/01 Eximstats V1.43
1665 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1667 SC/02 Eximstats V1.44
1668 Use a glob alias rather than an array ref in the generated
1669 parser. This improves both readability and performance.
1671 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1672 Collect SpamAssassin and rejection statistics.
1673 Don't display local sender or destination tables unless
1674 there is data to show.
1675 Added average volumes into the top table text output.
1677 SC/04 Eximstats V1.46
1678 Collect data on the number of addresses (recipients)
1679 as well as the number of messages.
1681 SC/05 Eximstats V1.47
1682 Added 'Message too big' to the list of mail rejection
1683 reasons (thanks to Marco Gaiarin).
1685 SC/06 Eximstats V1.48
1686 Mainlog lines which have GMT offsets and are too short to
1687 have a flag are now skipped.
1689 SC/07 Eximstats V1.49 (Alain Williams)
1690 Added the -emptyok flag.
1692 SC/08 Eximstats V1.50
1693 Fixes for obtaining the IP address from reject messages.
1695 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1696 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1697 whitesspace changes from 4.64-PH/27
1699 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1702 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1703 are found, allow negative numbers in numeric criteria)
1705 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1707 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1708 to match changes made in 4.64-PH/43
1710 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1712 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1713 "domain" is actually an IP address, and force a failure. This locks out
1714 those revolvers/nameservers that support "A-for-A" lookups, in
1715 contravention of the specifications.
1717 PH/48 When a host name was looked up from an IP address, and the subsequent
1718 forward lookup of the name timed out, the host name was left in
1719 $sender_host_name, contrary to the specification.
1721 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1722 restricted to single-key lookups, Exim was not diagnosing an error if
1723 * or *@ was used with a query-style lookup.
1725 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1727 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1728 long ago, but noone interested enough thought of it. Let's just say that
1729 the "1.1" means that there are some new functions that weren't there at
1730 some point in the past.
1732 PH/51 Error processing for expansion failure of helo_data from an smtp
1733 transport during callout processing was broken.
1735 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1736 tested/used via the -bh/-bhc/-bs options.
1738 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1739 bug, fixed in subsequent PCRE releases).
1741 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1742 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1744 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1750 SC/01 Use a glob alias rather than an array ref in eximstats generated
1751 parser. This improves both readability and performance.
1753 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1754 Don't display local sender or destination tables in eximstats unless
1755 there is data to show.
1756 Added average volumes into the eximstats top table text output.
1758 SC/03 Collect data on the number of addresses (recipients) as well
1759 as the number of messages in eximstats.
1761 TF/01 Correct an error in the documentation for the redirect router. Exim
1762 does (usually) call initgroups() when daemonizing.
1764 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1765 with consistent privilege compared to when running as a daemon.
1767 TF/03 Note in the spec that $authenticated_id is not set for local
1768 submissions from trusted users.
1770 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1771 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1773 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1774 by adding some example configuration directives to the default
1775 configuration file. A little bit of work is required to uncomment the
1776 directives and define how usernames and passwords are checked, but
1777 there is now a framework to start from.
1779 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1780 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1781 without this. I don't know how relevant this is to other LDAP libraries.
1783 PH/02 Add the verb name to the "unknown ACL verb" error.
1785 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1787 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1789 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1790 directory not expanded when it should be if an expanded home directory
1791 was set for the address (which is overridden by the transport).
1793 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1796 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1797 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1798 because it is too late at that time, and has no effect.
1800 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1801 security issue with \' (bugzilla #107). I could not use the
1802 PQescapeStringConn() function, because it needs a PGconn value as one of
1805 PH/08 When testing addresses using -bt, indicate those final addresses that
1806 are duplicates that would not cause an additional delivery. At least one
1807 person was confused, thinking that -bt output corresponded to deliveries.
1808 (Suppressing duplicates isn't a good idea as you lose the information
1809 about possibly different redirections that led to the duplicates.)
1811 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1812 systems where poll() doesn't work, in particular OS X.
1814 PH/10 Added more information to debugging output for retry time not reached.
1816 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1817 operations in malware.c.
1819 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1822 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1823 syslog_duplication set false, log lines that would normally be written
1824 both the the main log and to the reject log were not written to syslog at
1827 PH/14 In the default configuration, change the use of "message" in ACL warn
1828 statements to "add_header".
1830 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1831 not followed by a command (e.g. "seen endif").
1833 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1834 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1837 PH/17 Added extra conditions to the default value of delay_warning_condition
1841 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1842 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1843 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1846 The Auto-Submitted: and various List- headers are standardised, whereas I
1847 don't think Precedence: ever was.
1849 PH/18 Refactored debugging code in route_finduser() to show more information,
1850 in particular, the error code if getpwnam() issues one.
1852 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1853 This is apparently needed in addition to the PH/07 change above to avoid
1854 any possible encoding problems.
1856 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1857 but not after initializing Perl.
1859 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1860 output them only if debugging. By default they are written stderr,
1861 apparently, which is not desirable.
1863 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1866 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1869 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1871 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1872 authenticated or an ident call has been made. Suppress the default
1873 values for $authenticated_id and $authenticated_sender (but permit -oMai
1874 and -oMas) when testing with -bh.
1876 PH/24 Re-jigged the order of the tests in the default configuration so that the
1877 tests for valid domains and recipients precede the DNS black list and CSA
1878 tests, on the grounds that those ones are more expensive.
1880 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1881 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1882 This bug exists in every version of Exim that I still have, right back to
1885 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1886 However, an attempt to turn on case-sensitivity in a regex key by
1887 including (?-i) didn't work because the subject string was already
1888 lowercased, and the effects were non-intuitive. It turns out that a
1889 one-line patch can be used to allow (?-i) to work as expected.
1895 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1896 other effects) broke the use of negated acl sub-conditions.
1898 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1901 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1902 "Deny" causes Exim to reject the incoming connection with a 554 error.
1903 Unfortunately, if there is a major crisis, such as a disk failure,
1904 tcp-wrappers gives "deny", whereas what one would like would be some
1905 kind of temporary error. A kludge has been added to help with this.
1906 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1907 554 error is used if errno is still zero or contains ENOENT (which occurs
1908 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1911 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1913 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1914 errors. Otherwise a message that provokes a temporary error (when other
1915 messages do not) can cause a whole host to time out.
1917 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1918 addresses were routed directly to files or pipes from a redirect router.
1919 File deliveries just didn't batch; pipe deliveries might have suffered
1922 PH/06 A failure to get a lock for a hints database would erroneously always say
1923 "Failed to get write lock", even when it was really a read lock.
1925 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1926 of 0600. This has been changed to use the value of the lockfile_mode
1927 option (which defaults to 0600).
1929 PH/08 Applied small patch from the Sieve maintainer.
1931 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1932 folder from quota calculations, a direct delivery into this folder messed
1933 up the contents of the maildirsize file. This was because the regex was
1934 used only to exclude .Trash (or whatever) when the size of the mailbox
1935 was calculated. There was no check that a delivery was happening into an
1936 excluded directory. This bug has been fixed by ignoring all quota
1937 processing for deliveries into excluded directories.
1939 PH/10 Added the maildirfolder_create_regex option to appendfile.
1945 PH/01 The code for finding all the local interface addresses on a FreeBSD
1946 system running IPv6 was broken. This may well have applied to all BSD
1947 systems, as well as to others that have similar system calls. The broken
1948 code found IPv4 interfaces correctly, but gave incorrect values for the
1949 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1950 that it would not match correctly against @[] and not recognize the IPv6
1953 PH/02 The ipliteral router was not recognizing addresses of the form user@
1954 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1956 PH/03 Added disable_ipv6.
1958 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1959 decoded form, because it is most often used to construct To: headers
1960 lines in autoreplies, and the decoded form may well be syntactically
1961 invalid. However, $reply_address has leading white space removed, and all
1962 newlines turned into spaces so that the autoreply transport does not
1965 PH/05 If group was specified without a user on a router, and no group or user
1966 was specified on a transport, the group from the router was ignored.
1968 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1969 for visible compile-time settings that can be used to change these
1970 numbers, for those that want even more. Backwards compatibility with old
1971 spool files has been maintained. However, going back to a previous Exim
1972 release will lost any variables that are in spool files.
1974 PH/07 Two small changes when running in the test harness: increase delay when
1975 passing a TCP/IP connection to a new process, in case the original
1976 process has to generate a bounce, and remove special handling of
1977 127.0.0.2 (sic), which is no longer necessary.
1979 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1980 be the same on different OS.
1982 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1985 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1986 whether --show-vars was specified or not
1988 JJ/02 exipick: Added support for new ACL variable spool format introduced
1991 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1992 syntactically invalid From: or Reply-to: line, and a filter used this to
1993 generate an autoreply, and therefore failed to obtain an address for the
1994 autoreply, Exim could try to deliver to a non-existent relative file
1995 name, causing unrelated and misleading errors. What now happens is that
1996 it logs this as a hard delivery error, but does not attempt to create a
1999 PH/11 The exinext utility has a -C option for testing purposes, but although
2000 the given file was scanned by exinext itself; it wasn't being passed on
2001 when Exim was called.
2003 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2004 an end-of-file indication when reading a command response.
2006 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2007 compiled. In many other places in Exim, IPv6 addresses are always
2008 recognized, so I have changed this. It also means that IPv4 domain
2009 literals of the form [IPV4:n.n.n.n] are now always recognized.
2011 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2012 used if the router is not running as root, for example, when verifying at
2013 ACL time, or when using -bh. The debugging output from this situation was
2014 non-existent - all you got was a failure to exec. I have made two
2017 (a) Failures to set uid/gid, the current directory, or a process leader
2018 in a subprocess such as that created by queryprogram now generate
2019 suitable debugging ouput when -d is set.
2021 (b) The queryprogram router detects when it is not running as root,
2022 outputs suitable debugging information if -d is set, and then runs
2023 the subprocess without attempting to change uid/gid.
2025 PH/15 Minor change to Makefile for building test_host (undocumented testing
2028 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2029 additional section of a DNS packet that returns MX or SRV records.
2030 Instead, it always explicitly searches for A/AAAA records. This avoids
2031 major problems that occur when a DNS server includes only records of one
2032 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2033 fixed another bug: if SRV records were looked up and the corresponding
2034 address records were *not* found in the additional section, the port
2035 values from the SRV records were lost.
2037 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2038 using the correct key (the original address) when searching the retry
2039 rules in order to find which one to use for generating the retry hint.
2041 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2042 adding the default one. Similarly, if it contains a Reply-To: header, the
2043 errors_reply_to option, if set, is not used.
2045 PH/19 When calculating a retry time, Exim used to measure the "time since
2046 failure" by looking at the "first failed" field in the retry record. Now
2047 it does not use this if it is later than than the arrival time of the
2048 message. Instead it uses the arrival time. This makes for better
2049 behaviour in cases where some deliveries succeed, thus re-setting the
2050 "first failed" field. An example is a quota failure for a huge message
2051 when small messages continue to be delivered. Without this change, the
2052 "time since failure" will always be short, possible causing more frequent
2053 delivery attempts for the huge message than are intended.
2054 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2056 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2057 $1, $2, $3) because the numerical variables can be reset during some
2058 expansion items (e.g. "match"), thereby losing the authentication data.
2060 PH/21 Make -bV show the size of off_t variables so that the test suite can
2061 decide whether to run tests for quotas > 2G.
2063 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2064 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2065 filecount value is greater than 2G or if a quota value is greater than 2G
2066 on a system where the size of off_t is not greater than 4, a panic error
2069 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2070 never match. The debug and -bh output now contains an explicit error
2071 message indicating a malformed IPv4 address or mask.
2073 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2074 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2075 PH/23 above applies.
2077 PH/25 Do not write to syslog when running in the test harness. The only
2078 occasion when this arises is a failure to open the main or panic logs
2079 (for which there is an explicit test).
2081 PH/26 Added the /no_tell option to "control=freeze".
2083 PH/27 If a host name lookup failed very early in a connection, for example, if
2084 the IP address matched host_lookup and the reverse lookup yielded a name
2085 that did not have a forward lookup, an error message of the form "no IP
2086 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2087 could be logged. Now it outputs the IP address instead of "NULL".
2089 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2090 allows the sender and the authenticated sender to be set when
2091 submitting a message from within Exim. Since child_open_exim() is
2092 documented for local_scan(), the new function should be too.
2094 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2095 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2096 results in an empty string is now treated as unset.
2098 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2100 PH/31 Added sender_verify_fail logging option.
2102 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2103 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2104 tidied the source and removed it altogether.
2106 PH/33 When a queue run was abandoned because the load average was too high, a
2107 log line was always written; now it is written only if the queue_run log
2108 selector is set. In addition, the log line for abandonment now contains
2109 information about the queue run such as the pid. This is always present
2110 in "start" and "stop" lines but was omitted from the "abandon" line.
2112 PH/34 Omit spaces between a header name and the colon in the error message that
2113 is given when verify = headers_syntax fails (if there are lots of them,
2114 the message gets confusing).
2116 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2117 names, as there are now some PTR records that contain slashes. This check
2118 is only to protect against broken name servers that fall over on strange
2119 characters, so the fact that it applies to all lookups doesn't matter.
2121 PH/36 Now that the new test suite is complete, we can remove some of the
2122 special code in Exim that was needed for the old test suite. For example,
2123 sorting DNS records because real resolvers return them in an arbitrary
2124 order. The new test suite's fake resolver always returns records in the
2127 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2128 bounces) except when queue_only is set, to avoid logging races between
2129 the different processes.
2131 PH/38 Panic-die if .include specifies a non-absolute path.
2133 PH/39 A tweak to the "H" retry rule from its user.
2135 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2136 a label. They prevented compilation on older perls.
2138 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2139 a warning to be raised on newish perls.
2141 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2142 on queue. Changes to match documented behaviour of showing count of
2143 messages matching specified criteria.
2145 PH/40 Changed the default ident timeout from 30s to 5s.
2147 PH/41 Added support for the use of login_cap features, on those BSD systems
2148 that have them, for controlling the resources used by pipe deliveries.
2150 PH/42 The content-scanning code uses fopen() to create files in which to put
2151 message data. Previously it was not paying any attention to the mode of
2152 the files. Exim runs with umask(0) because the rest of the code creates
2153 files with open(), and sets the required mode explicitly. Thus, these
2154 files were ending up world-writeable. This was not a big issue, because,
2155 being within the spool directory, they were not world-accessible. I have
2156 created a function called modefopen, which takes an additional mode
2157 argument. It sets umask(777), creates the file, chmods it to the required
2158 mode, then resets the umask. All the relevant calls to fopen() in the
2159 content scanning code have been changed to use this function.
2161 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2162 to 24 hours. This avoids potential overflow problems when processing G
2163 and H retry rules. I suspect nobody ever tinkers with this value.
2165 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2167 PH/45 When the plaintext authenticator is running as a client, the server's
2168 challenges are checked to ensure they are valid base64 strings. By
2169 default, the authentication attempt is cancelled if an invalid string is
2170 received. Setting client_ignore_invalid_base64 true ignores these errors.
2171 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2172 they are received. Thus, the responses can be made to depend on the
2173 challenges. If an invalid string is ignored, an empty string is placed in
2176 PH/46 Messages that are created by the autoreply transport now contains a
2177 References: header, in accordance with RFCs 2822 and 3834.
2179 PH/47 Added authenticated_sender_force to the smtp transport.
2181 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2183 PH/49 Installed latest patch from the Sieve maintainer.
2185 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2186 was also set, the appendfile transport was unnecessarily scanning a
2187 directory of message files (e.g. for maildir delivery) to find the count
2188 of files (along with the size), even though it did not need this
2189 information. It now does the scan only if it needs to find either the
2190 size of the count of files.
2192 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2194 PH/52 Two bugs concerned with error handling when the smtp transport is
2197 (i) Exim was not creating retry information for temporary errors given
2198 for individual recipients after the DATA command when the smtp transport
2199 was used in LMTP mode. This meant that they could be retried too
2200 frequently, and not timed out correctly.
2202 (ii) Exim was setting the flag that allows error details to be returned
2203 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2204 recipients that were returned after the DATA command.
2206 PH/53 This is related to PH/52, but is more general: for any failing address,
2207 when detailed error information was permitted to be returned to the
2208 sender, but the error was temporary, then after the final timeout, only
2209 "retry timeout exceeded" was returned. Now it returns the full error as
2210 well as "retry timeout exceeded".
2212 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2213 do this, and (what is worse) MTAs that accept it.
2215 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2216 will now be deprecated.
2218 PH/56 New os.c-cygwin from the Cygwin maintainer.
2220 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2221 formats (previously only available in exim formats via -bpr, -bpru,
2222 and -bpra. Now also available in native and exiqgrep formats)
2224 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2225 with very large, slow to parse queues
2227 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2229 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2231 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2232 responses to authentication challenges, though it was showing the
2233 challenges; (ii) I've removed the CR characters from the debug output for
2236 PH/58 Allow for the insertion of a newline as well as a space when a string
2237 is turned into more than one encoded-word during RFC 2047 encoding. The
2238 Sieve code now uses this.
2240 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2241 data_4xx, lost_connection, tls_required.
2243 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2244 message was being sent as an SMTP response.
2246 PH/61 Add -l and -k options to exicyclog.
2248 PH/62 When verifying, if an address was redirected to one new address, so that
2249 verification continued, and the new address failed or deferred after
2250 having set something in $address_data, the value of $address_data was not
2251 passed back to the ACL. This was different to the case when no
2252 redirection occurred. The value is now passed back in both cases.
2254 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2255 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2256 use login_cap.h, so on its own it isn't the distinguishing feature. The
2257 new name refers directly to the setclassresources() function.
2259 PH/65 Added configuration files for NetBSD3.
2261 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2263 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2264 is preferred over IPv4.
2266 PH/68 The bounce_return_message and bounce_return_body options were not being
2267 honoured for bounces generated during the reception of non-SMTP messages.
2268 In particular, this applied to messages rejected by the ACL. This bug has
2269 been fixed. However, if bounce_return_message is true and bounce_return_
2270 body is false, the headers that are returned for a non-SMTP message
2271 include only those that have been read before the error was detected.
2272 (In the case of an ACL rejection, they have all been read.)
2274 PH/69 The HTML version of the specification is now built in a directory called
2275 spec_html instead of spec.html, because the latter looks like a path with
2276 a MIME-type, and this confuses some software.
2278 PH/70 Catch two compiler warnings in sieve.c.
2280 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2281 function verify_get_ident() calls ip_connect() to connect a socket, but
2282 if the "connect()" function timed out, ip_connect() used to close the
2283 socket. However, verify_get_ident() also closes the socket later, and in
2284 between Exim writes to the log, which may get opened at this point. When
2285 the socket was closed in ip_connect(), the log could get the same file
2286 descriptor number as the socket. This naturally causes chaos. The fix is
2287 not to close the socket in ip_connect(); the socket should be closed by
2288 the function that creates it. There was only one place in the code where
2289 this was missing, in the iplookup router, which I don't think anybody now
2290 uses, but I've fixed it anyway.
2292 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2293 well as to direct DNS lookups. Otherwise the handling of names in host
2294 lists is inconsistent and therefore confusing.
2300 PH/01 Two changes to the default runtime configuration:
2302 (1) Move the checks for relay_from_hosts and authenticated clients from
2303 after to before the (commented out) DNS black list checks.
2305 (2) Add control=submission to the relay_from_hosts and authenticated
2306 clients checks, on the grounds that messages accepted by these
2307 statements are most likely to be submissions.
2309 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2311 (1) Generate an error if the third argument for the ${prvs expansion is
2314 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2317 (3) Reset the variables that are obtained from the first argument of
2318 ${prvscheck and used in the second argument before leaving the code,
2319 because their memory is reclaimed, so using them afterwards may do
2322 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2323 one (it's much easier than Tom thought :-).
2325 (5) Because of (4), we can now allow for the use of $prvscheck_result
2326 inside the third argument.
2328 PH/03 For some reason, the default setting of PATH when running a command from
2329 a pipe transport was just "/usr/bin". I have changed it to
2332 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2333 anything to be listed in the output from -bV.
2335 PH/05 When a filter generated an autoreply, the entire To: header line was
2336 quoted in the delivery log line, like this:
2338 => >A.N.Other <ano@some.domain> <original@ddress> ...
2340 This has been changed so that it extracts the operative address. There
2341 may be more than one such address. If so, they are comma-separated, like
2344 => >ano@some.domain,ona@other.domain <original@ddress> ...
2346 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2347 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2348 not being looked up in the rDNS to get a host name, Exim was showing the
2349 IP address twice in Received: lines, even though the IP addresses were
2350 identical. For example:
2352 Received: from [1.2.3.4] (helo=[1.2.3.4])
2354 However, if the real host name was known, it was omitting the HELO data
2355 if it matched the actual IP address. This has been tidied up so that it
2356 doesn't show the same IP address twice.
2358 PH/07 When both +timestamp and +memory debugging was on, the value given by
2359 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2360 called by the debug printing, thereby overwriting the timestamp buffer.
2361 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2363 PH/08 When the original message was included in an autoreply transport, it
2364 always said "this is a copy of the message, including all the headers",
2365 even if body_only or headers_only was set. It now gives an appropriate
2368 PH/09 Applied a patch from the Sieve maintainer which:
2370 o fixes some comments
2371 o adds the (disabled) notify extension core
2372 o adds some debug output for the result of if/elsif tests
2373 o points to the current vacation draft in the documentation
2374 and documents the missing references header update
2378 o fixes a bug in processing the envelope test (when testing
2379 multiple envelope elements, the last element determinted the
2382 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2383 Electronic Mail") by including:
2385 Auto-submitted: auto-generated
2387 in the messages that it generates (bounce messages and others, such as
2388 warnings). In the case of bounce messages for non-SMTP mesages, there was
2389 also a typo: it was using "Auto_submitted" (underscore instead of
2390 hyphen). Since every message generated by Exim is necessarily in response
2391 to another message, thes have all been changed to:
2393 Auto-Submitted: auto-replied
2395 in accordance with these statements in the RFC:
2397 The auto-replied keyword:
2399 - SHOULD be used on messages sent in direct response to another
2400 message by an automatic process,
2402 - MUST NOT be used on manually-generated messages,
2404 - MAY be used on Delivery Status Notifications (DSNs) and Message
2405 Disposition Notifications (MDNs),
2407 - MUST NOT be used on messages generated by automatic or periodic
2408 processes, except for messages which are automatic responses to
2411 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2412 to the default Received: header definition.
2414 PH/12 Added log selector acl_warn_skipped (default on).
2416 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2417 variables because (a) they are in the wrong storage pool and (b) even if
2418 they were copied, it wouldn't work properly because of the caching.
2420 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2421 checking when decoding. Apparently there are clients that generate
2422 overlong encoded strings. Why am I not surprised?
2424 PH/15 If the first argument of "${if match_address" was not empty, but did not
2425 contain an "@" character, Exim crashed. Now it writes a panic log message
2426 and treats the condition as false.
2428 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2430 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2431 extension "envelope-auth". The code is finished and in agreement with
2432 other implementations, but there is no documentation so far and in fact,
2433 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2434 not changing the active code.
2436 Print executed "if" and "elsif" statements when debugging is used. This
2437 helps a great deal to understand what a filter does.
2439 Document more things not specified clearly in RFC3028. I had all this
2440 sorted out, when out of a sudden new issues came to my mind. Oops."
2442 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2445 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2446 canonical form (as documented). However, after a host name lookup from
2447 the IP address, check_host() was doing a simple string comparison with
2448 addresses acquired from the DNS when checking that the found name did
2449 have the original IP as one of its addresses. Since any found IPv6
2450 addresses are likely to be in abbreviated form, the comparison could
2451 fail. Luckily, there already exists a function for doing the comparison
2452 by converting both addresses to binary, so now that is used instead of
2453 the text comparison.
2455 PH/20 There was another similar case to PH/19, when a complete host name was
2456 given in a host list; looking up its IP address could give an abbreviated
2457 form, whereas the current host's name might or might not be abbreviated.
2458 The same fix has been applied.
2464 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2465 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2468 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2469 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2471 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2473 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2474 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2475 Also, for "Auto-Submitted", treat anything other than "no" as
2476 non-personal, in accordance with RFC 3834. (Previously it treated
2477 anything starting "auto-" as non-personal.)
2479 TF/01 The control=submission/name=... option had a problem with syntax
2480 errors if the name included a slash character. The /name= option
2481 now slurps the rest of the string, so it can include any characters
2482 but it must come last in the list of options (after /sender_retain
2485 PH/05 Some modifications to the interface to the fake nameserver for the new
2493 TK/01 Added the "success_on_redirect" address verification option. See
2494 NewStuff for rationale and an example.
2496 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2498 PH/02 Patch to exigrep to allow it to work on syslog lines.
2500 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2501 fread() to skip over the body file's header line, because in Cygwin the
2502 header line is locked and is inaccessible.
2504 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2505 co-exist for some time) to make it clear that it is the Exim ID that is
2506 referenced, not the Message-ID: header line.
2508 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2509 string_format() function, because snprintf() does not exist on all
2512 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2513 ${stat: expansion item.
2515 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2516 protocol synchronization error", to keep the pedants happy.
2518 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2519 well as for IRIX systems, when gcc is being used. See the host.c source
2522 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2524 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2527 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2528 transport and to the smtp transport in LMTP mode.
2530 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2532 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2533 run a filter in a subprocess. This could lead to confusion in subsequent
2534 lookups in the parent process. There should also be a search_tidyup() at
2535 the end of the subprocess.
2537 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2538 only if the host matched helo_try_verify_hosts, which caused the
2539 verification to occur when the EHLO/HELO command was issued. The ACL just
2540 tested the remembered result. Now, if a previous verification attempt has
2541 not happened, "verify = helo" does it there and then.
2543 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2545 TK/03 Fix log output including CR from clamd.
2547 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2548 exist provoked a memory error which could cause a segfault.
2550 PH/15 Installed PCRE 6.2
2552 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2554 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2555 of the problem. Specifically, suggested +O2 rather than +O1 for the
2558 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2560 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2561 record, the port was not correctly logged when the outgoing_port log
2562 selector was set (it logged the transort's default port).
2564 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2565 fallback_hosts, and "hosts" in the smtp transport.
2567 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2568 host errors such as "Connection refused".
2570 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2571 authentication with radiusclient 0.4.9:
2573 - Error returned from rc_read_config was caught wrongly
2574 - Username/password not passed on to radius server due to wrong length.
2576 The presumption is that some radiusclient API changes for 4.51/PH/17
2577 were not taken care of correctly. The code is still untested by me (my
2578 Linux distribution still has 0.3.2 of radiusclient), but it was
2579 contributed by a Radius user.
2581 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2582 expanding the "port" option of the smtp transport.
2584 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2585 while reading a MIME header. Thanks to Tom Hughes for a patch.
2587 PH/24 Include config.h inside local_scan.h so that configuration settings are
2590 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2591 that in an ACL for RCPT (for example), you can examine exactly what was
2594 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2595 commands, but it was not correctly comparing the address with the actual
2596 client host address. Thus, it would show the EHLO address in Received:
2597 header lines when this was not necessary.
2599 PH/27 Added the % operator to ${eval:}.
2601 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2602 it should be ignoring failures (because with -C, for example, it has lost
2603 privilege). It wasn't ignoring creation failures other than "already
2606 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2609 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2610 because an input line was too long, either on its own, or by virtue of
2611 too many continuations, the temporary file was not being removed, and the
2612 return code was incorrect.
2614 PH/31 Missing "BOOL" in function definition in filtertest.c.
2616 PH/32 Applied Sieve patches from the maintainer.
2618 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2620 PH/33 Added "verify = not_blind".
2622 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2623 Local/Makefile (with some defaults set). These are used in built scripts
2624 such as exicyclog, but they have never been used in the exim_install
2625 script (though there are many overriding facilities there). I have
2626 arranged that the exim_install script now takes note of these two
2629 PH/35 Installed configuration files for Dragonfly.
2631 PH/36 When a locally submitted message by a trusted user did not contain a
2632 From: header, and the sender address was obtained from -f or from an SMTP
2633 MAIL command, and the trusted user did not use -F to supply a sender
2634 name, $originator_name was incorrectly used when constructing a From:
2635 header. Furthermore, $originator_name was used for submission mode
2636 messages from external hosts without From: headers in a similar way,
2637 which is clearly wrong.
2639 PH/37 Added control=suppress_local_fixups.
2641 PH/38 When log_selector = +received_sender was set, and the addition of the
2642 sender made the log line's construction buffer exactly full, or one byte
2643 less than full, an overflow happened when the terminating "\n" was
2646 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2647 when the result of a list match is failure because a DNS lookup failed.
2649 PH/40 RM_COMMAND is now used in the building process.
2651 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2652 the "build-* directories that it finds.
2654 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2655 address in a domain literal was a prefix of an interface address.
2657 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2658 when verifying a sender address, unless rewrite_headers is false.
2660 PH/44 Wrote a long comment about why errors_to addresses are verified as
2661 recipients, not senders.
2663 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2664 the ratelimit ACL was added.
2666 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2668 PH/46 Added extra information about PostgreSQL errors to the error string.
2670 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2671 avoiding the need to install special zones in a real server. This is
2672 backwards compatible; if it can't find the fake resolver, it drops back.
2673 Thus, both old and new test suites can be run.
2675 TF/02 Added util/ratelimit.pl
2677 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2678 clock is set back in time.
2680 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2681 Candler <B.Candler@pobox.com>.
2683 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2684 OK for addresses that are the result of redirections.
2686 PH/48 A number of further additions for the benefit of the new test suite,
2687 including a fake gethostbyname() that interfaces to the fake DNS resolver
2690 TF/06 The fix for widen_domains has also been applied to qualify_single and
2691 search_parents which are the other dnslookup options that can cause
2694 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2697 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2699 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2700 service" if it encountered an MX record with an empty target hostname.
2701 The message is now "an MX or SRV record indicated no SMTP service".
2703 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2704 if the DNS of the sending site is misconfigured. This is quite a
2705 common situation. This change restores the behaviour of treating a
2706 helo verification defer as a failure.
2708 PH/51 If self=fail was set on a router, the bounce message did not include the
2709 actual error message.
2715 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2717 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2718 command itself ended in error, the underlying message about the transport
2719 filter timeout was being overwritten with the pipe command error. Now the
2720 underlying error message should be appended to the second error message.
2722 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2724 PH/02 Reduce the amount of output that "make" produces by default. Full output
2725 can still be requested.
2727 PH/03 The warning log line about a condition test deferring for a "warn" verb
2728 was being output only once per connection, rather than after each
2729 occurrence (because it was using the same function as for successful
2730 "warn" verbs). This seems wrong, so I have changed it.
2732 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2733 it should not have, which might have caused a crash in the right
2734 circumstances, but probably never did.
2736 PH/04 Installed a modified version of Tony Finch's patch to make submission
2737 mode fix the return path as well as the Sender: header line, and to
2738 add a /name= option so that you can make the user's friendly name appear
2741 TF/03 Added the control = fakedefer ACL modifier.
2743 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2744 Mark Lowes for thorough testing.
2746 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2748 TK/03 Merged latest SRS patch from Miles Wilton.
2750 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2751 with the definition in sysexits.h (which is #included earlier).
2752 Fortunately, Exim does not actually use EX_OK. The code used to try to
2753 preserve the sysexits.h value, by assumimg that macro definitions were
2754 scanned for macro replacements. I have been disabused of this notion,
2755 so now the code just undefines EX_OK before #including unistd.h.
2757 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2758 in the smtp transport. When a block could not be written in a single
2759 write() function, the timeout was being re-applied to each part-write.
2760 This seems wrong - if the receiver was accepting one byte at a time it
2761 would take for ever. The timeout is now adjusted when this happens. It
2762 doesn't have to be particularly precise.
2764 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2765 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2767 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2768 without a domain if the check to <postmaster@domain> fails.
2770 SC/01 Eximstats: added -xls and the ability to specify output files
2771 (patch written by Frank Heydlauf).
2773 SC/02 Eximstats: use FileHandles for outputing results.
2775 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2777 SC/04 Eximstats: fixed display of large numbers with -nvr option
2779 SC/05 Eximstats: fixed merging of reports with empty tables.
2781 SC/06 Eximstats: added the -include_original_destination flag
2783 SC/07 Eximstats: removed tabs and trailing whitespace.
2785 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2787 TK/06 MBOX spool code: Add real "From " MBOX separator line
2788 so the .eml file is really in mbox format (even though
2789 most programs do not really care). Patch from Alex Miller.
2791 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2792 The latter is generated from $received_to and is only set if the
2793 message has one envelope recipient. SA can use these headers,
2794 obviously out-of-the-box. Patch from Alex Miller.
2796 PH/08 The ${def test on a variable was returning false if the variable's
2797 value was "0", contrary to what the specification has always said!
2798 The result should be true unless the variable is empty.
2800 PH/09 The syntax error of a character other than { following "${if
2801 def:variable_name" (after optional whitespace) was not being diagnosed.
2802 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2803 accidental colon was present, for example, could give incorrect results.
2805 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2806 result is used (not including appendfile, where other changes are about
2809 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2810 This involved changing a lot of size variables from int to off_t. It
2811 should work with maildirs and everything.
2813 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2814 spamd dying while we are connected to it.
2816 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2819 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2820 function for BDB 4.3.
2822 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2824 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2825 the sources for PCRE have been re-arranged and more files are now
2828 PH/15 The code I had for printing potentially long long variables in PH/11
2829 above was not the best (it lost precision). The length of off_t variables
2830 is now inspected at build time, and an appropriate printing format (%ld
2831 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2832 to be "long long int" or "long int". This is needed for the internal
2833 formatting function string_vformat().
2835 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2836 the configuration file to be ":syslog", then the script "guesses" where
2837 the logs files are, rather than using the compiled in default. In our
2838 case the guess is not the same as the compiled default, so the script
2839 suddenly stopped working when I started to use syslog. The patch checks
2840 to see if log_file_path is "". If so, it attempts to read it from exim
2841 with no configuration file to get the compiled in version, before it
2842 falls back to the previous guessing code."
2844 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2845 implementing BATV in an Exim configuration. See NewStuff for the gory
2848 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2849 Makefile that are specific to HP-UX.
2851 PH/18 If the "use_postmaster" option was set for a recipient callout together
2852 with the "random" option, the postmaster address was used as the MAIL
2853 FROM address for the random test, but not for the subsequent recipient
2854 test. It is now used for both.
2856 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2857 patch removes a few documentation additions to RFC 3028, because the
2858 latest draft now contains them. It adds the new en;ascii-case comparator
2859 and a new error check for 8bit text in MIME parts. Comparator and
2860 require names are now matched exactly. I enabled the subaddress
2861 extension, but it is not well tested yet (read: it works for me)."
2863 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2864 rework some of the code of TK/09 above to avoid the hardwired use of
2865 "%lld" and "long long". Replaced the call to snprintf() with a call to
2868 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2869 records point to non-existent hosts", "retry timeout exceeded", and
2870 "retry time not reached for any host after a long failure period".
2872 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2873 experimental DomainKeys support:
2875 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2876 (2) On an error such as an illegally used "control", the wrong name for
2877 the control was given.
2879 These problems did NOT occur unless DomainKeys support was compiled.
2881 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2883 PH/24 Added ${if match_ip condition.
2885 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2886 databases so that it will be absolutely obvious if a crash occurs in the
2887 DB library. This is a regular occurrence (often caused by mis-matched
2890 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2891 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2892 picked up on a user's system that detects such things. There doesn't seem
2893 to be a gcc warning option for this - only an attribute that has to be
2894 put on the function's prototype. It seems that in Fedora Core 4 they have
2895 set this on a number of new functions. No doubt there will be more in due
2898 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2899 specify a transport. However, if an address that was verified by such a
2900 router was the subject of a callout, Exim crashed because it tried to
2901 read the rcpt_include_affixes from the non-existent transport. Now it
2902 just assumes that the setting of that option is false. This bug was
2903 introduced by 4.51/PH/31.
2905 PH/28 Changed -d+all to exclude +memory, because that information is very
2906 rarely of interest, but it makes the output a lot bigger. People tend to
2907 do -d+all out of habit.
2909 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2910 code in os-type was giving problems when libc.so lives in lib64, like on
2913 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2914 aren't the modern standard, and it seems that some systems' include files
2915 don't always have them. Exim was already checking for some of the newer
2916 ones like T_AAAA, and defining it itself. I've added checks for all the
2917 record types that Exim uses.
2919 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2920 not automatically generating a new one, as it is supposed to. This
2921 prevented TLS from working. If the file did exist, but contained invalid
2922 data, a new version was generated, as expected. It was only the case of a
2923 non-existent file that was broken.
2925 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2926 with a change in libdomainkeys > 0.64.
2928 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2929 from DNS. If the selector record carries the flag, it now has
2930 precedence over the domain-wide flag.
2932 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2934 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2935 the use of a port name that isn't defined in /etc/services) occurred, the
2936 message was deferred as in a normal delivery, and thus remained on the
2937 spool, instead of being failed because of the mua_wrapper setting. This
2938 is now fixed, and I tidied up some of the mua_wrapper messages at the
2941 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2942 the messages in a hash of arrays rather than using individual hashes.
2943 This is a bit cleaner and results in dramatic memory savings, albeit
2944 at a slight CPU cost.
2946 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2947 as requested by Marc Sherman.
2949 SC/10 Eximstats: added histograms for user specified patterns as requested
2952 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2954 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2955 fopen() in the content-scanning modules that did not already have it.
2961 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2962 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2964 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2966 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2968 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2969 to test Sieve filters that use "vacation".
2971 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2972 that changes the way the GnuTLS parameters are stored in the cache file.
2973 The new format can be generated externally. For backward compatibility,
2974 if the data in the cache doesn't make sense, Exim assumes it has read an
2975 old-format file, and it generates new data and writes a new file. This
2976 means that you can't go back to an older release without removing the
2979 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2980 work if there are any delivery delays because "one_time" forces the
2981 parent to be marked "delivered", so its unseen clone is never tried
2982 again. For this reason, Exim now forbids the simultaneous setting of
2985 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2986 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2987 entry if you want to know the details. The fix, however, neglected to
2988 consider the case when local delivery batching is involved. The test for
2989 "previously delivered" was not happening when checking to see if an
2990 address could be batched with a previous (undelivered) one; under
2991 certain circumstances this could lead to multiple deliveries to the same
2994 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2995 in its include files, and this causes problems building Exim.
2997 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2998 header_syntax) but Exim was just ignoring anything given after a slash.
2999 In particular, this caused confusion with an attempt to use "verify =
3000 reverse_host_lookup/defer_ok". An error is now given when options are
3001 supplied for verify items that do not have them. (Maybe reverse_host_
3002 lookup should have a defer_ok option, but that's a different point.)
3004 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3005 defined by RFC 821) to 2048, because there were problems with some AUTH
3006 commands, and RFC 1869 says the size should be increased for extended
3007 SMTP commands that take arguments.
3009 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3012 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3013 "unknown" error; now it says that the functionality isn't in the binary.
3015 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3016 an address' error message when a string expansion fails (syntax or
3017 whatever). Otherwise the password may appear in the log. Following change
3018 PH/42 below, there is no longer a chance of it appearing in a bounce
3021 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3023 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3024 of its addresses was ever tried. (Bugzilla bug #2.)
3026 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3027 the result incorrectly in the debug output. (It correctly added a newline
3028 to what was transported.)
3030 TF/01 Added $received_time.
3032 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3033 commented out examples of how to interface to a virus scanner and to
3034 SpamAssassin. Also added commented examples of av_scanner and
3035 spamd_address settings.
3037 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3038 and controls are allowed in which ACLs. There were a couple of minor
3039 errors. Some of the entries in the conditions table (which is a table of
3040 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3041 negation of where the condition IS allowed.
3043 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3045 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3046 header file does not have a version number, so I've had to invent a new
3047 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3048 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3049 radiusclient), but it was contributed by a Radius user.
3051 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3052 files or directories, for OpenSSL.
3054 PH/19 When an Exim process that is running as root has to create an Exim log
3055 file, it does so in a subprocess that runs as exim:exim so as to get the
3056 ownership right at creation (otherwise, other Exim processes might see
3057 the file with the wrong ownership). There was no test for failure of this
3058 fork() call, which would lead to the process getting stuck as it waited
3059 for a non-existent subprocess. Forks do occasionally fail when resources
3060 run out. I reviewed all the other calls to fork(); they all seem to check
3063 PH/20 When checking for unexpected SMTP input at connect time (before writing
3064 the banner), Exim was not dealing correctly with a non-positive return
3065 from the read() function. If the client had disconnected by this time,
3066 the result was a log entry for a synchronization error with an empty
3067 string after "input=" when read() returned zero. If read() returned -1
3068 (an event I could not check), uninitialized data bytes were printed.
3069 There were reports of junk text (parts of files, etc) appearing after
3072 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3074 PH/22 Added support for macro redefinition, and (re)definition in between
3075 driver and ACL definitions.
3077 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3078 forgetting to use the resulting value; it was using the unexpanded value.
3080 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3081 hadn't been configured. The fix is from Juergen Kreileder, who
3082 understands it better than I do:
3084 "Here's what I see happening with three configured cyrus_sasl
3085 authenticators configured (plain, login, cram-md5):
3087 On startup auth_cyrus_sasl_init() gets called for each of these.
3088 This means three calls to sasl_listmech() without a specified mech_list.
3089 => SASL tests which mechs of all available mechs actually work
3090 => three warnings about OTP not working
3091 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3093 With the patch, sasl_listmech() also gets called three times. But now
3094 SASL's mech_list option is set to the server_mech specified in the the
3095 authenticator. Or in other words, the answer from sasl_listmech()
3096 gets limited to just the mech you're testing for (which is different
3098 => the return list contains just 'plain' or 'login', 'cram-md5' or
3099 nothing depending on the value of ob->server_mech.
3101 I've just tested the patch: Authentication still works fine,
3102 unavailable mechs specified in the exim configuration are still
3103 caught, and the auth.log warnings about OTP are gone."
3105 PH/25 When debugging is enabled, the contents of the command line are added
3106 to the debugging output, even when log_selector=+arguments is not
3109 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3110 answer is "GNU", and only if the return is "GNU/something" is the answer
3113 PH/27 $acl_verify_message is now set immediately after the failure of a
3114 verification in an ACL, and so is available in subsequent modifiers. In
3115 particular, the message can be preserved by coding like this:
3117 warn !verify = sender
3118 set acl_m0 = $acl_verify_message
3120 Previously, $acl_verify_message was set only while expanding "message"
3121 and "log_message" when a very denied access.
3123 PH/28 Modified OS/os.c-Linux with
3125 -#ifndef OS_LOAD_AVERAGE
3126 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3128 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3129 nomenclature these days.)
3131 PH/29 Installed patch from the Sieve maintainer that adds the options
3132 sieve_useraddress and sieve_subaddress to the redirect router.
3134 PH/30 In these circumstances:
3135 . Two addresses routed to the same list of hosts;
3136 . First host does not offer TLS;
3137 . First host accepts first address;
3138 . First host gives temporary error to second address;
3139 . Second host offers TLS and a TLS session is established;
3140 . Second host accepts second address.
3141 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3142 and peerdn, if requested) that were in fact used only for the second
3145 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3146 attention to any local part prefix or suffix that was matched by the
3147 router that accepted the address. It now behaves in the same way as it
3148 does for delivery: the affixes are removed from the local part unless
3149 rcpt_include_affixes is set on the transport.
3151 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3152 timeout during the DATA phase of an incoming message.
3154 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3155 applied a patch sanctioned by the Sieve maintainer.
3157 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3158 the uid or gid is negative. A case of a negative gid caused this to be
3159 noticed. The fix allows for either to be negative.
3161 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3162 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3165 PH/36 The change PH/12 above was broken. Fixed it.
3167 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3168 the grounds that routing the same address twice would always produce the
3169 same answer. This might have been true once, but it is certainly no
3170 longer true now. Routing a child address may depend on the previous
3171 routing that produced that child. Some complicated redirection strategies
3172 went wrong when messages had multiple recipients, and made Exim's
3173 behaviour dependent on the order in which the addresses were given.
3175 I have moved the duplicate checking until after the routing is complete.
3176 Exim scans the addresses that are assigned to local and remote
3177 transports, and removes any duplicates. This means that more work will be
3178 done, as duplicates will always all be routed, but duplicates are
3179 presumably rare, so I don't expect this is of any significance.
3181 For deliveries to pipes, files, and autoreplies, the duplicate checking
3182 still happens during the routing process, since they are not going to be
3185 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3186 It corrects a timeout issue with spamd. This is Ian's comment: "The
3187 background is that sometimes spamd either never reads data from a
3188 connection it has accepted, or it never writes response data. The exiscan
3189 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3190 blindly assumes that writes won't block so it may never time out."
3192 PH/39 Allow G after quota size as well as K and M.
3194 PH/40 The value set for $authenticated_id in an authenticator may not contain
3195 binary zeroes or newlines because the value is written to log lines and
3196 to spool files. There was no check on this. Now the value is run through
3197 the string_printing() function so that such characters are converted to
3198 printable escape sequences.
3200 PH/41 $message_linecount is a new variable that contains the total number of
3201 lines in the message. Compare $body_linecount, which is the count for the
3204 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3205 bounce and delay warning messages, except in certain special cases, which
3208 (a) An SMTP error message from a remote host;
3209 (b) A message specified in a :fail: redirection;
3210 (c) A message specified in a "fail" command in a system filter;
3211 (d) A message specified in a FAIL return from the queryprogram router;
3212 (e) A message specified by the cannot_route_message router option.
3214 In these cases only, Exim does include the error details in bounce and
3215 warning messages. There are also a few cases where bland messages such
3216 as "unrouteable address" or "local delivery error" are given.
3218 PH/43 $value is now also set for the "else" part of a ${run expansion.
3220 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3221 being worked on, but at least Exim now implements the latest version to
3224 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3225 process to complete was treated as a delivery failure, a timeout while
3226 writing the message to the pipe was logged, but erroneously treated as a
3227 successful delivery. Such timeouts include transport filter timeouts. For
3228 consistency with the overall process timeout, these timeouts are now
3229 treated as errors, giving rise to delivery failures by default. However,
3230 there is now a new Boolean option for the pipe transport called
3231 timeout_defer, which, if set TRUE, converts the failures into defers for
3232 both kinds of timeout. A transport filter timeout is now identified in
3235 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3236 systems where "make" and "gmake" are different, calling "gmake" at top
3237 level broke things. I've arranged for the value of $(MAKE) to be passed
3238 from the Makefile to this script so that it can call the same version of
3242 A note about Exim versions 4.44 and 4.50
3243 ----------------------------------------
3245 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3246 changes of various kinds. As a consequence, a big documentation update was
3247 needed. This delayed the release for rather longer than seemed good, especially
3248 in the light of a couple of (minor) security issues. Therefore, the changes
3249 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3250 release. So 4.44 and 4.50 are in effect two different branches that both start
3253 I have left the 4.50 change log unchanged; it contains all the changes since
3254 4.43. The change log for 4.44 is below; many of its items are identical to
3255 those for 4.50. This seems to be the most sensible way to preserve the
3256 historical information.
3262 1. Minor wording change to the doc/README.SIEVE file.
3264 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3265 computation of the current number of files was incorrect.
3267 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3268 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3271 4. Give more explanation in the error message when the command for a transport
3272 filter fails to execute.
3274 5. There are several places where Exim runs a non-Exim command in a
3275 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3276 was being done only for the command run by the queryprogram router. It is
3277 now done for all such subprocesses. The other cases are: ${run, transport
3278 filters, and the commands run by the lmtp and pipe transports.
3280 6. Added CONFIGURE_GROUP build-time option.
3282 7. Some older OS have a limit of 256 on the maximum number of file
3283 descriptors. Exim was using setrlimit() to set 1000 as a large value
3284 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3285 systems. I've change it so that if it can't get 1000, it tries for 256.
3287 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3288 was an oversight, and furthermore, ever since the addition of extra
3289 controls (e.g. 4.43/32), the checks on when to allow different forms of
3290 "control" were broken. There should now be diagnostics for all cases when a
3291 control that does not make sense is encountered.
3293 9. Added the /retain_sender option to "control=submission".
3295 10. $recipients is now available in the predata ACL (oversight).
3297 11. Tidy the search cache before the fork to do a delivery from a message
3298 received from the command line. Otherwise the child will trigger a lookup
3299 failure and thereby defer the delivery if it tries to use (for example) a
3300 cached ldap connection that the parent has called unbind on.
3302 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3303 of $address_data from the recipient verification was clobbered by the
3304 sender verification.
3306 13. The value of address_data from a sender verification is now available in
3307 $sender_address_data in subsequent conditions in the ACL statement.
3309 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3311 15. Added a new option "connect=<time>" to callout options, to set a different
3314 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3315 was its contents. (It was OK if the option was not defined at all.)
3317 17. A "Completed" log line is now written for messages that are removed from
3318 the spool by the -Mrm option.
3320 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3321 information about exactly what failed.
3323 19. Added -dd to debug only the daemon process.
3325 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3326 handles timeouts, both on the server side and network timeouts. Renamed the
3327 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3329 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3330 It is now set to "smtps".
3332 22. $host_address is now set to the target address during the checking of
3333 ignore_target_hosts.
3335 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3336 being passed; this would have caused $sender_host_name to have been used if
3337 matching the list had actually called for a host name (not very likely,
3338 since this list is usually IP addresses). A host name is now passed as
3341 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3342 code that specifies a non-restarting handler (typically sigaction() in
3343 modern systems) in an attempt to fix a rare and obscure crash bug.
3345 25. Narrowed the window for a race in the daemon that could cause it to ignore
3346 SIGCHLD signals. This is not a major problem, because they are used only to
3347 wake it up if nothing else does.
3349 26. A malformed maildirsize file could cause Exim to calculate negative values
3350 for the mailbox size or file count. Odd effects could occur as a result.
3351 The maildirsize information is now recalculated if the size or filecount
3354 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3355 support for a long time. Removed HAVE_SYS_VFS_H.
3357 28. Installed the latest version of exipick from John Jetmore.
3359 29. In an address list, if the pattern was not a regular expression, an empty
3360 subject address (from a bounce message) matched only if the pattern was an
3361 empty string. Non-empty patterns were not even tested. This was the wrong
3362 because it is perfectly reasonable to use an empty address as part of a
3363 database query. An empty address is now tested by patterns that are
3364 lookups. However, all the other forms of pattern expect the subject to
3365 contain a local part and a domain, and therefore, for them, an empty
3366 address still always fails if the pattern is not itself empty.
3368 30. Exim went into a mad DNS loop when attempting to do a callout where the
3369 host was specified on an smtp transport, and looking it up yielded more
3370 than one IP address.
3372 31. Re-factored the code for checking spool and log partition space into a
3373 function that finds that data and another that does the check. The former
3374 is then used to implement four new variables: $spool_space, $log_space,
3375 $spool_inodes, and $log_inodes.
3377 32. The RFC2047 encoding function was originally intended for short strings
3378 such as real names; it was not keeping to the 75-character limit for
3379 encoded words that the RFC imposes. It now respects the limit, and
3380 generates multiple encoded words if necessary. To be on the safe side, I
3381 have increased the buffer size for the ${rfc2047: expansion operator from
3384 33. It is now permitted to omit both strings after an "if" condition; if the
3385 condition is true, the result is "true". As before, when the second string
3386 is omitted, a false condition yields an empty string. This makes it less
3387 cumbersome to write custom ACL and router conditions.
3389 34. Failure to deliver a bounce message always caused it to be frozen, even if
3390 there was an errors_to setting on the router. The errors_to setting is now
3393 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3394 canonical form (fully expanded) before being placed in
3395 $sender_host_address.
3397 36. The table in the code that translates DNS record types into text (T_A to
3398 "A" for instance) was missing entries for NS and CNAME. It is just possible
3399 that this could have caused confusion if both these types were looked up
3400 for the same domain, because the text type is used as part of Exim's
3401 per-process caching. But the chance of anyone hitting this buglet seems
3404 37. The dnsdb lookup has been extended in a number of ways.
3406 (1) There is a new type, "zns", which walks up the domain tree until it
3407 finds some nameserver records. It should be used with care.
3409 (2) There is a new type, "mxh", which is like "mx" except that it returns
3410 just the host names, not the priorities.
3412 (3) It is now possible to give a list of domains (or IP addresses) to be
3413 looked up. The behaviour when one of the lookups defers can be
3414 controlled by a keyword.
3416 (4) It is now possible to specify the separator character for use when
3417 multiple records are returned.
3419 38. The dnslists ACL condition has been extended: it is now possible to supply
3420 a list of IP addresses and/or domains to be looked up in a particular DNS
3423 39. Added log_selector=+queue_time_overall.
3425 40. When running the queue in the test harness, wait just a tad after forking a
3426 delivery process, to get repeatability of debugging output.
3428 41. Include certificate and key file names in error message when GnuTLS fails
3429 to set them up, because the GnuTLS error message doesn't include the name
3430 of the failing file when there is a problem reading it.
3432 42. Allow both -bf and -bF in the same test run.
3434 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3436 44. The "Exiscan patch" is now merged into the mainline Exim source.
3438 45. Sometimes the final signoff response after QUIT could fail to get
3439 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3440 < 0 before doing a fflush(). This bug looks as though it goes back to the
3441 introduction of TLS in release 3.20, but "sometimes" must have been rare
3442 because the tests only now provoked it.
3444 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3445 (this can affect the format of dates).
3447 47. exim_tidydb, when checking for the continued existence of a message for
3448 which it has found a message-specific retry record, was not finding
3449 messages that were in split spool directories. Consequently, it was
3450 deleting retry records that should have stayed in existence.
3452 48. Steve fixed some bugs in eximstats.
3454 49. The SPA authentication driver was not abandoning authentication and moving
3455 on to the next authenticator when an expansion was forced to fail,
3456 contradicting the general specification for all authenticators. Instead it
3457 was generating a temporary error. It now behaves as specified.
3459 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3460 (the order specifies the preference for clients). The order is now AES256,
3461 AES128, 3DES, ARCFOUR128.
3463 51. Small patch to Sieve code - explicitly set From: when generating an
3466 52. Exim crashed if a remote delivery caused a very long error message to be
3467 recorded - for instance if somebody sent an entire SpamAssassin report back
3468 as a large number of 550 error lines. This bug was coincidentally fixed by
3469 increasing the size of one of Exim's internal buffers (big_buffer) that
3470 happened as part of the Exiscan merge. However, to be on the safe side, I
3471 have made the code more robust (and fixed the comments that describe what
3474 53. Now that there can be additional text after "Completed" in log lines (if
3475 the queue_time_overall log selector is set), a one-byte patch to exigrep
3476 was needed to allow it to recognize "Completed" as not the last thing in
3479 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3480 patch that reportedly fixes this has been added. I am not expert enough to
3481 create a test for it. This is what the patch creator wrote:
3483 "I found a little strange behaviour of ldap code when working with
3484 Windows 2003 AD Domain, where users was placed in more than one
3485 Organization Units. When I tried to give exim partial DN, the exit code
3486 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3487 But simultaneously result of request was absolutely normal ldap result,
3488 so I produce this patch..."
3490 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3491 REFERENCE, so I have modified the code to exclude the patch when that macro
3494 55. Some experimental protocols are using DNS PTR records for new purposes. The
3495 keys for these records are domain names, not reversed IP addresses. The
3496 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3497 leaves it alone. Component reversal etc. now happens only for IP addresses.
3500 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3502 57. Double the size of the debug message buffer (to 2048) so that more of very
3503 long debug lines gets shown.
3505 58. The exicyclog utility now does better if the number of log files to keep
3506 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3508 59. Two changes related to the smtp_active_hostname option:
3510 (1) $smtp_active_hostname is now available as a variable.
3511 (2) The default for smtp_banner uses $smtp_active_hostname instead
3512 of $primary_hostname.
3514 60. The host_aton() function is supposed to be passed a string that is known
3515 to be a valid IP address. However, in the case of IPv6 addresses, it was
3516 not checking this. This is a hostage to fortune. Exim now panics and dies
3517 if the condition is not met. A case was found where this could be provoked
3518 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3519 components; fortuitously, this particular loophole had already been fixed
3520 by change 4.50/55 above.
3522 If there are any other similar loopholes, the new check in host_aton()
3523 itself should stop them being exploited. The report I received stated that
3524 data on the command line could provoke the exploit when Exim was running as
3525 exim, but did not say which command line option was involved. All I could
3526 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3527 running as the user.
3530 61. There was a buffer overflow vulnerability in the SPA authentication code
3531 (which came originally from the Samba project). I have added a test to the
3532 spa_base64_to_bits() function which I hope fixes it.
3535 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3536 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3538 63. The daemon start-up calls getloadavg() while still root for those OS that
3539 need the first call to be done as root, but it missed one case: when
3540 deliver_queue_load_max is set with deliver_drop_privilege. This is
3541 necessary for the benefit of the queue runner, because there is no re-exec
3542 when deliver_drop_privilege is set.
3544 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3545 This has been fixed.
3547 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3548 was in use, was not putting the data itself into the right store pool;
3549 consequently, it could be overwritten for a subsequent message in the same
3550 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3553 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3555 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3556 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3557 the function were treating the return as a boolean value, which happened to
3558 work because 0=false and not-0=true, but is not correct code.
3560 68. The host_aton() function was not handling scoped IPv6 addresses (those
3561 with, for example, "%eth0" on the end) correctly.
3563 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3564 negated items (that is, ~something) in unsigned ints. Some compilers
3565 apparently mutter when there is no cast.
3567 70. If an address verification called from an ACL failed, and did not produce a
3568 user-specific message (i.e. there was only a "system" message), nothing was
3569 put in $acl_verify_message. In this situation, it now puts the system
3572 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3573 session; change 4.31/43 added the unwanted input to the log line - except
3574 that it did not do this in the start of session case. It now does.
3576 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3577 This is wrong and can cause the other end to generate a synchronization
3578 error if it is another Exim or anything else that does the synchronization
3579 check. A QUIT command is no longer sent after a timeout.
3581 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3582 during host lookups.
3584 74. The defer_ok option of callout verification was not working if it was used
3585 when verifying addresses in header lines, that is, for this case:
3587 verify = header_sender/callout=defer_ok
3589 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3590 those file descriptors could be used for SMTP connections. If anything
3591 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3592 could be sent to the SMTP client, causing chaos. The daemon now opens
3593 stdin, stdout, and stderr to /dev/null when it puts itself into the
3596 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3597 log by default. The user can override this with suitable Perl magic.
3599 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3600 the log message when discard triggers, was not working for the DATA ACL or
3601 for the non-SMTP ACL.
3603 78. Error message wording change in sieve.c.
3605 79. If smtp_accept_max_per_host was set, the number of connections could be
3606 restricted to fewer than expected, because the daemon was trying to set up
3607 a new connection before checking whether the processes handling previous
3608 connections had finished. The check for completed processes is now done
3609 earlier. On busy systems, this bug wouldn't be noticed because something
3610 else would have woken the daemon, and it would have reaped the completed
3613 80. If a message was submitted locally by a user whose login name contained one
3614 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3615 It caused a spool format error. I have fixed the spool reading code. A
3616 related problem was that the "from" clause in the Received: line became
3617 illegal because of the space(s). It is now covered by ${quote_local_part.
3619 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3622 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3624 83. Patch from Sieve maintainer to fix unterminated string problem in
3625 "vacation" handling.
3627 84. Some minor changes to the Linux configuration files to help with other
3628 OS variants using glibc.
3630 85. One more patch for Sieve to update vacation handling to latest spec.
3633 ----------------------------------------------------
3634 See the note above about the 4.44 and 4.50 releases.
3635 ----------------------------------------------------
3641 1. Change 4.43/35 introduced a bug that caused file counts to be
3642 incorrectly computed when quota_filecount was set in an appendfile
3645 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3646 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3649 3. Give more explanation in the error message when the command for a transport
3650 filter fails to execute.
3652 4. There are several places where Exim runs a non-Exim command in a
3653 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3654 was being done only for the command run by the queryprogram router. It is
3655 now done for all such subprocesses. The other cases are: ${run, transport
3656 filters, and the commands run by the lmtp and pipe transports.
3658 5. Some older OS have a limit of 256 on the maximum number of file
3659 descriptors. Exim was using setrlimit() to set 1000 as a large value
3660 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3661 systems. I've change it so that if it can't get 1000, it tries for 256.
3663 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3664 was an oversight, and furthermore, ever since the addition of extra
3665 controls (e.g. 4.43/32), the checks on when to allow different forms of
3666 "control" were broken. There should now be diagnostics for all cases when a
3667 control that does not make sense is encountered.
3669 7. $recipients is now available in the predata ACL (oversight).
3671 8. Tidy the search cache before the fork to do a delivery from a message
3672 received from the command line. Otherwise the child will trigger a lookup
3673 failure and thereby defer the delivery if it tries to use (for example) a
3674 cached ldap connection that the parent has called unbind on.
3676 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3677 of $address_data from the recipient verification was clobbered by the
3678 sender verification.
3680 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3681 was its contents. (It was OK if the option was not defined at all.)
3683 11. A "Completed" log line is now written for messages that are removed from
3684 the spool by the -Mrm option.
3686 12. $host_address is now set to the target address during the checking of
3687 ignore_target_hosts.
3689 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3690 being passed; this would have caused $sender_host_name to have been used if
3691 matching the list had actually called for a host name (not very likely,
3692 since this list is usually IP addresses). A host name is now passed as
3695 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3696 code that specifies a non-restarting handler (typically sigaction() in
3697 modern systems) in an attempt to fix a rare and obscure crash bug.
3699 15. Narrowed the window for a race in the daemon that could cause it to ignore
3700 SIGCHLD signals. This is not a major problem, because they are used only to
3701 wake it up if nothing else does.
3703 16. A malformed maildirsize file could cause Exim to calculate negative values
3704 for the mailbox size or file count. Odd effects could occur as a result.
3705 The maildirsize information is now recalculated if the size or filecount
3708 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3709 support for a long time. Removed HAVE_SYS_VFS_H.
3711 18. Updated exipick to current release from John Jetmore.
3713 19. Allow an empty sender to be matched against a lookup in an address list.
3714 Previously the only cases considered were a regular expression, or an
3717 20. Exim went into a mad DNS lookup loop when doing a callout where the
3718 host was specified on the transport, if the DNS lookup yielded more than
3721 21. The RFC2047 encoding function was originally intended for short strings
3722 such as real names; it was not keeping to the 75-character limit for
3723 encoded words that the RFC imposes. It now respects the limit, and
3724 generates multiple encoded words if necessary. To be on the safe side, I
3725 have increased the buffer size for the ${rfc2047: expansion operator from
3728 22. Failure to deliver a bounce message always caused it to be frozen, even if
3729 there was an errors_to setting on the router. The errors_to setting is now
3732 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3733 canonical form (fully expanded) before being placed in
3734 $sender_host_address.
3736 24. Updated eximstats to version 1.33
3738 25. Include certificate and key file names in error message when GnuTLS fails
3739 to set them up, because the GnuTLS error message doesn't include the name
3740 of the failing file when there is a problem reading it.
3742 26. Expand error message when OpenSSL has problems setting up cert/key files.
3745 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3746 (this can affect the format of dates).
3748 28. exim_tidydb, when checking for the continued existence of a message for
3749 which it has found a message-specific retry record, was not finding
3750 messages that were in split spool directories. Consequently, it was
3751 deleting retry records that should have stayed in existence.
3753 29. eximstats updated to version 1.35
3754 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3755 1.35 - bugfix such that pie charts by volume are generated correctly
3757 30. The SPA authentication driver was not abandoning authentication and moving
3758 on to the next authenticator when an expansion was forced to fail,
3759 contradicting the general specification for all authenticators. Instead it
3760 was generating a temporary error. It now behaves as specified.
3762 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3763 (the order specifies the preference for clients). The order is now AES256,
3764 AES128, 3DES, ARCFOUR128.
3766 31. Small patch to Sieve code - explicitly set From: when generating an
3769 32. Exim crashed if a remote delivery caused a very long error message to be
3770 recorded - for instance if somebody sent an entire SpamAssassin report back
3771 as a large number of 550 error lines. This bug was coincidentally fixed by
3772 increasing the size of one of Exim's internal buffers (big_buffer) that
3773 happened as part of the Exiscan merge. However, to be on the safe side, I
3774 have made the code more robust (and fixed the comments that describe what
3777 33. Some experimental protocols are using DNS PTR records for new purposes. The
3778 keys for these records are domain names, not reversed IP addresses. The
3779 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3780 leaves it alone. Component reversal etc. now happens only for IP addresses.
3783 34. The host_aton() function is supposed to be passed a string that is known
3784 to be a valid IP address. However, in the case of IPv6 addresses, it was
3785 not checking this. This is a hostage to fortune. Exim now panics and dies
3786 if the condition is not met. A case was found where this could be provoked
3787 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3788 components; fortuitously, this particular loophole had already been fixed
3789 by change 4.50/55 or 4.44/33 above.
3791 If there are any other similar loopholes, the new check in host_aton()
3792 itself should stop them being exploited. The report I received stated that
3793 data on the command line could provoke the exploit when Exim was running as
3794 exim, but did not say which command line option was involved. All I could
3795 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3796 running as the user.
3799 35. There was a buffer overflow vulnerability in the SPA authentication code
3800 (which came originally from the Samba project). I have added a test to the
3801 spa_base64_to_bits() function which I hope fixes it.
3804 36. The daemon start-up calls getloadavg() while still root for those OS that
3805 need the first call to be done as root, but it missed one case: when
3806 deliver_queue_load_max is set with deliver_drop_privilege. This is
3807 necessary for the benefit of the queue runner, because there is no re-exec
3808 when deliver_drop_privilege is set.
3810 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3811 was in use, was not putting the data itself into the right store pool;
3812 consequently, it could be overwritten for a subsequent message in the same
3813 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3816 38. Sometimes the final signoff response after QUIT could fail to get
3817 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3818 < 0 before doing a fflush(). This bug looks as though it goes back to the
3819 introduction of TLS in release 3.20, but "sometimes" must have been rare
3820 because the tests only now provoked it.
3826 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3827 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3828 Now it returns an integer. A number of calls were still expecting a T/F
3829 return. Fortuitously, in all cases, the tests worked in OK situations,
3830 which is the norm. However, things would have gone wrong on any write
3831 failures on the smtp file descriptor. This function is used when sending
3832 messages over SMTP and also when doing verify callouts.
3834 2. When Exim is called to do synchronous delivery of a locally submitted
3835 message (the -odf or -odi options), it no longer closes stderr before doing
3838 3. Implemented the mua_wrapper option.
3840 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3842 5. Implemented the functions header_remove(), header_testname(),
3843 header_add_at_position(), and receive_remove_recipient(), and exported them
3846 6. If an ACL "warn" statement specified the addition of headers, Exim already
3847 inserted X-ACL-Warn: at the start if there was no header name. However, it
3848 was not making this test for the second and subsequent header lines if
3849 there were newlines in the string. This meant that an invalid header could
3850 be inserted if Exim was badly configured.
3852 7. Allow an ACL "warn" statement to add header lines at the start or after all
3853 the Received: headers, as well as at the end.
3855 8. Added the rcpt_4xx retry error code.
3857 9. Added postmaster_mailfrom=xxx to callout verification option.
3859 10. Added mailfrom=xxxx to the callout verification option, for verify=
3862 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3863 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3865 12. Inserted some casts to stop certain compilers warning when using pointer
3866 differences as field lengths or precisions in printf-type calls (mostly
3867 affecting debugging statements).
3869 13. Added optional readline() support for -be (dynamically loaded).
3871 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3872 same clock tick as a message's arrival, so that its received time was the
3873 same as the "first fail" time on the retry record, and that message
3874 remained on the queue past the ultimate address timeout, every queue runner
3875 would try a delivery (because it was past the ultimate address timeout) but
3876 after another failure, the ultimate address timeout, which should have then
3877 bounced the address, did not kick in. This was a "< instead of <=" error;
3878 in most cases the first failure would have been in the next clock tick
3879 after the received time, and all would be well.
3881 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3882 being recognized when the domain list was tested by the match_domain
3883 condition in an expansion string.
3885 16. Added the ${str2b64: operator.
3887 17. Exim was always calling setrlimit() to set a large limit for the number of
3888 processes, without checking whether the existing limit was already
3889 adequate. (It did check for the limit on file descriptors.) Furthermore,
3890 errors from getrlimit() and setrlimit() were being ignored. Now they are
3891 logged to the main and panic logs, but Exim does carry on, to try to do its
3892 job under whatever limits there are.
3894 18. Imported PCRE 5.0.
3896 19. Trivial typo in log message " temporarily refused connection" (the leading
3899 20. If the log selector return_path_on_delivery was set and an address was
3900 redirected to /dev/null, the delivery process crashed because it assumed
3901 that a return path would always be set for a "successful" delivery. In this
3902 case, the whole delivery is bypassed as an optimization, and therefore no
3905 21. Internal re-arrangement: the function for sending a challenge and reading
3906 a response while authentication was assuming a zero-terminated challenge
3907 string. It's now changed to take a pointer and a length, to allow for
3908 binary data in such strings.
3910 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3912 23. Exim was not respecting finduser_retries when seeking the login of the
3913 uid under which it was called; it was always trying 10 times. (The default
3914 setting of finduser_retries is zero.) Also, it was sleeping after the final
3915 failure, which is pointless.
3917 24. Implemented tls_on_connect_ports.
3919 25. Implemented acl_smtp_predata.
3921 26. If the domain in control=submission is set empty, Exim assumes that the
3922 authenticated id is a complete email address when it generates From: or
3923 Sender: header lines.
3925 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3926 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3927 chown and chgrp in /bin and hostname in /usr/bin.
3929 28. Exim was keeping the "process log" file open after each use, just as it
3930 does for the main log. This opens the possibility of it remaining open for
3931 long periods when the USR1 signal hits a daemon. Occasional processlog
3932 errors were reported, that could have been caused by this. Anyway, it seems
3933 much more sensible not to leave this file open at all, so that is what now
3936 29. The long-running daemon process does not normally write to the log once it
3937 has entered its main loop, and it closes the log before doing so. This is
3938 so that log files can straightforwardly be renamed and moved. However,
3939 there are a couple of unusual error situations where the daemon does write
3940 log entries, and I had neglected to close the log afterwards.
3942 30. The text of an SMTP error response that was received during a remote
3943 delivery was being truncated at 512 bytes. This is too short for some of
3944 the long messages that one sometimes sees. I've increased the limit to
3947 31. It is now possible to make retry rules that apply only when a message has a
3948 specific sender, in particular, an empty sender.
3950 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3951 it possible to be selective about when SMTP synchronization is enforced.
3953 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3955 32. Implemented hosts_connection_nolog.
3957 33. Added an ACL for QUIT.
3959 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3962 35. Added mailbox_size and mailbox_filecount to appendfile.
3964 36. Added control = no_multiline_responses to ACLs.
3966 37. There was a bug in the logic of the code that waits for the clock to tick
3967 in the case where the clock went backwards by a substantial amount such
3968 that the microsecond fraction of "now" was more than the microsecond
3969 fraction of "then" (but the whole seconds number was less).
3971 38. Added support for the libradius Radius client library this is found on
3972 FreeBSD (previously only the radiusclient library was supported).
3978 1. When certain lookups returned multiple values in the form name=value, the
3979 quoting of the values was not always being done properly. Specifically:
3980 (a) If the value started with a double quote, but contained no whitespace,
3982 (b) If the value contained whitespace other than a space character (i.e.
3983 tabs or newlines or carriage returns) it was not quoted.
3984 This fix has been applied to the mysql and pgsql lookups by writing a
3985 separate quoting function and calling it from the lookup code. The fix
3986 should probably also be applied to nisplus, ibase and oracle lookups, but
3987 since I cannot test any of those, I have not disturbed their existing code.
3989 2. A hit in the callout cache for a specific address caused a log line with no
3990 reason for rejecting RCPT. Now it says "Previous (cached) callout
3991 verification failure".
3993 3. There was an off-by-one bug in the queryprogram router. An over-long
3994 return line was truncated at 256 instead of 255 characters, thereby
3995 overflowing its buffer with the terminating zero. As well as fixing this, I
3996 have increased the buffer size to 1024 (and made a note to document this).
3998 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3999 when Exim is waiting for an SMTP response from a remote server, Exim
4000 restarts its select() call on the socket, thereby resetting its timeout.
4001 This is not a problem when such interrupts are rare. Somebody set up a cron
4002 job to run exiwhat every 2 minutes, which is less than the normal select()
4003 timeout (5 or 10 minutes). This meant that the select() timeout never
4004 kicked in because it was always reset. I have fixed this by comparing the
4005 time when an interrupt arrives with the time at the start of the first call
4006 to select(). If more time than the timeout has elapsed, the interrupt is
4007 treated as a timeout.
4009 5. Some internal re-factoring in preparation for the addition of Sieve
4010 extensions (by MH). In particular, the "personal" test is moved to a
4011 separate function, and given an option for scanning Cc: and Bcc: (which is
4012 not set for Exim filters).
4014 6. When Exim created an email address using the login of the caller as the
4015 local part (e.g. when creating a From: or Sender: header line), it was not
4016 quoting the local part when it contained special characters such as @.
4018 7. Installed new OpenBSD configuration files.
4020 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4021 try to make them clearer.
4023 9. Callout options, other than the timeout value, were being ignored when
4024 verifying sender addresses in header lines. For example, when using
4026 verify = header_sender/callout=no_cache
4028 the cache was (incorrectly) being used.
4030 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4031 only the Cygwin environment.
4033 11. When return_path_on_delivery was set as a log selector, if different remote
4034 addresses in the same message used different return paths and parallel
4035 remote delivery occurred, the wrong values would sometimes be logged.
4036 (Whenever a remote delivery process finished, the return path value from
4037 the most recently started remote delivery process was logged.)
4039 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4040 lines when AUTH and/or TLS are in use. This is the "received protocol"
4041 field. Exim used to use "asmtp" for authenticated SMTP, without any
4042 indication (in the protocol name) for TLS use. Now it follows the RFC and
4043 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4044 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4045 names appear in log lines as well as in Received: header lines.
4047 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4048 extensions, and comparison tests, and to fix some bugs.
4050 14. Changes to the "personal" filter test:
4052 (1) The test was buggy in that it was just doing the equivalent of
4053 "contains" tests on header lines. For example, if a user's address was
4054 anne@some.where, the "personal" test would incorrectly be true for
4056 To: susanne@some.where
4058 This test is now done by extracting each address from the header in turn,
4059 and checking the entire address. Other tests that are part of "personal"
4060 are now done using regular expressions (for example, to check local parts
4061 of addresses in From: header lines).
4063 (2) The list of non-personal local parts in From: addresses has been
4064 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4065 taken from the Sieve specification recommendations.
4067 (3) If the message contains any header line starting with "List-" it is
4068 treated as non-personal.
4070 (4) The test for "circular" in the Subject: header line has been removed
4071 because it now seems ill-conceived.
4073 15. Minor typos in src/EDITME comments corrected.
4075 16. Installed latest exipick from John Jetmore.
4077 17. If headers_add on a router specified a text string that was too long for
4078 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4079 of string_sprintf() is now avoided.
4081 18. $message_body_size was not set (it was always zero) when running the DATA
4082 ACL and the local_scan() function.
4084 19. For the "mail" command in an Exim filter, no default was being set for
4085 the once_repeat time, causing a random time value to be used if "once" was
4086 specified. (If the value happened to be <= 0, no repeat happened.) The
4087 default is now 0s, meaning "never repeat". The "vacation" command was OK
4088 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4089 (I found it when inspecting the code).
4091 20. There is now an overall timeout for performing a callout verification. It
4092 defaults to 4 times the callout timeout, which applies to individual SMTP
4093 commands during the callout. The overall timeout applies when there is more
4094 than one host that can be tried. The timeout is checked before trying the
4095 next host. This prevents very long delays if there are a large number of
4096 hosts and all are timing out (e.g. when the network connections are timing
4097 out). The value of the overall timeout can be changed by specifying an
4098 additional sub-option for "callout", called "maxwait". For example:
4100 verify = sender/callout=5s,maxwait=20s
4102 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4103 to the end before writing, but this should make it even safer).
4105 22. Exim was forgetting that it had advertised PIPELINING for the second and
4106 subsequent messages on an SMTP connection. It was also not resetting its
4107 memory on STARTTLS and an internal HELO.
4109 23. When Exim logs an SMTP synchronization error within a session, it now
4110 records whether PIPELINING has been advertised or not.
4112 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4113 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4114 rather than long int.
4116 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4118 26. Added the never_mail option to autoreply.
4124 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4125 crash if the getsockname() call failed; this can happen if a connection is
4126 closed very soon after it is established. The problem was simply in the
4127 order in which certain operations were done, causing Exim to try to write
4128 to the SMTP stream before it had set up the file descriptor. The bug has
4129 been fixed by making things happen in the correct order.
4135 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4136 before the connection was closed, thus losing the rejection response.
4138 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4139 some early Solaris releases, but causes trouble in current releases where
4140 socklen_t is defined.
4142 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4145 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4148 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4149 that releasing the top of it at the end releases what was used for sub-
4150 expansions (unless the block got too big). However, discard this block if
4151 the first thing is a variable or header, so that we can use its block when
4152 it is dynamic (useful for very large $message_headers, for example).
4154 6. Lookups now cache *every* query, not just the most recent. A new, separate
4155 store pool is used for this. It can be recovered when all lookup caches are
4156 flushed. Lookups now release memory at the end of their result strings.
4157 This has involved some general refactoring of the lookup sources.
4159 7. Some code has been added to the store_xxx() functions to reduce the amount
4160 of flapping under certain conditions.
4162 8. log_incoming_interface used to affect only the <= reception log lines. Now
4163 it causes the local interface and port to be added to several more SMTP log
4164 lines, for example "SMTP connection from", and rejection lines.
4166 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4168 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4170 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4171 could be overwritten at the end of the current message (or the start of a
4172 new message if it was set in a HELO ACL). The value is now preserved for
4173 the duration of the SMTP connection.
4175 12. If a transport had a headers_rewrite setting, and a matching header line
4176 contained an unqualified address, that address was qualified, even if it
4177 did not match any rewriting rules. The underlying bug was that the values
4178 of the flags that permit the existence of unqualified sender and recipient
4179 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4180 non-local messages, and by -bnq for local messages) were not being
4181 preserved with the message after it was received.
4183 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4184 "next input=" as part of the text comprising the host identity instead of
4185 the correct text. The code was using the same buffer for two different
4186 strings. However, depending on which order the printing function evaluated
4187 its arguments, the bug did not always show up. Under Linux, for example, my
4188 test suite worked just fine.
4190 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4191 doesn't work with some older versions of Perl. It has been changed to "my",
4192 which in any case is probably the better facility to use.
4194 15. A really picky compiler found some instances of statements for creating
4195 error messages that either had too many or two few arguments for the format
4198 16. The size of the buffer for calls to the DNS resolver has been increased
4199 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4200 for addresses that have a lot of PTR records. This alleviates a problem; it
4201 does not fully solve it.
4203 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4204 buffer now truncates the list and logs the incident, which is the same
4205 action as happens when Exim is looking up a host name and its aliases.
4206 Previously in this situation something unpredictable would happen;
4207 sometimes it was "internal error: store_reset failed".
4209 18. If a server dropped the connection unexpectedly when an Exim client was
4210 using GnuTLS and trying to read a response, the client delivery process
4211 crashed while trying to generate an error log message.
4213 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4214 string, for example:
4216 warn message = H1: something\nH2: something
4218 the text was added as a single header line from Exim's point of view
4219 though it ended up OK in the delivered message. However, searching for the
4220 second and subsequent header lines using $h_h2: did not work. This has been
4221 fixed. Similarly, if a system filter added multiple headers in this way,
4222 the routers could not see them.
4224 20. Expanded the error message when iplsearch is called with an invalid key to
4225 suggest using net-iplsearch in a host list.
4227 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4228 ACLs are no longer actually imposed (and a message to that effect is
4231 22. If a "gecos" field in a passwd entry contained escaped characters, in
4232 particular, if it contained a \" sequence, Exim got it wrong when building
4233 a From: or a Sender: header from that name. A second bug also caused
4234 incorrect handling when an unquoted " was present following a character
4235 that needed quoting.
4237 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4238 was not being matched caselessly.
4240 24. Arranged for all hyphens in the exim.8 source to be escaped with
4243 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4244 properly. Recipient callout cache records were still being keyed to include
4245 the sender, even when use_sender was set false. This led to far more
4246 callouts that were necessary. The sender is no longer included in the key
4247 when use_sender is false.
4249 26. Added "control = submission" modifier to ACLs.
4251 27. Added the ${base62d: operator to decode base 62 numbers.
4253 28. dnsdb lookups can now access SRV records.
4255 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4256 the configuration file.
4258 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4259 (-v) mode. This makes the output for a verbose queue run more intelligible.
4261 31. Added a use_postmaster feature to recipient callouts.
4263 32. Added the $body_zerocount variable, containing the number of binary zero
4264 bytes in the message body.
4266 33. The time of last modification of the "new" subdirectory is now used as the
4267 "mailbox time last read" when there is a quota error for a maildir
4270 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4272 35. Added +ignore_unknown as a special item in host lists.
4274 36. Code for decoding IPv6 addresses in host lists is now included, even if
4275 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4276 address was recognized as an IP address, but was then not correctly decoded
4277 into binary, causing unexpected and incorrect effects when compared with
4284 1. Very minor rewording of debugging text in manualroute to say "list of
4285 hosts" instead of "hostlist".
4287 2. If verify=header_syntax was set, and a header line with an unqualified
4288 address (no domain) and a large number of spaces between the end of the
4289 name and the colon was received, the reception process suffered a buffer
4290 overflow, and (when I tested it) crashed. This was caused by some obsolete
4291 code that should have been removed. The fix is to remove it!
4293 3. When running in the test harness, delay a bit after writing a bounce
4294 message to get a bit more predictability in the log output.
4296 4. Added a call to search_tidyup() just before forking a reception process. In
4297 theory, someone could use a lookup in the expansion of smtp_accept_max_
4298 per_host which, without the tidyup, could leave open a database connection.
4300 5. Added the variables $recipient_data and $sender_data which get set from a
4301 lookup success in an ACL "recipients" or "senders" condition, or a router
4302 "senders" option, similar to $domain_data and $local_part_data.
4304 6. Moved the writing of debug_print from before to after the "senders" test
4307 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4308 problems for message scanning, either using a data ACL, or using
4309 local_scan() because the Received: header was not generated till after they
4310 were called (in order to set the time as the time of reception completion).
4311 I have revised the way this works. The header is now generated after the
4312 body is received, but before the ACL or local_scan() are called. After they
4313 are run, the timestamp in the header is updated.
4319 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4320 before starting a queue runner without re-exec. This happened only when
4321 deliver_drop_privilege was set or when the Exim user was set to root. The
4322 effect of the bug was that timeouts during subsequent deliveries caused
4323 crashes instead of being properly handled. The handler is now left at its
4324 default (and expected) setting.
4326 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4327 message, again when deliver_drop_privilege is set or Exim is run as root.
4328 The bug described in (1) was not present in this case, but the tidying up
4329 of the other signals was missing. I have made the two cases consistent.
4331 3. The ignore_target_hosts setting on a manualroute router was being ignored
4332 for hosts that were looked up using the /MX notation.
4334 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4337 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4338 operated on the sender address. After changing the $sender_address to <>
4339 for the sender address verify, Exim was re-instated it as the original
4340 (before rewriting) address, but remembering that it had rewritten it, so it
4341 wasn't rewriting it again. This bug also had the effect of breaking the
4342 sender address verification caching when the sender address was rewritten.
4344 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4345 This has been changed so that if the ip literal address matches
4346 ignore_target_hosts, the router declines.
4348 7. Added expansion conditions match_domain, match_address, and match_local_
4349 part (NOT match_host).
4351 8. The placeholder for the Received: header didn't have a length field set.
4353 9. Added code to Exim itself and to exim_lock to test for a specific race
4354 condition that could lead to file corruption when using MBX delivery. The
4355 issue is with the lockfile that is created in /tmp. If this file is removed
4356 after a process has opened it but before that process has acquired a lock,
4357 there is the potential for a second process to recreate the file and also
4358 acquire a lock. This could lead to two Exim processes writing to the file
4359 at the same time. The added code performs the same test as UW imapd; it
4360 checks after acquiring the lock that its file descriptor still refers to
4361 the same named file.
4363 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4364 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4367 11. Added the smtp_active_hostname option. If used, this will typically be made
4368 to depend on the incoming interface address. Because $interface_address is
4369 not set up until the daemon has forked a reception process, error responses
4370 that can happen earlier (such as "too many connections") no longer contain
4373 12. If an expansion in a condition on a "warn" statement fails because a lookup
4374 defers, the "warn" statement is abandoned, and the next ACL statement is
4375 processed. Previously this caused the whole ACL to be aborted.
4377 13. Added the iplsearch lookup type.
4379 14. Added ident_timeout as a log selector.
4381 15. Added tls_certificate_verified as a log selector.
4383 16. Added a global option tls_require_ciphers (compare the smtp transport
4384 option of the same name). This controls incoming TLS connections.
4386 17. I finally figured out how to make tls_require_ciphers do a similar thing
4387 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4388 before starting the TLS session.
4390 18. Tabs are now shown as \t in -bP output.
4392 19. If the log selector return_path_on_delivery was set, Exim crashed when
4393 bouncing a message because it had too many Received: header lines.
4395 20. If two routers both had headers_remove settings, and the first one included
4396 a superfluous trailing colon, the final name in the first list and the
4397 first name in the second list were incorrectly joined into one item (with a
4398 colon in the middle).
4404 1. Added -C and -D options to the exinext utility, mainly to make it easier
4405 to include in the automated testing, but these could be helpful when
4406 multiple configurations are in use.
4408 2. The exinext utility was not formatting the output nicely when there was
4409 an alternate port involved in the retry record key, nor when there was a
4410 message id as well (for retries that were specific to a specific message
4411 and a specific host). It was also confused by IPv6 addresses, because of
4412 the additional colons they contain. I have fixed the IPv4 problem, and
4413 patched it up to do a reasonable job for IPv6.
4415 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4416 delivery, the log line now contains "pipelined" if PIPELINING was used.
4418 4. An SMTP transport process used to panic and die if the bind() call to set
4419 an explicit outgoing interface failed. This has been changed; it is now
4420 treated in the same way as a connect() failure.
4422 5. A reference to $sender_host_name in the part of a conditional expansion
4423 that was being skipped was still causing a DNS lookup. This no longer
4426 6. The def: expansion condition was not recognizing references to header lines
4427 that used bh_ and bheader_.
4429 7. Added the _cache feature to named lists.
4431 8. The code for checking quota_filecount in the appendfile transport was
4432 allowing one more file than it should have been.
4434 9. For compatibility with Sendmail, the command line option
4442 and sets the incoming protocol and host name (for trusted callers). The
4443 host name and its colon can be omitted when only the protocol is to be set.
4444 Note the Exim already has two private options, -pd and -ps, that refer to
4445 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4446 "s", but I don't think that's a major issue.
4448 10. A number of refactoring changes to the code, none of which should affect
4451 (a) The number of logging options was getting close to filling up the
4452 32-bit word that was used as a bit map. I have split them into two classes:
4453 those that are passed in the argument to log_write(), and those that are
4454 only ever tested independently outside of that function. These are now in
4455 separate 32-bit words, so there is plenty of room for expansion again.
4456 There is no change in the user interface or the logging behaviour.
4458 (b) When building, for example, log lines, the code previously used a
4459 macro that called string_cat() twice, in order to add two strings. This is
4460 not really sufficiently general. Furthermore, there was one instance where
4461 it was actually wrong because one of the argument was used twice, and in
4462 one call a function was used. (As it happened, calling the function twice
4463 did not affect the overall behaviour.) The macro has been replaced by a
4464 function that can join an arbitrary number of extra strings onto a growing
4467 (c) The code for expansion conditions now uses a table and a binary chop
4468 instead of a serial search (which was left over from when there were very
4469 few conditions). Also, it now recognizes conditions like "pam" even when
4470 the relevant support is not compiled in: a suitably worded error message is
4471 given if an attempt is made to use such a condition.
4473 11. Added ${time_interval:xxxxx}.
4475 12. A bug was causing one of the ddress fields not to be passed back correctly
4476 from remote delivery subprocesses. The field in question was not being
4477 subsequently used, so this caused to problems in practice.
4479 13. Added new log selectors queue_time and deliver_time.
4481 14. Might have fixed a bug in maildirsizefile handling that threw up
4482 "unexpected character" debug warnings, and recalculated the data
4483 unnecessarily. In any case, I expanded the warning message to give more
4486 15. Added the message "Restricted characters in address" to the statements in
4487 the default ACL that block characters like @ and % in local parts.
4489 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4490 Three changes have been made:
4492 (a) There was a serious bug; a negative response to MAIL caused the whole
4493 recipient domain to be cached as invalid, thereby blocking all messages
4494 to all local parts at the same domain, from all senders. This bug has
4495 been fixed. The domain is no longer cached after a negative response to
4496 MAIL if the sender used is not empty.
4498 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4501 (c) A new callout option, "use_sender" has been added for people who want
4502 the modified behaviour.
4508 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4511 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4512 indeed breaks things for older releases.
4514 3. Added additional logging to the case where there is a problem reading data
4515 from a filter that is running in a subprocess using a pipe, in order to
4516 try to track down a specific problem.
4518 4. Testing facility fudge: when running in the test harness and attempting
4519 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4520 getting "No route to host". Convert this to a timeout.
4522 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4525 6. Some OS don't have socklen_t but use size_t instead. This affects the
4526 fifth argument of getsockopt() amongst other things. This is now
4527 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4528 can be set for individual OS. I have set it for SunOS5, OSF1, and
4529 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4530 some earlier ones do not.
4532 7. Change 4.30/15 was not doing the test caselessly.
4534 8. The standard form for an IPv6 address literal was being rejected by address
4535 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4536 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4537 this, as well as the form without the "IPv6" on the front (but only when
4538 address literals are enabled, of course).
4540 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4542 10. Exim crashed if a message with an empty sender address specified by -f
4543 encountered a router with an errors_to setting. This could be provoked only
4544 by a command such as
4548 where an empty string was supplied; "<>" did not hit this bug.
4550 11. Installed PCRE release 4.5.
4552 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4553 remained set. It is now erased.
4555 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4556 times from message ids (which are base 36 rather than the normal 62).
4558 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4559 were being counted as actual protocol errors, and logged if the log
4560 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4561 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4562 and DATA following a set of rejected RCPTs do not count as protocol errors.
4563 In other words, Exim assumes they were pipelined, though this may not
4564 actually be the case. Of course, in all cases the client gets an
4565 appropriate error code.
4567 15. If a lookup fails in an ACL condition, a message about the failure may
4568 be available; it is used if testing the ACL cannot continue, because most
4569 such messages specify what the cause of the deferral is. However, some
4570 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4571 that caused an old message to be retained and used if a later statement
4572 caused a defer, replacing the real cause of the deferral.
4574 16. If an IP address had so many PTR records that the DNS lookup buffer
4575 was not large enough to hold them, Exim could crash while trying to process
4576 the truncated data. It now detects and logs this case.
4578 17. Further to 4.21/58, another change has been made: if (and only if) the
4579 first line of a message (the first header line) ends with CRLF, a bare LF
4580 in a subsequent header line has a space inserted after it, so as not to
4581 terminate the header.
4583 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4584 unnecessarily, used atoi() instead of strtol(), and didn't check the
4585 termination when getting file sizes from file names by regex.
4587 19. Completely re-implemented the support for maildirsize files, in the light
4588 of a number of problems with the previous contributed implementation
4589 (4.30/29). In particular:
4591 . If the quota is zero, the maildirsize file is maintained, but no quota is
4594 . If the maildir directory does not exist, it is created before any attempt
4595 to write a maildirsize file.
4597 . The quota value in the file is just a cache; if the quota is changed in
4598 the transport, the new value overrides.
4600 . A regular expression is available for excluding directories from the
4603 20. The autoreply transport checks the characters in options that define the
4604 message's headers; it allows continued headers, but it was checking with
4605 isspace() after an embedded newline instead of explicitly looking for a
4608 21. If all the "regular" hosts to which an address was routed had passed their
4609 expiry times, and had not reached their retry times, the address was
4610 bounced, even if fallback hosts were defined. Now Exim should go on to try
4613 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4614 equivalent code in the SMTP transport. Some hosts send humungous responses
4615 to HELO/EHLO, more than 1024 it seems.
4617 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4618 gives compiler warnings in some environments. I've now done it "properly",
4621 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4622 (because of problems with the built-in one) was declared to return uschar *
4623 instead of char *, causing compiler failure.
4625 25. Fixed a file descriptor leak when processing alias/forward files.
4627 26. Fixed a minor format string issue in dbfn.c.
4629 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4631 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4632 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4635 29. When a sender address is verified, it is cached, to save repeating the test
4636 when there is more than one recipient in a message. However, when the
4637 verification involves a callout, it is possible for different callout
4638 options to be set for different recipients. It is too complicated to keep
4639 track of this in the cache, so now Exim always runs a verification when a
4640 callout is required, relying on the callout cache for the optimization.
4641 The overhead is duplication of the address routing, but this should not be
4644 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4645 to be verified with callout=postmaster, and the main callout worked but the
4646 postmaster check failed, the verification correctly failed. However, if a
4647 subsequent RCPT command asked for sender verification *without* the
4648 postmaster check, incorrect caching caused this verification also to fail,
4651 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4652 it was not caching the DNS options (qualify_single, search_parents) that
4653 were used when the lookup failed. A subsequent lookup with different
4654 options therefore always gave the same answer, though there were cases
4655 where it should not have. (Example: a "domains = !$mx_any" option on a
4656 dnslookup router: the "domains" option is always processed without any
4657 widening, but the router might have qualify_single set.) Now Exim uses the
4658 cached value only when the same options are set.
4660 32. Added John Jetmore's "exipick" utility to the distribution.
4662 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4663 than a timeout (e.g. a certificate is required, and is not provided), an
4664 Exim server now closes the connection immediately. Previously it waited for
4665 the client to close - but if the client is SSL, it seems that they each
4666 wait for each other, leading to a delay before one of them times out.
4668 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4669 maintained 0.8.x compatibility because I don't think many are using it, and
4670 it is clearly obsolete.
4672 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4675 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4676 client certificate was expired. A simple patch fixes this, though I don't
4677 understand the full logic of why the verify callback is called multiple
4680 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4681 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4682 which causes problems with some clients (such as the Certicom SSL Plus
4683 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4684 disables the coutermeasure allowing Eudora to connect."
4686 38. Exim was not checking that a write() to a log file succeeded. This could
4687 lead to Bad Things if a log got too big, in particular if it hit a file
4688 size limit. Exim now panics and dies if it cannot write to a log file, just
4689 as it does if it cannot open a log file.
4691 39. Modified OS/Makefile-Linux so that it now contains
4693 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4695 The two -D definitions ensure that Exim is compiled with large file
4696 support, which makes it possible to handle log files that are bigger than
4699 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4700 instance) a domain was checked against a named list that involved a lookup,
4701 causing $domain_data to be set, then another domain was checked against the
4702 same list, then the first domain was re-checked, the value of $domain_data
4703 after the final check could be wrong. In particular, if the second check
4704 failed, it could be set empty. This bug probably also applied to
4707 41. The strip_trailing_dot option was not being applied to the address given
4708 with the -f command-line option.
4710 42. The code for reading a message's header from the spool was incrementing
4711 $received_count, but never initializing it. This meant that the value was
4712 incorrect (doubled) while delivering a message in the same process in which
4713 it was received. In the most common configuration of Exim, this never
4714 happens - a fresh exec is done - but it can happen when
4715 deliver_drop_privilege is set.
4717 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4718 it now includes up to 150 characters of the unexpected data in the log
4721 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4722 and building data strings. The size of both of these buffers was 10 000
4723 bytes - far larger than anybody would *ever* want, thought I. Needless to
4724 say, somebody hit the limit. I have increased the maximum line length to
4725 20 000 and the maximum data length of concatenated lines to 100 000. I have
4726 also fixed two bugs, because there was no checking on these buffers. Tsk,
4727 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4728 buffer is too small.
4730 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4731 lsearch lookups. Now it does.
4733 46. When parsing a route_list item in a manualroute router, a fixed-length
4734 buffer was used for the list of hosts. I made this 1024 bytes long,
4735 thinking that nobody would ever have a list of hosts that long. Wrong.
4736 Somebody had a whole pile of complicated expansion conditions, and the
4737 string was silently truncated, leading to an expansion error. It turns out
4738 that it is easier to change to an unlimited length (owing to other changes
4739 that have happened since this code was originally written) than to build
4740 structure for giving a limitation error. The length of the item that
4741 expands into the list of hosts is now unlimited.
4743 47. The lsearch lookup could not handle data where the length of text line was
4744 more than 4095 characters. Such lines were truncated, leading to shortened
4745 data being returned. It should now handle lines of any length.
4747 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4748 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4751 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4752 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4753 no longer happens in comment lines. A list of replacements is now placed
4754 at the head of all of the source files, except those whose only change is
4755 to replace PERL_COMMAND in the very first #! line.
4757 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4758 messages on the queue, with a bottom-up merge sort, using code contributed
4759 by Michael Haardt. This should make operations like -bp somewhat faster on
4760 large queues. It won't affect queue runners, except when queue_run_in_order
4763 51. Installed eximstats 1.31 in the distribution.
4765 52. Added support for SRV lookups to the dnslookup router.
4767 53. If an ACL referred to $message_body or $message_body_end, the value was not
4768 reset for any messages that followed in the same SMTP session.
4770 54. The store-handling optimization for building very long strings was not
4771 differentiating between the different store pools. I don't think this
4772 actually made any difference in practice, but I've tidied it.
4774 55. While running the routers to verify a sender address, $sender_address
4775 was still set to the sender address. This is wrong, because when routing to
4776 send a bounce to the sender, it would be empty. Therefore, I have changed
4777 it so that, while verifying a sender address, $sender_address is set to <>.
4778 (There is no change to what happens when verifying a recipient address.)
4780 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4781 target A or AAAA records (if not already returned) without resetting the
4782 qualify_single or search_parents options of the DNS resolver. These are
4783 inappropriate in this case because the targets of MX and SRV records must
4784 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4785 target that, when qualified, matched something in the local domain. These
4786 two options are now turned off when doing these lookups.
4788 57. It seems that at least some releases of Reiserfs (which does not have the
4789 concept of a fixed number of inodes) returns zero and not -1 for the
4790 number of available inodes. This interacted badly with check_spool_inodes,
4791 which assumed that -1 was the "no such thing" setting. What I have done is
4792 to check that the total number of inodes is greater than zero before doing
4793 the test of how many are available.
4795 58. When a "warn" ACL statement has a log_message modifier, the message is
4796 remembered, and not repeated. This is to avoid a lot of repetition when a
4797 message has many recipients that cause the same warning to be written.
4798 Howewer, Exim was preserving the list of already written lines for an
4799 entire SMTP session, which doesn't seem right. The memory is now reset if a
4800 new message is started.
4802 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4803 debugging output unless log_rewrite was also set.
4805 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4806 of (int)(handle) when we know that handle contains (void *)(-1).
4808 61. The Exim daemon panic-logs an error return when it closes the incoming
4809 connection. However "connection reset by peer" seems to be common, and
4810 isn't really an error worthy of noting specially, so that particular error
4813 62. When Exim is trying to find all the local interfaces, it used to panic and
4814 die if the ioctl to get the interface flags failed. However, it seems that
4815 on at least one OS (Solaris 9) it is possible to have an interface that is
4816 included in the list of interfaces, but for which you get a failure error
4817 for this call. This happens when the interface is not "plumbed" into a
4818 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4819 failure of the "get flags" call assumes that the interface is down.
4821 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4822 makes life easier for people who are doing arithmetic on fields extracted
4823 from dates, where you often get leading zeros that should not be
4824 interpreted as octal.
4826 64. Added qualify_domain to the redirect router, to override the global
4829 65. If a pathologically long header line contained very many addresses (the
4830 report of this problem mentioned 10 000) and each of them was rewritten,
4831 Exim could use up a very large amount of memory. (It kept on making new
4832 copies of the header line as it rewrote, and never released the old ones.)
4833 At the expense of a bit more processing, the header rewriting function has
4834 been changed so that it no longer eats memory in this way.
4836 66. The generation of the Received: header has been moved from the time that a
4837 message starts to be received, to the time that it finishes. The timestamp
4838 in the Received: header should now be very close to that of the <= log
4839 line. There are two side-effects of this change:
4841 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4842 logged header lines no longer include the local Received: line, because
4843 it has not yet been created. The same applies to a copy of the message
4844 that is returned to a non-SMTP sender when a message is rejected.
4846 (b) When a filter file is tested using -bf, no additional Received: header
4847 is added to the test message. After some thought, I decided that this
4850 This change does not affect the value of $received_for. It is still set
4851 after address rewriting, but before local_scan() is called.
4853 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4855 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4856 gave an unhelpful panic error message, and a defer error. I have managed to
4857 change this behaviour so that it now rejects any supplied certificate,
4858 which seems right, as the list of acceptable certificates is empty.
4860 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4861 gave an unhelpful defer error. I have not managed to make this reject any
4862 supplied certificates, but the error message it gives is "no certificate
4863 supplied", which is not helpful.
4865 70. exigrep's output now also includes lines that are not associated with any
4866 message, but which match the given pattern. Implemented by a patch from
4867 Martin Sluka, which also tidied up the Perl a bit.
4869 71. Recipient callout verification, like sender verification, was using <> in
4870 the MAIL FROM command. This isn't really the right thing, since the actual
4871 sender may affect whether the remote host accepts the recipient or not. I
4872 have changed it to use the actual sender in the callout; this means that
4873 the cache record is now keyed on a recipient/sender pair, not just the
4874 recipient address. There doesn't seem to be a real danger of callout loops,
4875 since a callout by the remote host to check the sender would use <>.
4876 [SEE ABOVE: changed after hitting problems.]
4878 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4879 temporary errors. However, in the case of such a code being given after
4880 the end of a data transmission (i.e. after ".") Exim was failing to write
4881 a retry record for the message. (Yes, there was some broken host that was
4882 actually sending 8xx at this point.)
4884 73. An unknown lookup type in a host list could cause Exim to panic-die when
4885 the list was checked. (An example that provoked this was putting <; in the
4886 middle of a list instead of at the start.) If this happened during a DATA
4887 ACL check, a -D file could be left lying around. This kind of configuration
4888 error no longer causes Exim to die; instead it causes a defer errror. The
4889 incident is still logged to the main and panic logs.
4891 74. Buglet left over from Exim 3 conversion. The message "too many messages
4892 in one connection" was written to the rejectlog but not the mainlog, except
4893 when address rewriting (yes!) was being logged.
4895 75. Added write_rejectlog option.
4897 76. When a system filter was run not as root (that is, when system_filter_user
4898 was set), the values of the $n variables were not being returned to the
4899 main process; thus, they were not subsequently available in the $sn
4902 77. Added +return_path_on_delivery log selector.
4904 78. A connection timeout was being treated differently from recipients deferred
4905 when testing hosts_max_try with a message that was older than the host's
4906 retry timeout. (The host should not be counted, thus allowing all hosts to
4907 be tried at least once before bouncing.) This may have been the cause of an
4908 occasionally reported bug whereby a message would remain on the queue
4909 longer than the retry timeout, but would be bounced if a delivery was
4910 forced. I say "may" because I never totally pinned down the problem;
4911 setting up timeout/retry tests is difficult. See also the next item.
4913 79. The ultimate address timeout was not being applied to errors that involved
4914 a combination of host plus message (for example, a timeout on a MAIL
4915 command). When an address resolved to a number of possible hosts, and they
4916 were not all tried for each delivery (e.g. because of hosts_max_try), a
4917 message could remain on the queue longer than the retry timeout.
4919 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4922 81. Fixed an obscure SMTP outgoing bug which required at least the following
4923 conditions: (a) there was another message waiting for the same server;
4924 (b) the server returned 5xx to all RCPT commands in the first message so
4925 that the message was not completed; (c) the server dropped the connection
4926 or gave a negative response to the RSET that Exim sends to abort the
4927 transaction. The observed case was a dropped connection after DATA that had
4928 been sent in pipelining mode. That is, the server had advertised PIPELINING
4929 but was not implementing it correctly. The effect of the bug was incorrect
4930 behaviour, such as trying another host, and this could lead to a crash.
4936 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4937 and daemon.c were passed as pointers to ints; they should have been
4938 pointers to socklen_t variables (which are typically unsigned ints).
4940 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4943 3. Fixed a really odd bug that affected only the testing scheme; patching a
4944 certain fixed string in the binary changed the value of another string that
4945 happened to be identical to the end of the original first string.
4947 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4948 name", it returns that address as the IP address. On some operating
4949 systems (e.g. Solaris), it also passes back the IP address string as the
4950 "host name". However, on others (e.g. Linux), it passes back an empty
4951 string. Exim wasn't checking for this, and was changing the host name to an
4952 empty string, assuming it had been canonicized.
4954 5. Although rare, it is permitted to have more than one PTR record for a given
4955 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4956 all the names associated with an address, because they do in Solaris.
4957 However, it seems that they do not in Linux for data that comes from the
4958 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4959 I found this out when I moved to a new Linux workstation and tried to run
4960 the Exim test suite.
4962 To get round this problem I have changed the code so that it now does its
4963 own call to the DNS to look up PTR records when searching for a host name.
4964 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4965 addresses that are only in /etc/hosts are still found.
4967 This behaviour is, however, controlled by an option called host_lookup_
4968 order, which defaults to "bydns:byaddr". If people want to use the other
4969 order, or indeed, just use one or the other means of lookup, they can
4970 specify it in this variable.
4972 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4973 some operating systems, this comes back from gethostbyaddr() as an empty
4974 string, and this is what Exim used to test for. However, it seems that in
4975 other systems, "." is yielded. Exim now tests for this case too.
4977 7. The values of check_spool_space and check_log_space are now held internally
4978 as a number of kilobytes instead of an absolute number of bytes. If a
4979 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4980 kilobyte. This means that much larger values can be stored.
4982 8. Exim monitor: an attempt to get the action menu when not actually pointing
4983 at a message produces an empty menu entitled "No message selected". This
4984 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4985 no entries in it ("Shell widget menu has zero width and/or height"). So I
4986 have added a single, blank menu entry in this case.
4988 9. Added ${quote_local_part.
4990 10. MIME decoding is now applied to the contents of Subject: header lines when
4993 11. Now that a reference to $sender_host_address automatically causes a reverse
4994 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4995 host lookup before query-style lookups in lists that might use this
4996 variable. This has therefore been abolished, and the "net-" prefix is no
4997 longer necessary for query-style lookups.
4999 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5000 have been a typo for LFLAGS, so it has been changed.
5002 13. The install script calls Exim with "-C /dev/null" in order to find the
5003 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5004 to be output. Howeve, since Exim outputs its version number before the
5005 error, it didn't break the script. It just looked ugly. I fixed this by
5006 always allowing "-C /dev/null" if the caller is root.
5008 14. Ignore overlarge ACL variable number when reading spool file - insurance
5009 against a later release with more variables having written the file.
5011 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5012 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5013 this, as well as the form without the "IPv6" on the front.
5015 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5016 OS/Makefile-Darwin file.
5018 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5019 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5021 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5022 variable. However, some OS use ints for inodes. I've added cast to long int
5023 to get rid of the compiler warning.
5025 19. I had forgotten to lock out "/../" in configuration file names when
5026 ALT_CONFIG_PREFIX was set.
5028 20. Routers used for verification do not need to specify transports. However,
5029 if such a router generated a host list, and callout was configured, Exim
5030 crashed, because it could not find a port number from the (non-existent)
5031 transport. It now assumes port 25 in this circumstance.
5033 21. Added the -t option to exigrep.
5035 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5036 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5037 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5038 them is set, LOOKUP_LSEARCH is forced.
5040 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5042 24. Added sender and host information to the "rejected by local_scan()" log
5043 line; previously there was no indication of these.
5045 25. Added .include_if_exists.
5047 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5048 Linux. It turns out that not all file systems support this. Apparently some
5049 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5050 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5051 not supported on the file descriptor, is now ignored when Exim is trying to
5052 sync a directory. This applies only to Linux.
5054 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5056 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5057 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5060 29. Added support for maildirsize files from supplied patch (modified a bit).
5062 30. The use of :fail: followed by an empty string could lead Exim to respond to
5063 sender verification failures with (e.g.):
5065 550 Verification failed for <xxx>
5066 550 Sender verify failed
5068 where the first response line was missing the '-' that indicates it is not
5069 the final line of the response.
5071 31. The loop for finding the name of the user that called Exim had a hardwired
5072 limit of 10; it now uses the value of finduser_retries, which is used for
5073 all other user lookups.
5075 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5078 33. Exim was neglecting to zero errno before one call of strtol() when
5079 expanding a string and expecting an integer value. On some systems this
5080 resulted in spurious "integer overflow" errors. Also, it was casting the
5081 result into an int without checking.
5083 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5084 did not work. The code looks as if it has *never* worked, though it appears
5085 to have been documented since at least releast 1.62. I have made it work.
5087 35. The "timeout_DNS" error in retry rules, also documented since at least
5088 1.62, also never worked. As it isn't clear exactly what this means, and
5089 clearly it isn't a major issue, I have abolished the feature by treating it
5090 as "timeout", and writing a warning to the main and panic logs.
5092 36. The display of retry rules for -brt wasn't always showing the error code
5095 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5096 timeout_connect_A, timeout_connect_MX.
5098 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5099 to the empty sender.
5101 39. The daemon was not analysing the content of -oX till after it had closed
5102 stderr and disconnected from the controlling terminal. This meant that any
5103 syntax errors were only noted on the panic log, and the return code from
5104 the command was 0. By re-arranging the code a little, I've made the
5105 decoding happen first, so such errors now appear on stderr, and the return
5106 code is 1. However, the actual setting up of the sockets still happens in
5107 the disconnected process, so errors there are still only recorded on the
5110 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5111 connections (as happens on some IP stacks) was logged at start up time as
5112 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5113 it from "IPv6 and IPv4", which means that two separate sockets are being
5116 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5117 says whether AF_INET or AF_INET6 was passed as an argument.
5119 42. Exiwhat output was messed up when time zones were included in log
5122 43. Exiwhat now gives more information about the daemon's listening ports,
5123 and whether -tls-on-connect was used.
5125 44. The "port" option of the smtp transport is now expanded.
5127 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5128 silently ignored. Now an error message is written to the main and panic
5131 46. There's a new ACL modifier called "logwrite" which writes to a log file
5132 as soon as it is encountered.
5134 47. Added $local_user_uid and $local_user_gid at routing time.
5136 48. Exim crashed when trying to verify a sender address that was being
5139 49. Exim was recognizing only a space character after ".include". It now also
5140 recognizes a tab character.
5142 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5143 extracting the relevant information from the specification. The man page no
5144 longer contains scrambled data for the -d option, and I've added a section
5145 at the front about calling Exim under different names.
5147 51. Added "extra_headers" argument to the "mail" command in filter files.
5149 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5152 53. Installed eximstats 1.29.
5154 54. Added transport_filter_timeout as a generic transport option.
5156 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5157 Cc: header lines. This was required by RFC 822, but it not required by RFC
5160 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5161 incoming messages that did not have them. Now it does so only if the
5162 message originates locally, that is, if there is no associated remote host
5163 address. When Resent- header lines are present, this applies to the Resent-
5164 lines rather than the non-Resent- lines.
5166 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5167 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5169 58. Messages for configuration errors now include the name of the main
5170 configuration file - useful now that there may be more than one file in a
5171 list (.included file names were always shown).
5173 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5174 for those rare installations that do not start the daemon as root or run it
5175 setuid root. I've cut out the call to initgroups() if the daemon is not
5178 60. The Exim user and group can now be bound into the binary as text strings
5179 that are looked up at the start of Exim's processing.
5181 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5183 62. Added $mailstore_basename variable.
5185 63. Installed patch to sieve.c from Michael Haardt.
5187 64. When Exim failed to open the panic log after failing to open the main log,
5188 the original message it was trying to log was written to stderr and debug
5189 output, but if they were not available (the usual case in production), it
5190 was lost. Now it is written to syslog before the two lines that record the
5191 failures to open the logs.
5193 65. Users' Exim filters run in subprocesses under the user's uid. It is
5194 possible for a "deliver" command or an alias in a "personal" command to
5195 provoke an address rewrite. If logging of address rewriting is configured,
5196 this fails because the process is not running as root or exim. There may be
5197 a better way of dealing with this, but for the moment (because 4.30 needs
5198 to be released), I have disabled address rewrite logging when running a
5199 filter in a non-root, non-exim process.
5205 1. The buildconfig auxiliary program wasn't quoting the value set for
5206 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5207 not defined. This bug was masked in 4.22 by the effect that was fixed in
5210 2. Some messages that were rejected after a message id was allocated were
5211 shown as "incomplete" by exigrep. It no longer does this for messages that
5212 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5214 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5215 have allow_domain_literals set, the ID did not get logged in the <= line.
5216 Domain literals are now always recognized in Message-ID: header lines.
5218 4. The first argument for a ${extract expansion item is the key name or field
5219 number. Leading and trailing spaces in this item were not being ignored,
5220 causing some misleading effects.
5222 5. When deliver_drop_privilege was set, single queue runner processes started
5223 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5224 same command in the process it spins off) were not dropping privilege.
5226 6. When the daemon running as "exim" started a queue runner, it always
5227 re-executed Exim in the spun-off process. This is a waste of effort when
5228 deliver_drop_privilege is set. The new process now just calls the
5229 queue-runner function directly.
5235 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5238 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5239 ignored. Though the use of -oP was forcing the writing of a pid file, it
5240 was always written to the default place.
5242 3. If the message "no IP address found for host xxxx" is generated during
5243 incoming verification, it is now followed by identification of the incoming
5244 connection (so you can more easily find what provoked it).
5246 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5248 5. Added some features to "harden" Exim a bit more against certain attacks:
5250 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5251 be put in Local/Makefile. This is like the never_users runtime option,
5252 but it cannot be overridden. The default setting is "root".
5254 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5255 prefix string with which any file named in a -C command line option
5258 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5259 is retained for -C and -D only if the caller of Exim is root. Without
5260 it, the exim user may also use -C and -D and retain privilege.
5262 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5263 command line option is disabled.
5265 6. Macro names set by the -D option must start with an upper case letter, just
5266 like macro names defined in the configuration file.
5268 7. Added "dereference=" facility to LDAP.
5270 8. Two instances of the typo "uknown" in the source files are fixed.
5272 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5273 the Configure-Makefile script screwed up while processing it.
5275 10. Incorporated PCRE 4.4.
5277 11. The SMTP synchronization check was not operating right at the start of an
5278 SMTP session. For example, it could not catch a HELO sent before the client
5279 waited for the greeting. There is now a check for outstanding input at the
5280 point when the greeting is written. Because of the duplex, asynchronous
5281 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5282 way, but not yet received, when the check is performed.
5284 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5285 on TCP/IP sockets, because this apparently causes some broken clients to
5288 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5289 unchanged) from the Cygwin maintainer.
5291 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5292 was supported instead of testing for mbx. Effectively a typo.
5294 15. The spa authenticator server code was not checking that the input it
5295 received was valid base64.
5297 16. The debug output line for the "set" modifier in ACLs was not showing the
5298 name of the variable that was being set.
5300 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5302 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5303 lookup on its own. Something else was needed to trigger the lookup. For
5304 example, a match in host_lookup or the need for a host name in a host list.
5305 Now, if $sender_host_name is referenced and the host name has not yet been
5306 looked up, a lookup is performed. If the lookup fails, the variable remains
5307 empty, and $host_lookup_failed is set to "1".
5309 19. Added "eqi" as a case-independent comparison operator.
5311 20. The saslauthd authentication condition could segfault if neither service
5312 nor realm was specified.
5314 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5315 error message that was logged was misleading, and incoming SMTP
5316 connections were dropped. The message is now more accurate, and temporary
5317 errors are given to SMTP connections.
5319 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5320 (except RSET) with a 421 error, until QUIT is received. However, it was
5321 failing to send a response to QUIT.
5323 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5324 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5325 ACL is now run after the helo_try_verify_hosts code.
5327 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5328 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5329 case-independent, so other case variants are also recognized). Apparently
5330 some systems use these upper case variants.
5332 25. If more than two messages were waiting for the same host, and a transport
5333 filter was specified for the transport, Exim sent two messages over the
5334 same TCP/IP connection, and then failed with "socket operation on non-
5335 socket" when it tried to send the third.
5337 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5339 27. The extern definition of crypt16() in expand.c was not being excluded when
5340 the OS had its own crypt16() function.
5342 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5343 as a preferred synonym for return_size_limit, both as an option and as an
5346 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5348 30. Changed the default configuration ACL to relax the local part checking rule
5349 for addresses that are not in any local domains. For these addresses,
5350 slashes and pipe symbols are allowed within local parts, but the sequence
5351 /../ is explicitly forbidden.
5353 31. SPA server authentication was not clearing the challenge buffer before
5356 32. log_message in a "warn" ACL statement was writing to the reject log as
5357 well as to the main log, which contradicts the documentation and doesn't
5358 seem right (because no rejection is happening). So I have stopped it.
5360 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5361 However, I am unable to do any testing of this.
5363 34. Fixed an infelicity in the appendfile transport. When checking directories
5364 for a mailbox, to see if any needed to be created, it was accidentally
5365 using path names with one or more superfluous leading slashes; tracing
5366 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5368 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5369 contents are added to the log line that is written for every discarded
5370 recipient. (Previously a log_message setting was ignored.)
5372 36. The ${quote: operator now quotes the string if it is empty.
5374 37. The install script runs exim in order to find its version number. If for
5375 some reason other than non-existence or emptiness, which it checks, it
5376 could not run './exim', it was installing it with an empty version number,
5377 i.e. as "exim-". This error state is now caught, and the installation is
5380 38. An argument was missing from the function that creates an error message
5381 when Exim fails to connect to the socket for saslauthd authentication.
5382 This could cause Exim to crash, or give a corrupted message.
5384 39. Added isip, isip4, and isip6 to ${if conditions.
5386 40. The ACL variables $acl_xx are now saved with the message, and can be
5387 accessed later in routers, transports, and filters.
5389 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5390 strings in the file are not string-expanded.
5392 42. If a MAIL command specified a SIZE value that was too large to fit into an
5393 int variable, the check against message_size_limit failed. Such values are
5394 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5395 day this will have to be increased, but I don't think I want to be around
5396 when emails are that large.
5403 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5404 iconv() is not standard in FreeBSD.
5406 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5407 IPv6 enabled. The observed symptom was a segmentation fault on return from
5408 the function os_common_find_running_interfaces() in src/os.c.
5410 3. In the check_special_case() function in daemon.c I had used "errno" as an
5411 argument name, which causes warnings on some systems. This was basically a
5412 typo, since it was named "eno" in the comments!
5414 4. The code that waits for the clock to tick (at a resolution of some fraction
5415 of a second) so as to ensure message-id uniqueness was always waiting for
5416 at least one whole tick, when it could have waited for less. [This is
5417 almost certainly not relevant at current processor speeds, where it is
5418 unlikely to ever wait at all. But we try to future-proof.]
5420 5. The function that sleeps for a time interval that includes fractions of a
5421 second contained a race. It did not block SIGALRM between setting the
5422 timer, and suspending (a couple of lines later). If the interval was short
5423 and the sigsuspend() was delayed until after it had expired, the suspension
5424 never ended. On busy systems this could lead to processes getting stuck for
5427 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5428 process, before it forks any delivery processes. The open lookup caching
5429 mechanism meant that the open file or database connection was passed into
5430 the delivery process. The problem was that delivery processes always tidy
5431 up cached lookup data. This could cause a problem for the next delivery
5432 process started by the queue runner, because the external queue runner
5433 process does not know about the closure. So the next delivery process
5434 still has data in the lookup cache. In the case of a file lookup, there was
5435 no problem because closing a file descriptor in a subprocess doesn't affect
5436 the parent. However, if the lookup was caching a connection to a database,
5437 the connection was closed, and the second delivery process was likely to
5438 see errors such as "PGSQL: query failed: server closed the connection
5439 unexpectedly". The problem has been fixed by closing all cached lookups
5440 in a queue runner before running a delivery process.
5442 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5443 seem to have the "const" qualifier which it has on other OS. I've
5446 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5447 *of the same type* (client or server) with the same public name that an
5448 error should be diagnosed.
5450 9. When Exim looked up a host name for an IP address, but failed to find the
5451 original IP address when looking up the host name (a safety check), it
5452 output the message "<ip address> does not match any IP for NULL", which was
5453 confusing, to say the least. The bug was that the host name should have
5454 appeared instead of "NULL".
5456 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5457 user that is built into the binary, and the -C or -D options is used, root
5458 privilege is dropped before the configuration file is read. In addition,
5459 logging is switched to stderr instead of the normal log files. If the
5460 configuration then re-defines the Exim user, the unprivileged environment
5461 is probably not what is expected, so Exim logs a panic warning message (but
5464 However, if deliver_drop_privilege is set, the unprivileged state may well
5465 be exactly what is intended, so the warning has been cut out in that case,
5466 and Exim is allowed to try to write to its normal log files.
5472 1. smtp_return_error_details was not giving details for temporary sender
5473 or receiver verification errors.
5475 2. Diagnose a configuration error if two authenticators have the same public
5478 3. Exim used not to create the message log file for a message until the first
5479 delivery attempt. This could be confusing when incoming messages were held
5480 for policy or load reasons. The message log file is now created at the time
5481 the message is received, and an initial "Received" line is written to it.
5483 4. The automatically generated man page for command line options had a minor
5484 bug that caused no ill effects; however, a more serious problem was that
5485 the procedure for building the man page automatically didn't always
5486 operate. Consequently, release 4.20 contains an out-of-date version. This
5487 shouldn't happen again.
5489 5. When building Exim with embedded Perl support, the script that builds the
5490 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5491 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5493 6. The freeze_tell option was not being used for messages that were frozen on
5494 arrival, either by an ACL or by local_scan().
5496 7. Added the smtp_incomplete_transaction log selector.
5498 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5499 was accepting AUTH without a new EHLO.
5501 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5502 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5505 10. Logging of TCP/IP connections (when configured) now happens in the main
5506 daemon process instead of the child process, so that the TCP/IP connection
5507 count is more accurate (but it can never be perfect).
5509 11. The use of "drop" in a nested ACL was not being handled correctly in the
5510 outer ACL. Now, if condition failure induced by the nested "drop" causes
5511 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5512 or "require"), the connection is dropped.
5514 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5515 that yield "discard" can now be used with an "accept" or a "discard" verb,
5516 but an error is generated for any others (because I can't see a useful way
5517 to define what should happen).
5519 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5520 are now processed in the same way as lines in the Exim configuration file.
5521 In particular, continuation lines are supported.
5523 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5525 15. Added -ti meaning -t -i.
5527 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5528 domains, and warn by logging if others are found.
5530 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5531 returned from GIFCONF when Exim is trying to find the list of interfaces on
5532 a host. The code in os.c has been modified to copy each ifreq to an aligned
5533 structure in all cases.
5535 Also, in some cases, the returned ifreq's were being copied to a 'struct
5536 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5537 means the last couple of bytes of an IPv6 address could be chopped if the
5538 ifreq contained only a normal sockaddr (14 bytes storage).
5540 18. Named domain lists were not supported in the hosts_treat_as_local option.
5541 An entry such as +xxxx was not recognized, and was treated as a literal
5544 19. Ensure that header lines added by a DATA ACL are included in the reject log
5545 if the ACL subsequently rejects the message.
5547 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5548 MD5 (which is deprecated).
5550 21. When testing a filter file using -bf, Exim was writing a message when it
5551 took the sender from a "From " line in the message, but it was not doing so
5552 when it took $return_path from a Return-Path: header line. It now does.
5554 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5555 with a valid header line field name (a series of printing characters
5556 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5558 23. Changed "disc" in the source to "disk" to conform to the documentation and
5559 the book and for uniformity.
5561 24. Ignore Sendmail's -Ooption=value command line item.
5563 25. When execve() failed while trying to run a command in a pipe transport,
5564 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5565 could be confused with a return value of 69 from the command itself. This
5566 has been changed to 127, the value the shell returns if it is asked to run
5567 a non-existent command. The wording for the related log line suggests a
5568 non-existent command as the problem.
5570 26. If received_header_text expands to an empty string, do not add a Received:
5571 header line to the message. (Well, it adds a token one on the spool, but
5572 marks it "old" so that it doesn't get used or transmitted.)
5574 27. Installed eximstats 1.28 (addition of -nt option).
5576 28. There was no check for failure on the call to getsockname() in the daemon
5577 code. This can fail if there is a shortage of resources on the system, with
5578 ENOMEM, for example. A temporary error is now given on failure.
5580 29. Contrary to the C standard, it seems that in some environments, the
5581 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5582 program. Exim now does this explicitly; it affects the formatting of
5583 timestamps using strftime().
5585 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5586 complaints. I've now initialized all the variables, to avoid this.
5588 32. Header lines added by a system filter were not being "seen" during
5589 transport-time rewrites.
5591 33. The info_callback() function passed to OpenSSL is set up with type void
5592 (*)(SSL *, int, int), as described somewhere. However, when calling the
5593 function (actually a macro) that sets it up, the type void(*)() is
5594 expected. I've put in a cast to prevent warnings from picky compilers.
5596 34. If a DNS black list lookup found a CNAME record, but there were no A
5597 records associated with the domain it pointed at, Exim crashed.
5599 35. If a DNS black list lookup returned more than one A record, Exim ignored
5600 all but the first. It now scans all returned addresses if a particular IP
5601 value is being sought. In this situation, the contents of the
5602 $dnslist_value variable are a list of all the addresses, separated by a
5605 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5606 to accept a host name and all its aliases if the forward lookup for any of
5607 them yielded the IP address of the incoming connection. Now it accepts only
5608 those names whose forward lookup yields the correct IP address. Any other
5609 names are discarded. This closes a loophole whereby a rogue DNS
5610 administrator could create reverse DNS records to break through a
5611 wildcarded host restriction in an ACL.
5613 37. If a user filter or a system filter that ran in a subprocess used any of
5614 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5615 the wrong values were passed to the pipe command ($thisaddress had the
5616 value of $0, $0 had the value of $1, etc). This bug was introduced by
5617 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5619 38. Improved the line breaking for long SMTP error messages from ACLs.
5620 Previously, if there was no break point between 40 and 75 characters, Exim
5621 left the rest of the message alone. Two changes have been made: (a) I've
5622 reduced the minimum length to 35 characters; (b) if it can't find a break
5623 point between 35 and 75 characters, it looks ahead and uses the first one
5624 that it finds. This may give the occasional overlong line, but at least the
5625 remaining text gets split now.
5627 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5628 file descriptors might be low, and that setting 1000 would always raise it.
5629 It turns out that in some environments, the limit is already over 1000 and
5630 that lowering it causes trouble. So now Exim takes care not to decrease it.
5632 40. When delivering a message, the value of $return_path is set to $sender_
5633 address at the start of routing (routers may change the value). By an
5634 oversight, this default was not being set up when an address was tested by
5635 -bt or -bv, which affected the outcome if any router or filter referred to
5638 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5639 log only when "message" or "log_message" are set. However, if one of the
5640 conditions was an address verification, or a call to a nested ACL, the
5641 messages generated by the underlying test were being passed through. This
5642 no longer happens. The underlying message is available in $acl_verify_
5643 message for both "message" and "log_message" expansions, so it can be
5644 passed through if needed.
5646 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5647 new expansion $bh_ to give the encoded byte string without charset
5648 translation. Translation happens only if iconv() is available; HAVE_ICONV
5649 indicates this at build time. HEADERS_CHARSET gives the charset to
5650 translate to; headers_charset can change it in the configuration, and
5651 "headers charset" can change it in an individual filter file.
5653 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5654 that creates RFC 2047 encoded "words" labels them as that charset instead
5655 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5656 expansion operator; (ii) when Exim creates a From: line for a local
5657 message; (iii) when a header line is rewritten to include a "phrase" part.
5659 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5660 buggy, causing it to skip the first lines of messages whose message ID
5661 ended in 'D'. This would not have bitten before Exim release 4.14, because
5662 message IDs were unlikely to end in 'D' before then. The effect was to have
5663 incorrect size information for certain domains.
5665 45. #include "config.h" was missing at the start of the crypt16.c module. This
5666 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5669 46. If there was a timeout during a "random" callout check, Exim treated it as
5670 a failure of the random address, and carried on sending RSET and the real
5671 address. If the delay was just some slowness somewhere, the response to the
5672 original RCPT would be taken as a response to RSET and so on, causing
5673 mayhem of various kinds.
5675 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5676 when I implemented it. It didn't allow for the fact that some option values
5677 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5678 the right test for positive values.
5680 48. Domain names in DNS records are case-independent. Exim always looks them up
5681 in lower case. Some resolvers return domain names in exactly the case they
5682 appear in the zone file, that is, they may contain uppercase letters. Not
5683 all resolvers do this - some return always lower case. Exim was treating a
5684 change of case by a resolver as a change of domain, similar to a widening
5685 of a domain abbreviation. This triggered its re-routing code and so it was
5686 trying to route what was effectively the same domain again. This normally
5687 caused routing to fail (because the router wouldn't handle the domain
5688 twice). Now Exim checks for this case specially, and just changes the
5689 casing of the domain that it ultimately uses when it transmits the message
5692 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5695 50. If a filter generated a file delivery with a non-absolute name (possible if
5696 no home directory exists for the router), the forbid_file option was not
5699 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5700 the existing equality matching.
5702 52. Exim was using ints instead of ino_t variables in some places where it was
5703 dealing with inode numbers.
5705 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5706 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5707 called TMPDIR, and if it finds it is different, it changes its value.
5709 54. The smtp_printf() function is now made available to local_scan() so
5710 additional output lines can be written before returning. There is also an
5711 smtp_fflush() function to enable the detection of a dropped connection.
5712 The variables smtp_input and smtp_batched_input are exported to
5715 55. Changed the default runtime configuration: the message "Unknown user"
5716 has been removed from the ACL, and instead placed on the localuser router,
5717 using the cannot_route_message feature. This means that any verification
5718 failures that generate their own messages won't get overridden. Similarly,
5719 the "Unrouteable address" message that was in the ACL for unverifiable
5720 relay addresses has also been removed.
5722 56. Added hosts_avoid_esmtp to the smtp transport.
5724 57. The exicyclog script was not checking for the esoteric option
5725 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5726 will work only if exicyclog is run under the appropriate euid.
5728 58. Following a discussion on the list, the rules by which Exim recognises line
5729 endings on incoming messages have been changed. The -dropcr and drop_cr
5730 options are now no-ops, retained only for backwards compatibility. The
5731 following line terminators are recognized: LF CRLF CR. However, special
5732 processing applies to CR:
5734 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5735 nor a local message in the state where . is a terminator.
5737 (ii) If a bare CR is encountered in a header line, an extra space is added
5738 after the line terminator so as not to end the header. The reasoning
5739 behind this is that bare CRs in header lines are most likely either
5740 to be mistakes, or people trying to play silly games.
5742 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5743 was being incorrectly given as 18 bytes larger than it should have been.
5744 This is a VOB (very old bug).
5746 60. This may never have affected anything current, but just in case it has:
5747 When the local host is found other than at the start of a list of hosts,
5748 the local host, those with the same MX, and any that follow, are discarded.
5749 When the list in question was part of a longer list of hosts, the following
5750 hosts (not currently being processed) were also being discarded. This no
5751 longer happens. I'm not sure if this situation could ever has previously
5754 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5757 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5758 header. This is something that is recommended in a new Internet Draft, and
5759 is something that is documented as being done by Sendmail. There are two
5760 possible values. For messages generated by the autoreply transport, Exim
5763 Auto-Submitted: auto-replied
5765 whereas for all other generated messages (e.g. bounces) it adds
5767 Auto-Submitted: auto-generated
5769 63. The "personal" condition in filters now includes a test for the
5770 Auto-Submitted: header. If it contains the string "auto-" the message it
5771 not considered personal.
5773 64. Added rcpt_include_affixes as a generic transport option.
5775 65. Added queue_only_override (default true).
5777 66. Added the syslog_duplication option.
5779 67. If what should have been the first header line of a message consisted of
5780 a space followed by a colon, Exim was mis-interpreting it as a header line.
5781 It isn't of course - it is syntactically invalid and should therefore be
5782 treated as the start of the message body. The misbehaviour could have
5783 caused a number of strange effects, including loss of data in subsequent
5784 header lines, and spool format errors.
5786 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5787 client host had authenticated. This control can now be exercised by an ACL
5788 for more flexibility.
5790 69. By default, callouts do not happen when testing with -bh. There is now a
5791 variant, -bhc, which does actually run the callout code, including
5792 consulting and updating the callout cache.
5794 70. Added support for saslauthd authentication, courtesy of Alexander
5797 71. If statvfs() failed on the spool or log directories while checking their
5798 size for availability, Exim confusingly gave the error "space shortage".
5799 Furthermore, in debugging mode it crashed with a floating point exception.
5800 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5801 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5802 serious problem, Exim now writes to the main and panic logs when this
5803 happens, with details of the failure. It then refuses to accept the
5804 incoming message, giving the message "spool directory problem" or "log
5805 directory problem" with a 421 code for SMTP messages.
5807 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5808 0, 1, and 2 exist, because some OS complain for execs without them (see
5809 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5810 descriptors. However, the code omitted to check that the open succeeded,
5811 causing mysterious errors if for some reason the permissions on /dev/null
5812 got screwed. Now Exim writes a message to the main and panic logs, and
5813 bombs out if it can't open /dev/null.
5815 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5816 interact so that it is all more flexible. It is supposed to remain
5817 backwards compatible. Also added extra_local_interfaces.
5819 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5820 to bomb out with an assertion failure - to the client this appears as a
5821 connection drop. This problem occurs in the part of the code that was taken
5822 from the Samba project. Fortunately, the assertion is in a very simple
5823 function, so I have fixed this by reproducing the function inline in the
5824 one place where it is called, and arranging for authentication to fail
5825 instead of killing the process with assert().
5827 75. The SPA client code was not working when the server requested OEM rather
5828 than Unicode encoding.
5830 76. Added code to make require_files with a specific uid setting more usable in
5831 the case where statting the file as root fails - usually a non-root-mounted
5832 NFS file system. When this happens and the failure is EACCES, Exim now
5833 forks a subprocess and does the per-uid checking as the relevant uid.
5835 77. Added process_log_path.
5837 78. If log_file_path was not explicitly set, a setting of check_log_space or
5838 check_log_inodes was ignored.
5840 79. If a space check for the spool or log partitions fails, the incident is now
5841 logged. Of course, in the latter case the data may get lost...
5843 80. Added the %p formatting code to string_format() so that it can be used to
5844 print addresses in debug_print(). Adjusted all the address printing in the
5845 debugging in store.c to use %p rather than %d.
5847 81. There was a concern that a line of code in smtp_in.c could overflow a
5848 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5849 initially expressed, the concern was not well-founded, because trailing
5850 spaces are removed early. However, if the trailing spaces were followed by
5851 a NULL, they did not get removed, so the overflow was possible. Two fixes
5854 (a) I re-wrote the offending code in a cleaner fashion.
5855 (b) If an incoming SMTP command contains a NULL character, it is rejected
5858 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5859 runs initgroups(), so that if the Exim user is in any additional groups,
5860 they will be used during message reception.
5866 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob