1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
9 PP/02 Make -n do something, by making it not do something.
10 When combined with -bP, the name of an option is not output.
12 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
15 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
16 $sender_host_name and config options to manage this, and basic check
19 PP/05 DSCP support for outbound connections and control modifier for inbound.
21 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
22 (Only plugin which currently uses this is kerberos4, which nobody should
23 be using, but we should make it available and other future plugins might
24 conceivably use it, even though it would break NAT; stuff *should* be
25 using channel bindings instead).
27 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
28 name; added for Sendmail compatibility; requires admin caller.
29 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
30 just ignore it); requires trusted caller.
31 Also parse but ignore: -Ac -Am -X<logfile>
34 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
36 TL/02 Add +smtp_confirmation as a default logging option.
38 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
39 Patch by Magnus Holmgren from 2007-02-20.
41 TL/04 Bugzilla 1281 - Spec typo.
42 Bugzilla 1283 - Spec typo.
43 Bugzilla 1290 - Spec grammar fixes.
45 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
47 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
49 JH/02 Support "G" suffix to numbers in ${if comparisons.
51 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
53 NM/01 Bugzilla 1197 - Spec typo
54 Bugzilla 1196 - Spec examples corrections
56 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
58 PP/09 Add gnutls_enable_pkcs11 option.
60 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
61 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
63 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
64 "acl {{name}{arg}...}", and optional args on acl condition
67 JH/05 Permit multiple router/transport headers_add/remove lines.
69 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
71 JH/07 Avoid using a waiting database for a single-message-only transport.
72 Performance patch from Paul Fisher. Bugzilla 1262.
74 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
77 JH/09 Add $headers_added variable, with content from use of ACL modifier
78 add_header (but not yet added to the message). Bugzilla 199.
80 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
81 Pulled from Bugzilla 817 by Wolfgang Breyha.
83 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
86 JH/11 Add A= logging on delivery lines, and a client_set_id option on
89 JH/12 Add optional authenticated_sender logging to A= and a log_selector
92 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
94 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
95 advertise SMTP AUTH mechanism to us, instead of a generic
96 protocol violation error. Also, make Exim more robust to bad
97 data from the Dovecot auth socket.
99 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
101 When a queue runner is handling a message, Exim first routes the
102 recipient addresses, during which it prunes them based on the retry
103 hints database. After that it attempts to deliver the message to
104 any remaining recipients. It then updates the hints database using
107 So if a recipient address works intermittently, it can get repeatedly
108 deferred at routing time. The retry hints record remains fresh so the
109 address never reaches the final cutoff time.
111 This is a fairly common occurrence when a user is bumping up against
112 their storage quota. Exim had some logic in its local delivery code
113 to deal with this. However it did not apply to per-recipient defers
114 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
116 This change adds a proper retry rule check during routing so that the
117 final cutoff time is checked against the message's age. We only do
118 this check if there is an address retry record and there is not a
119 domain retry record; this implies that previous attempts to handle
120 the address had the retry_use_local_parts option turned on. We use
121 this as an approximation for the destination being like a local
122 delivery, as in LMTP.
124 I suspect this new check makes the old local delivery cutoff check
125 redundant, but I have not verified this so I left the code in place.
127 TF/02 Correct gecos expansion when From: is a prefix of the username.
129 Test 0254 submits a message to Exim with the header
133 When I ran the test suite under the user fanf2, Exim expanded
134 the header to contain my full name, whereas it should have added
135 a Resent-Sender: header. It erroneously treats any prefix of the
136 username as equal to the username.
138 This change corrects that bug.
140 GF/01 DCC debug and logging tidyup
141 Error conditions log to paniclog rather than rejectlog.
142 Debug lines prefixed by "DCC: " to remove any ambiguity.
144 TF/03 Avoid unnecessary rebuilds of lookup-related code.
150 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
152 This, or similar/improved, will also be change PP/11 of 4.82.
158 PP/01 Handle short writes when writing local log-files.
159 In practice, only affects FreeBSD (8 onwards).
160 Bugzilla 1053, with thanks to Dmitry Isaikin.
162 NM/01 Bugzilla 949 - Documentation tweak
164 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
167 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
169 PP/02 Implemented gsasl authenticator.
171 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
173 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
174 `pkg-config foo` for cflags/libs.
176 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
177 with rest of GSASL and with heimdal_gssapi.
179 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
180 `pkg-config foo` for cflags/libs for the TLS implementation.
182 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
183 properties get this fed in as external SSF. A number of robustness
184 and debugging improvements to the cyrus_sasl authenticator.
186 PP/08 cyrus_sasl server now expands the server_realm option.
188 PP/09 Bugzilla 1214 - Log authentication information in reject log.
189 Patch by Jeremy Harris.
191 PP/10 Added dbmjz lookup type.
193 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
195 PP/12 MAIL args handles TAB as well as SP, for better interop with
196 non-compliant senders.
197 Analysis and variant patch by Todd Lyons.
199 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
200 Bug report from Lars Müller <lars@samba.org> (via SUSE),
201 Patch from Dirk Mueller <dmueller@suse.com>
203 PP/13 tls_peerdn now print-escaped for spool files.
204 Observed some $tls_peerdn in wild which contained \n, which resulted
205 in spool file corruption.
207 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
208 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
209 or write after TLS renegotiation, which otherwise led to messages
212 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
213 as a tracking header (ie: a signed header comes before the signature).
214 Patch from Wolfgang Breyha.
216 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
217 comma-sep list; embedded commas doubled.
219 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
221 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
223 Report and patch from Dmitry Banschikov.
225 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
226 Removed SSL_clear() after SSL_new() which led to protocol negotiation
227 failures. We appear to now support TLS1.1+ with Exim.
229 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
230 lets Exim select keys and certificates based upon TLS SNI from client.
231 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
232 before an outbound SMTP session. New log_selector, +tls_sni.
234 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
235 NULL dereference. Report and patch from Alun Jones.
237 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
238 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
239 Not seeing resolver debug output on NetBSD, but suspect this is a
240 resolver implementation change.
242 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
243 Left warnings. Added "eximon gdb" invocation mode.
245 PP/21 Defaulting "accept_8bitmime" to true, not false.
247 PP/22 Added -bw for inetd wait mode support.
249 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
250 locate the relevant includes and libraries. Made this the default.
252 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
253 Bugzilla 1246, report and most of solution from Tomasz Kusy.
255 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
256 This may cause build issues on older platforms.
258 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
259 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
260 gnutls_require_mac & gnutls_require_protocols (no longer supported).
261 Added SNI support via GnuTLS too.
262 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
264 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
266 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
267 Applied second patch from Janne, implementing suggestion to default
268 multiple-strings-in-record handling to match SPF spec.
270 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
272 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
273 read-only, out of scope).
274 Patch from Wolfgang Breyha, report from Stuart Northfield.
276 PP/29 Fix three issues highlighted by clang analyser static analysis.
277 Only crash-plausible issue would require the Cambridge-specific
278 iplookup router and a misconfiguration.
279 Report from Marcin Mirosław.
281 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
283 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
284 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
285 As part of this, removing so much warning spew let me fix some minor
286 real issues in debug logging.
288 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
289 assignment on my part. Fixed.
291 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
292 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
293 Janne Snabb (who went above and beyond: thank you).
295 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
296 string otherwise requires a connection and a bunch more work and it's
297 relatively easy to get wrong. Should also expose TLS library linkage
300 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
301 64-bit ${eval} (JH/03).
303 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
304 GNU libc to support some of the 64-bit stuff, should not lead to
305 conflicts. Defined before os.h is pulled in, so if a given platform
306 needs to override this, it can.
308 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
309 protection layer was required, which is not implemented.
310 Bugzilla 1254, patch from Wolfgang Breyha.
312 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
313 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
314 tls_dhparam take prime identifiers. Also unbreak combination of
315 OpenSSL+DH_params+TLSSNI.
317 PP/39 Disable SSLv2 by default in OpenSSL support.
323 PP/01 Solaris build fix for Oracle's LDAP libraries.
324 Bugzilla 1109, patch from Stephen Usher.
326 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
328 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
331 TF/02 Fix a couple more cases where we did not log the error message
332 when unlink() failed. See also change 4.74-TF/03.
334 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
335 lock up or crash if it happened to be inside a call to libc when it
336 got a SIGUSR1 from exiwhat.
338 The SIGUSR1 handler appends the current process status to the process
339 log which is later printed by exiwhat. It used to use the general
340 purpose logging code to do this, but several functions it calls are
341 not safe for signals.
343 The new output code in the SIGUSR1 handler is specific to the process
344 log, and simple enough that it's easy to inspect for signal safety.
345 Removing some special cases also simplifies the general logging code.
346 Removing the spurious timestamps from the process log simplifies
349 TF/04 Improved ratelimit ACL condition.
351 The /noupdate option has been deprecated in favour of /readonly which
352 has clearer semantics. The /leaky, /strict, and /readonly update modes
353 are mutually exclusive. The update mode is no longer included in the
354 database key; it just determines when the database is updated. (This
355 means that when you upgrde Exim will forget old rate measurements.)
357 Exim now checks that the per_* options are used with an update mode that
358 makes sense for the current ACL. For example, when Exim is processing a
359 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
360 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
361 must specify per_mail/readonly. If you omit the update mode it defaults to
362 /leaky where that makes sense (as before) or /readonly where required.
364 The /noupdate option is now undocumented but still supported for
365 backwards compatibility. It is equivalent to /readonly except that in
366 ACLs where /readonly is required you may specify /leaky/noupdate or
367 /strict/noupdate which are treated the same as /readonly.
369 A useful new feature is the /count= option. This is a generalization
370 of the per_byte option, so that you can measure the throughput of other
371 aggregate values. For example, the per_byte option is now equivalent
372 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
374 The per_rcpt option has been generalized using the /count= mechanism
375 (though it's more complicated than the per_byte equivalence). When it is
376 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
377 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
378 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
379 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
380 Note that using per_rcpt with a non-readonly update mode in more than
381 one ACL will cause the recipients to be double-counted. (The per_mail
382 and per_byte options don't have this problem.)
384 The handling of very low rates has changed slightly. If the computed rate
385 is less than the event's count (usually one) then this event is the first
386 after a long gap. In this case the rate is set to the same as this event's
387 count, so that the first message of a spam run is counted properly.
389 The major new feature is a mechanism for counting the rate of unique
390 events. The new per_addr option counts the number of different
391 recipients that someone has sent messages to in the last time period. It
392 behaves like per_rcpt if all the recipient addresses are different, but
393 duplicate recipient addresses do not increase the measured rate. Like
394 the /count= option this is a general mechanism, so the per_addr option
395 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
396 example, measure the rate that a client uses different sender addresses
397 with the options per_mail/unique=$sender_address. There are further
398 details in the main documentation.
400 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
402 TF/06 Removed a few PCRE remnants.
404 TF/07 Automatically extract Exim's version number from tags in the git
405 repository when doing development or release builds.
407 PP/02 Raise smtp_cmd_buffer_size to 16kB.
408 Bugzilla 879. Patch from Paul Fisher.
410 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
411 Heavily based on revision 40f9a89a from Simon Arlott's tree.
414 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
416 PP/05 Variable $av_failed, true if the AV scanner deferred.
417 Bugzilla 1078. Patch from John Horne.
419 PP/06 Stop make process more reliably on build failure.
420 Bugzilla 1087. Patch from Heiko Schlittermann.
422 PP/07 Make maildir_use_size_file an _expandable_ boolean.
423 Bugzilla 1089. Patch from Heiko Schlittermann.
425 PP/08 Handle ${run} returning more data than OS pipe buffer size.
426 Bugzilla 1131. Patch from Holger Weiß.
428 PP/09 Handle IPv6 addresses with SPF.
429 Bugzilla 860. Patch from Wolfgang Breyha.
431 PP/10 GnuTLS: support TLS 1.2 & 1.1.
433 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
436 PP/11 match_* no longer expand right-hand-side by default.
437 New compile-time build option, EXPAND_LISTMATCH_RHS.
438 New expansion conditions, "inlist", "inlisti".
440 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
442 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
444 PP/14 fix log_write() format string regression from TF/03.
445 Bugzilla 1152. Patch from Dmitry Isaikin.
451 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
453 PP/02 Harmonised TLS library version reporting; only show if debugging.
454 Layout now matches that introduced for other libraries in 4.74 PP/03.
456 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
458 PP/04 New "dns_use_edns0" global option.
460 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
463 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
464 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
466 TK/01 Updated PolarSSL code to 0.14.2.
467 Bugzilla 1097. Patch from Andreas Metzler.
469 PP/07 Catch divide-by-zero in ${eval:...}.
472 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
475 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
476 format-string attack -- SECURITY: remote arbitrary code execution.
478 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
479 time unintentionally subject to list matching rules, letting the header
480 cause arbitrary Exim lookups (of items which can occur in lists, *not*
481 arbitrary string expansion). This allowed for information disclosure.
483 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
484 INT_MIN/-1 -- value coerced to INT_MAX.
490 NM/01 Workround for PCRE version dependency in version reporting
493 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
494 This fixes portability to compilers other than gcc, notably
495 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
497 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
498 makefiles for portability to HP-UX and POSIX correctness.
500 PP/01 Permit LOOKUP_foo enabling on the make command-line.
501 Also via indented variable definition in the Makefile.
502 (Debugging by Oliver Heesakkers).
504 PP/02 Restore caching of spamd results with expanded spamd_address.
505 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
507 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
508 Improves build reliability. Fix from: Frank Elsner
510 NM/02 Fix wide character breakage in the rfc2047 coding
511 Fixes bug 1064. Patch from Andrey N. Oktyabrski
513 NM/03 Allow underscore in dnslist lookups
514 Fixes bug 1026. Patch from Graeme Fowler
516 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
517 Code patches from Adam Ciarcinski of NetBSD.
519 NM/04 Fixed exiqgrep to cope with mailq missing size issue
522 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
523 is logged, to avoid truncation. Patch from John Horne.
525 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
526 Patch from Jakob Hirsch.
528 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
529 SQL string expansion failure details.
530 Patch from Andrey Oktyabrski.
532 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
533 Patch from Simon Arlott.
535 PP/09 New lookups functionality failed to compile on old gcc which rejects
536 extern declarations in function scope.
537 Patch from Oliver Fleischmann
539 PP/10 Use sig_atomic_t for flags set from signal handlers.
540 Check getgroups() return and improve debugging.
541 Fixed developed for diagnosis in bug 927 (which turned out to be
544 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
545 Patch from Mark Zealey.
547 PP/12 Bugzilla 1056: Improved spamd server selection.
548 Patch from Mark Zealey.
550 PP/13 Bugzilla 1086: Deal with maildir quota file races.
551 Based on patch from Heiko Schlittermann.
553 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
554 Patch from Uwe Doering, sign-off by Michael Haardt.
556 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
557 variable declaration deep within a block. Bug and patch from
560 PP/15 lookups-Makefile IRIX compatibilty coercion.
562 PP/16 Make DISABLE_DKIM build knob functional.
564 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
565 Patch by Simon Arlott
567 TF/03 Fix valgrind.h portability to C89 compilers that do not support
568 variable argument macros. Our copy now differs from upstream.
574 TF/01 Failure to get a lock on a hints database can have serious
575 consequences so log it to the panic log.
577 TF/02 Log LMTP confirmation messages in the same way as SMTP,
578 controlled using the smtp_confirmation log selector.
580 TF/03 Include the error message when we fail to unlink a spool file.
582 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
583 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
584 for maintaining out-of-tree patches for some time.
586 PP/01 Bugzilla 139: Documentation and portability issues.
587 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
588 Handle per-OS dynamic-module compilation flags.
590 PP/02 Let /dev/null have normal permissions.
591 The 4.73 fixes were a little too stringent and complained about the
592 permissions on /dev/null. Exempt it from some checks.
593 Reported by Andreas M. Kirchwitz.
595 PP/03 Report version information for many libraries, including
596 Exim version information for dynamically loaded libraries. Created
597 version.h, now support a version extension string for distributors
598 who patch heavily. Dynamic module ABI change.
600 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
601 privilege escalation vulnerability whereby the Exim run-time user
602 can cause root to append content of the attacker's choosing to
605 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
608 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
609 If dropping privileges for untrusted macros, we disabled normal logging
610 on the basis that it would fail; for the Exim run-time user, this is not
611 the case, and it resulted in successful deliveries going unlogged.
612 Fixed. Reported by Andreas Metzler.
618 PP/01 Date: & Message-Id: revert to normally being appended to a message,
619 only prepend for the Resent-* case. Fixes regression introduced in
620 Exim 4.70 by NM/22 for Bugzilla 607.
622 PP/02 Include check_rfc2047_length in configure.default because we're seeing
623 increasing numbers of administrators be bitten by this.
625 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
627 PP/03 Bugzilla 994: added openssl_options main configuration option.
629 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
631 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
633 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
635 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
636 without a peer certificate, leading to a segfault because of an
637 assumption that peers always have certificates. Be a little more
638 paranoid. Problem reported by Martin Tscholak.
640 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
641 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
642 NB: ClamAV planning to remove STREAM in "middle of 2010".
643 CL also introduces -bmalware, various -d+acl logging additions and
644 more caution in buffer sizes.
646 PP/09 Implemented reverse_ip expansion operator.
648 PP/10 Bugzilla 937: provide a "debug" ACL control.
650 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
652 PP/12 Bugzilla 973: Implement --version.
654 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
656 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
658 PP/15 Bugzilla 816: support multiple condition rules on Routers.
660 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
661 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
662 ignore trailing whitespace.
664 JJ/02 prevent non-panic DKIM error from being sent to paniclog
666 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
669 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
670 Notification from Dr Andrew Aitchison.
672 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
673 ExtendedDetectionInfo response format.
674 Notification from John Horne.
676 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
679 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
680 XSL and documented dependency on system catalogs, with examples of how
683 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
686 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
687 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
688 configuration file which is writeable by the Exim user or group.
690 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
691 of configuration files to cover files specified with the -C option if
692 they are going to be used with root privileges, not just the default
695 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
696 option (effectively making it always true).
698 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
699 files to be used while preserving root privileges.
701 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
702 that rogue child processes cannot use them.
704 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
705 run-time user, instead of root.
707 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
708 Exim run-time user without dropping privileges.
710 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
711 result string, instead of calling string_vformat() twice with the same
714 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
715 for other users. Others should always drop root privileges if they use
716 -C on the command line, even for a whitelisted configure file.
718 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
720 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
726 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
727 $data_path, and $header_path variables; fixed documentation bugs and
730 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
731 exipick to access non-standard spools, including the "frozen" queue
734 NM/01 Bugzilla 965: Support mysql stored procedures.
735 Patch from Alain Williams
737 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
739 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
740 Patch from Andreas Metzler
742 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
743 Patch from Kirill Miazine
745 NM/05 Bugzilla 671: Added umask to procmail example.
747 JJ/03 installed exipick 20100323.0, fixing doc bug
749 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
750 directory. Notification and patch from Dan Rosenberg.
752 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
754 TK/02 Improve log output when DKIM signing operation fails.
756 MH/01 Treat the transport option dkim_domain as a colon separated
757 list, not as a single string, and sign the message with each element,
758 omitting multiple occurences of the same signer.
760 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
761 Bugzilla 985, 986. Patch by Simon Arlott
763 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
764 Patch by Simon Arlott
766 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
767 MBX locking. Notification from Dan Rosenberg.
773 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
775 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
777 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
779 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
781 NM/04 Bugzilla 915: Flag broken perl installation during build.
787 TK/01 Added patch by Johannes Berg that expands the main option
788 "spamd_address" if it starts with a dollar sign.
790 TK/02 Write list of recipients to X-Envelope-Sender header when building
791 the mbox-format spool file for content scanning (suggested by Jakob
794 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
795 (http://www.dcc-servers.net/) support via dccifd. Activated by
796 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
798 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
799 by Mark Daniel Reidel <mr@df.eu>.
801 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
802 When building exim an external PCRE library is now needed -
803 PCRE is a system library on the majority of modern systems.
804 See entry on PCRE_LIBS in EDITME file.
806 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
807 conversation. Added nologin parameter to request.
808 Patch contributed by Kirill Miazine.
810 TF/01 Do not log submission mode rewrites if they do not change the address.
812 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
814 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
815 log files in place. Contributed by Roberto Lima.
817 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
819 TF/03 Bugzilla 615: When checking the local_parts router precondition
820 after a local_part_suffix or local_part_prefix option, Exim now
821 does not use the address's named list lookup cache, since this
822 contains cached lookups for the whole local part.
824 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
825 Robert Millan. Documentation is in experimental-spec.txt.
827 TF/04 Bugzilla 668: Fix parallel build (make -j).
829 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
831 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
832 Patch provided by Jan Srzednicki.
834 TF/05 Leading white space used to be stripped from $spam_report which
835 wrecked the formatting. Now it is preserved.
837 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
838 that they are available at delivery time.
840 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
842 TF/08 TLS error reporting now respects the incoming_interface and
843 incoming_port log selectors.
845 TF/09 Produce a more useful error message if an SMTP transport's hosts
846 setting expands to an empty string.
848 NM/06 Bugzilla 744: EXPN did not work under TLS.
849 Patch provided by Phil Pennock.
851 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
852 Patch provided by Richard Godbee.
854 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
855 acl_smtp_notquit, added index entry.
857 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
858 Patch provided by Eugene Bujak.
860 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
861 accept(). Patch provided by Maxim Dounin.
863 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
864 Patch provided by Phil Pennock.
866 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
868 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
869 Patch provided by Brad "anomie" Jorsch.
871 NM/14 Bugzilla 622: Added timeout setting to transport filter.
872 Patch provided by Dean Brooks.
874 TK/05 Add native DKIM support (does not depend on external libraries).
876 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
877 Patch provided by Graeme Fowler.
879 NM/16 Bugzilla 851: Documentation example syntax fix.
881 NM/17 Changed NOTICE file to remove references to embedded PCRE.
883 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
886 NM/19 Bugzilla 745: TLS version reporting.
887 Patch provided by Phil Pennock.
889 NM/20 Bugzilla 167: bool: condition support.
890 Patch provided by Phil Pennock.
892 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
893 clients. Patch provided by Phil Pennock.
895 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
896 Patch provided by Brad "anomie" Jorsch.
898 NM/23 Bugzilla 687: Fix misparses in eximstats.
899 Patch provided by Heiko Schlittermann.
901 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
902 Patch provided by Heiko Schlittermann.
904 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
905 plus update to original patch.
907 NM/26 Bugzilla 799: Documentation correction for ratelimit.
909 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
910 Patch provided by David Brownlee.
912 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
914 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
916 NM/30 Bugzilla 888: TLS documentation bugfixes.
918 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
920 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
921 Unlike the original bugzilla I have changed all shell scripts in src tree.
923 NM/33 Bugzilla 898: Transport filter timeout fix.
924 Patch by Todd Rinaldo.
926 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
927 Patch by Serge Demonchaux.
929 NM/35 Bugzilla 39: Base64 decode bug fixes.
930 Patch by Jakob Hirsch.
932 NM/36 Bugzilla 909: Correct connect() call in dcc code.
934 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
936 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
938 NM/39 Bugzilla 911: Fixed MakeLinks build script.
944 TK/01 Add preliminary DKIM support. Currently requires a forked version of
945 ALT-N's libdkim that I have put here:
946 http://duncanthrax.net/exim-experimental/
948 Note to Michael Haardt: I had to rename some vars in sieve.c. They
949 were called 'true' and it seems that C99 defines that as a reserved
950 keyword to be used with 'bool' variable types. That means you could
951 not include C99-style headers which use bools without triggering
952 build errors in sieve.c.
954 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
955 as mailq or other aliases. Changed the --help handling significantly
956 to do whats expected. exim_usage() emits usage/help information.
958 SC/01 Added the -bylocaldomain option to eximstats.
960 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
962 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
964 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
970 PH/01 Another patch from the Sieve maintainer.
972 PH/02 When an IPv6 address is converted to a string for single-key lookup
973 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
974 dots are used instead of colons so that keys in lsearch files need not
975 contain colons. This was done some time before quoting was made available
976 in lsearch files. However, iplsearch files do require colons in IPv6 keys
977 (notated using the quote facility) so as to distinguish them from IPv4
978 keys. This meant that lookups for IP addresses in host lists did not work
979 for iplsearch lookups.
981 This has been fixed by arranging for IPv6 addresses to be expressed with
982 colons if the lookup type is iplsearch. This is not incompatible, because
983 previously such lookups could never work.
985 The situation is now rather anomolous, since one *can* have colons in
986 ordinary lsearch keys. However, making the change in all cases is
987 incompatible and would probably break a number of configurations.
989 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
992 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
993 conversion specification without a maximum field width, thereby enabling
994 a rogue spamd server to cause a buffer overflow. While nobody in their
995 right mind would setup Exim to query an untrusted spamd server, an
996 attacker that gains access to a server running spamd could potentially
997 exploit this vulnerability to run arbitrary code as the Exim user.
999 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1000 $primary_hostname instead of what libspf2 thinks the hosts name is.
1002 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1003 a directory entry by the name of the lookup key. Previously, if a
1004 symlink pointed to a non-existing file or a file in a directory that
1005 Exim lacked permissions to read, a lookup for a key matching that
1006 symlink would fail. Now it is enough that a matching directory entry
1007 exists, symlink or not. (Bugzilla 503.)
1009 PH/03 The body_linecount and body_zerocount variables are now exported in the
1012 PH/04 Added the $dnslist_matched variable.
1014 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1015 This means they are set thereafter only if the connection becomes
1018 PH/06 Added the client_condition to authenticators so that some can be skipped
1019 by clients under certain conditions.
1021 PH/07 The error message for a badly-placed control=no_multiline_responses left
1022 "_responses" off the end of the name.
1024 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1026 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1027 (without spaces) instead of just copying the configuration text.
1029 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1031 PH/11 Added $max_received_linelength.
1033 PH/12 Added +ignore_defer and +include_defer to host lists.
1035 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1036 way in which PCRE > 7.0 is built.
1038 PH/14 Implemented queue_only_load_latch.
1040 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1041 MAIL command. The effect was to mangle the value on 64-bit systems.
1043 PH/16 Another patch from the Sieve maintainer.
1045 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1047 PH/18 If a system quota error occurred while trying to create the file for
1048 a maildir delivery, the message "Mailbox is full" was not appended to the
1049 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1050 only to a quota excession during the actual writing of the file.
1052 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1053 characters?) which causes problems in log lines. The DN values are now
1054 passed through string_printing() before being added to log lines.
1056 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1057 and InterBase are left for another time.)
1059 PH/21 Added message_body_newlines option.
1061 PH/22 Guard against possible overflow in moan_check_errorcopy().
1063 PH/23 POSIX allows open() to be a macro; guard against that.
1065 PH/24 If the recipient of an error message contained an @ in the local part
1066 (suitably quoted, of course), incorrect values were put in $domain and
1067 $local_part during the evaluation of errors_copy.
1073 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1074 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1077 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1078 issue a MAIL command.
1080 PH/02 In an ACL statement such as
1082 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1084 if a client was not listed at all, or was listed with a value other than
1085 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1086 the condition was not true (as it should be), so access was not denied.
1087 The bug was that the ! inversion was incorrectly passed on to the second
1088 item. This has been fixed.
1090 PH/03 Added additional dnslists conditions == and =& which are different from
1091 = and & when the dns lookup returns more than one IP address.
1093 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1094 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1096 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1097 FSYNC, which compiles an option called disable_fsync that allows for
1098 bypassing fsync(). The documentation is heavily laced with warnings.
1100 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1102 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1103 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1104 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1105 including adding "make clean"; (3) Added -fPIC when compiling the test
1106 dynamically loaded module, to get rid of a warning.
1108 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1109 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1110 = 0s. The bug is otherwise harmless.
1112 PH/07 There was a bug in the dovecot authenticator such that the value of
1113 $auth1 could be overwritten, and so not correctly preserved, after a
1114 successful authentication. This usually meant that the value preserved by
1115 the server_setid option was incorrect.
1117 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1119 PH/09 Installed PCRE release 7.0.
1121 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1122 run for batched SMTP input. It is now run at the start of every message
1123 in the batch. While fixing this I discovered that the process information
1124 (output by running exiwhat) was not always getting set for -bs and -bS
1125 input. This is fixed, and it now also says "batched" for BSMTP.
1127 PH/11 Added control=no_pipelining.
1129 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1130 patch, slightly modified), and move the expansion of helo_data till after
1131 the connection is made in the smtp transport (so it can use these
1134 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1136 PH/14 Added log_selector = +pid.
1138 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1140 PH/16 Add ${if forany and ${if forall.
1142 PH/17 Added dsn_from option to vary the From: line in DSNs.
1144 PH/18 Flush SMTP output before performing a callout, unless control =
1145 no_callout_flush is set.
1147 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1148 was true (the default) a successful delivery failed to delete the retry
1149 item, thus causing premature timeout of the address. The bug is now
1152 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1154 PH/21 Long custom messages for fakedefer and fakereject are now split up
1155 into multiline reponses in the same way that messages for "deny" and
1156 other ACL rejections are.
1158 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1159 with slight modification.
1161 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1162 draft, changing the syntax and factoring some duplicate code".
1164 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1165 for deliveries of the second and subsequent messages over the same SMTP
1168 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1169 ${reduce, with only minor "tidies".
1171 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1173 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1174 expansion side effects.
1176 PH/27 When a message times out after an over-quota error from an Exim-imposed
1177 quota, the bounce message says "mailbox is full". This message was not
1178 being given when it was a system quota that was exceeded. It now should
1181 MH/03 Made $recipients available in local_scan(). local_scan() already has
1182 better access to the recipient list through recipients_list[], but
1183 $recipients can be useful in postmaster-provided expansion strings.
1185 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1186 in the case of a MAIL command with additional options following the
1187 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1188 were accidentally chopped off.
1190 PH/29 SMTP synchronization checks are implemented when a command is read -
1191 there is a check that no more input is waiting when there shouldn't be
1192 any. However, for some commands, a delay in an ACL can mean that it is
1193 some time before the response is written. In this time, more input might
1194 arrive, invalidly. So now there are extra checks after an ACL has run for
1195 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1196 pipelining has not been advertised.
1198 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1200 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1201 returned in the SMTP response when smtp_return_error_details was false.
1202 This has been fixed.
1204 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1205 instead of the C I/O that was originally supplied, because problems were
1206 reported on Solaris.
1208 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1209 Exim which did not show up earlier: it was assuming that a call to
1210 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1211 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1212 was a macro that became an assignment, so it seemed to work. This has
1213 changed to a proper function call with a void return, hence the compile
1214 error. Exim's code has been fixed.
1216 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1219 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1221 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1223 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1224 because it turns out to be too big or there is a timeout) there is no
1225 "Completed" line in the log. When some messages of this type were
1226 selected by exigrep, they were listed as "not completed". Others were
1227 picked up by some special patterns. I have improved the selection
1228 criteria to be more general.
1230 PH/38 The host_find_failed option in the manualroute router can now be set
1231 to "ignore", to completely ignore a host whose IP address cannot be
1232 found. If all hosts are ignored, the behaviour is controlled by the new
1233 host_all_ignored option.
1235 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1236 homing or because of multiple MX records with /mx) generated more than
1237 one IP address, and the following item turned out to be the local host,
1238 all the secondary addresses of the first item were incorrectly removed
1239 from the list, along with the local host and any following hosts (which
1240 is what is supposed to happen).
1242 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1243 whoever called Exim into the -H file. In the case of the daemon it was
1244 behaving confusingly. When first started, it used values for whoever
1245 started the daemon, but after a SIGHUP it used the Exim user (because it
1246 calls itself on a restart). I have changed the code so that it now always
1249 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1250 message are rejected with the same error (e.g. no authentication or bad
1251 sender address), and a DATA command is nevertheless sent (as can happen
1252 with PIPELINING or a stupid MUA), the error message that was given to the
1253 RCPT commands is included in the rejection of the DATA command. This is
1254 intended to be helpful for MUAs that show only the final error to their
1257 PH/42 Another patch from the Sieve maintainer.
1259 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1260 Eximstats - Fixed some broken HTML links and added missing column headers
1262 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1263 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1265 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1267 PH/43 Yet another patch from the Sieve maintainer.
1269 PH/44 I found a way to check for a TCP/IP connection going away before sending
1270 the response to the final '.' that terminates a message, but only in the
1271 case where the client has not sent further data following the '.'
1272 (unfortunately, this is allowed). However, in many cases there won't be
1273 any further data because there won't be any more messages to send. A call
1274 to select() can be used: if it shows that the input is "ready", there is
1275 either input waiting, or the socket has been closed. An attempt to read
1276 the next input character can distinguish the two cases. Previously, Exim
1277 would have sent an OK response which the client would never have see.
1278 This could lead to message repetition. This fix should cure that, at
1279 least in a lot of common cases.
1281 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1282 advertised in response to EHLO.
1288 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1289 fixed by 4.65/MH/01 (is this a record?) are fixed:
1291 (i) An empty string was always treated as zero by the numeric comparison
1292 operators. This behaviour has been restored.
1294 (ii) It is documented that the numeric comparison operators always treat
1295 their arguments as decimal numbers. This was broken in that numbers
1296 starting with 0 were being interpreted as octal.
1298 While fixing these problems I realized that there was another issue that
1299 hadn't been noticed. Values of message_size_limit (both the global option
1300 and the transport option) were treated as octal if they started with 0.
1301 The documentation was vague. These values are now always treated as
1302 decimal, and I will make that clear in the documentation.
1308 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1309 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1312 MH/01 Don't check that the operands of numeric comparison operators are
1313 integers when their expansion is in "skipping" mode (fixes bug
1314 introduced by 4.64-PH/07).
1316 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1317 child addresses, Exim now panics and dies. Previously, because the count
1318 is held in a short int, deliveries were likely to be lost. As such a
1319 large number of recipients for a single message is ridiculous
1320 (performance will be very, very poor), I have chosen to impose a limit
1321 rather than extend the field.
1327 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1328 leftover -K file (the existence of which was triggered by #402).
1329 While we were at it, introduced process PID as part of the -K
1330 filename. This should rule out race conditions when creating
1333 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1334 processing considerably. Previous code took too long for large mails,
1335 triggering a timeout which in turn triggers #401.
1337 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1338 in the DK code in transports.c. sendfile() is not really portable,
1339 hence the _LINUX specificness.
1341 TF/01 In the add_headers option to the mail command in an Exim filter,
1342 there was a bug that Exim would claim a syntax error in any
1343 header after the first one which had an odd number of characters
1346 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1347 callout verification, Exim cached a "reject" for the entire domain. This
1348 is correct for most verifications, but it is not correct for a recipient
1349 verification with use_sender or use_postmaster set, because in that case
1350 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1351 case of MAIL FROM:<> rejection from other early rejections (e.g.
1352 rejection of HELO). When verifying a recipient using a non-null MAIL
1353 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1354 Whatever the result of the callout, the value of the domain cache is
1355 left unchanged (for any other kind of callout, getting as far as trying
1356 RCPT means that the domain itself is ok).
1358 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1361 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1362 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1363 session, the daemon ignores ECONNECT errors and logs others; it now
1364 ignores EPIPE as well.
1366 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1367 (quoted-printable decoding).
1369 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1370 later the small subsequent patch to fix an introduced bug.
1372 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1374 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1376 PH/08 An error is now given if message_size_limit is specified negative.
1378 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1379 to be given (somewhat) arbitrary names.
1381 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1384 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1385 miscellaneous code fixes
1387 PH/10 Added the log_reject_target ACL modifier to specify where to log
1390 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1391 hostname. This is wrong, because it relates to the incoming message (and
1392 probably the interface on which it is arriving) and not to the outgoing
1393 callout (which could be using a different interface). This has been
1394 changed to use the value of the helo_data option from the smtp transport
1395 instead - this is what is used when a message is actually being sent. If
1396 there is no remote transport (possible with a router that sets up host
1397 addresses), $smtp_active_hostname is used.
1399 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1400 tweaks were necessary in order to get it to work (see also 21 below):
1401 (a) The code assumed that strncpy() returns a negative number on buffer
1402 overflow, which isn't the case. Replaced with Exim's string_format()
1404 (b) There were several signed/unsigned issues. I just did the minimum
1405 hacking in of casts. There is scope for a larger refactoring.
1406 (c) The code used strcasecmp() which is not a standard C function.
1407 Replaced with Exim's strcmpic() function.
1408 (d) The code set only $1; it now sets $auth1 as well.
1409 (e) A simple test gave the error "authentication client didn't specify
1410 service in request". It would seem that Dovecot has changed its
1411 interface. Fortunately there's a specification; I followed it and
1412 changed what the client sends and it appears to be working now.
1414 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1417 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1418 address A is aliased to B and C, where B exists and C does not. Without
1419 -v the output is "A verified" because verification stops after a
1420 successful redirection if more than one address is generated. However,
1421 with -v the child addresses are also verified. Exim was outputting "A
1422 failed to verify" and then showing the successful verification for C,
1423 with its parentage. It now outputs "B failed to verify", showing B's
1424 parentage before showing the successful verification of C.
1426 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1427 look up a TXT record in a specific list after matching in a combined
1430 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1431 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1432 they consult the DNS. I had assumed they would set it the way they
1433 wanted; and indeed my experiments on Linux seem to show that in some
1434 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1435 To be on the safe side, however, I have now made the interface to
1436 host_find_byname() similar to host_find_bydns(), with an argument
1437 containing the DNS resolver options. The host_find_byname() function now
1438 sets these options at its start, just as host_find_bydns() does. The smtp
1439 transport options dns_qualify_single and dns_search_parents are passed to
1440 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1441 of host_find_byname() use the default settings of RES_DEFNAMES
1442 (qualify_single) but not RES_DNSRCH (search_parents).
1444 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1445 spool_read_header() do less string testing, by means of a preliminary
1446 switch on the second character of optional "-foo" lines. (This is
1447 overdue, caused by the large number of possibilities that now exist.
1448 Originally there were few.) While I was there, I also converted the
1449 str(n)cmp tests so they don't re-test the leading "-" and the first
1450 character, in the hope this might squeeze out yet more improvement.
1452 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1453 flag allowing group syntax was set by the header_syntax check but not
1454 turned off, possible causing trouble later; (2) The flag was not being
1455 set at all for the header_verify test, causing "group"-style headers to
1456 be rejected. I have now set it in this case, and also caused header_
1457 verify to ignore an empty address taken from a group. While doing this, I
1458 came across some other cases where the code for allowing group syntax
1459 while scanning a header line wasn't quite right (mostly, not resetting
1460 the flag correctly in the right place). These bugs could have caused
1461 trouble for malformed header lines. I hope it is now all correct.
1463 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1464 with the "reply" argument non-NULL. The code, however (which originally
1465 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1466 but it didn't always do it. This confused somebody who was copying the
1467 code for some other use. I have removed all the tests.
1469 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1470 feature that was used to support insecure browsers during the U.S. crypto
1471 embargo. It requires special client support, and Exim is probably the
1472 only MTA that supported it -- and would never use it because real RSA is
1473 always available. This code has been removed, because it had the bad
1474 effect of slowing Exim down by computing (never used) parameters for the
1475 RSA_EXPORT functionality.
1477 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1478 authenticator to fail if there's a tab character in the incoming data
1479 (there should never be unless someone is messing about, as it's supposed
1480 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1481 if the connection is using TLS or if the remote IP is the same as the
1482 local IP, and the "valid-client-cert option" if a client certificate has
1485 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1486 authenticators. This can be used for authorization after authentication
1487 succeeds. (In the case of plaintext, it servers for both authentication
1490 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1491 if any retry times were supplied.
1493 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1494 connection, where there is no client IP address to check. In this
1495 situation, the verify now always succeeds.
1497 PH/25 Applied John Jetmore's -Mset patch.
1499 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1501 PH/27 In a string expansion for a processed (not raw) header when multiple
1502 headers of the same name were present, leading whitespace was being
1503 removed from all of them, but trailing whitespace was being removed only
1504 from the last one. Now trailing whitespace is removed from each header
1505 before concatenation. Completely empty headers in a concatenation (as
1506 before) are ignored.
1508 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1509 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1511 PH/29 [Removed. This was a change that I later backed out, and forgot to
1512 correct the ChangeLog entry (that I had efficiently created) before
1513 committing the later change.]
1515 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1516 address errors (4xx response to RCPT) over the same connection as other
1517 messages routed to the same hosts. Such deliveries are always "forced",
1518 so retry times are not inspected. This resulted in far too many retries
1519 for the affected addresses. The effect occurred only when there were more
1520 hosts than the hosts_max_try setting in the smtp transport when it had
1521 the 4xx errors. Those hosts that it had tried were not added to the list
1522 of hosts for which the message was waiting, so if all were tried, there
1523 was no problem. Two fixes have been applied:
1525 (i) If there are any address or message errors in an SMTP delivery, none
1526 of the hosts (tried or untried) are now added to the list of hosts
1527 for which the message is waiting, so the message should not be a
1528 candidate for sending over the same connection that was used for a
1529 successful delivery of some other message. This seems entirely
1530 reasonable: after all the message is NOT "waiting for some host".
1531 This is so "obvious" that I'm not sure why it wasn't done
1532 previously. Hope I haven't missed anything, but it can't do any
1533 harm, as the worst effect is to miss an optimization.
1535 (ii) If, despite (i), such a delivery is accidentally attempted, the
1536 routing retry time is respected, so at least it doesn't keep
1537 hammering the server.
1539 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1540 in ${readsocket because some servers need this prod.
1542 PH/32 Added some extra debug output when updating a wait-xxx database.
1544 PH/33 The hint "could be header name not terminated by colon", which has been
1545 given for certain expansion errors for a long time, was not being given
1546 for the ${if def:h_colon_omitted{... case.
1548 PH/34 The spec says: "With one important exception, whenever a domain list is
1549 being scanned, $domain contains the subject domain." There was at least
1550 one case where this was not true.
1552 PH/35 The error "getsockname() failed: connection reset by peer" was being
1553 written to the panic log as well as the main log, but it isn't really
1554 panic-worthy as it just means the connection died rather early on. I have
1555 removed the panic log writing for the ECONNRESET error when getsockname()
1558 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1559 runs only) independently of the message's sender address. This meant
1560 that, if the 4xx error was in fact related to the sender, a different
1561 message to the same recipient with a different sender could confuse
1562 things. In particualar, this can happen when sending to a greylisting
1563 server, but other circumstances could also provoke similar problems.
1564 I have changed the default so that the retry time for these errors is now
1565 based a combination of the sender and recipient addresses. This change
1566 can be overridden by setting address_retry_include_sender=false in the
1569 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1570 remote server are returned as part of bounce messages. This was not
1571 happening for LMTP over a pipe (the lmtp transport), but now it is the
1572 same for both kinds of LMTP.
1574 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1575 in header lines that were in fact CNAMEs. This is no longer the case.
1577 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1578 and queue runs started by the daemon processed all messages. This has
1579 been fixed so that -R and -S can now usefully be given with -q<time>.
1581 PH/40 Import PCRE release 6.7 (fixes some bugs).
1583 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1585 PH/42 Give an error if -q is specified more than once.
1587 PH/43 Renamed the variables $interface_address and $interface_port as
1588 $received_ip_address and $received_port, to make it clear that these
1589 values apply to message reception, and not to the outgoing interface when
1590 a message is delivered. (The old names remain recognized, of course.)
1592 PH/44 There was no timeout on the connect() call when using a Unix domain
1593 socket in the ${readsocket expansion. There now is.
1595 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1596 be meaningful with "accept".
1598 SC/01 Eximstats V1.43
1599 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1601 SC/02 Eximstats V1.44
1602 Use a glob alias rather than an array ref in the generated
1603 parser. This improves both readability and performance.
1605 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1606 Collect SpamAssassin and rejection statistics.
1607 Don't display local sender or destination tables unless
1608 there is data to show.
1609 Added average volumes into the top table text output.
1611 SC/04 Eximstats V1.46
1612 Collect data on the number of addresses (recipients)
1613 as well as the number of messages.
1615 SC/05 Eximstats V1.47
1616 Added 'Message too big' to the list of mail rejection
1617 reasons (thanks to Marco Gaiarin).
1619 SC/06 Eximstats V1.48
1620 Mainlog lines which have GMT offsets and are too short to
1621 have a flag are now skipped.
1623 SC/07 Eximstats V1.49 (Alain Williams)
1624 Added the -emptyok flag.
1626 SC/08 Eximstats V1.50
1627 Fixes for obtaining the IP address from reject messages.
1629 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1630 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1631 whitesspace changes from 4.64-PH/27
1633 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1636 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1637 are found, allow negative numbers in numeric criteria)
1639 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1641 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1642 to match changes made in 4.64-PH/43
1644 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1646 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1647 "domain" is actually an IP address, and force a failure. This locks out
1648 those revolvers/nameservers that support "A-for-A" lookups, in
1649 contravention of the specifications.
1651 PH/48 When a host name was looked up from an IP address, and the subsequent
1652 forward lookup of the name timed out, the host name was left in
1653 $sender_host_name, contrary to the specification.
1655 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1656 restricted to single-key lookups, Exim was not diagnosing an error if
1657 * or *@ was used with a query-style lookup.
1659 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1661 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1662 long ago, but noone interested enough thought of it. Let's just say that
1663 the "1.1" means that there are some new functions that weren't there at
1664 some point in the past.
1666 PH/51 Error processing for expansion failure of helo_data from an smtp
1667 transport during callout processing was broken.
1669 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1670 tested/used via the -bh/-bhc/-bs options.
1672 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1673 bug, fixed in subsequent PCRE releases).
1675 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1676 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1678 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1684 SC/01 Use a glob alias rather than an array ref in eximstats generated
1685 parser. This improves both readability and performance.
1687 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1688 Don't display local sender or destination tables in eximstats unless
1689 there is data to show.
1690 Added average volumes into the eximstats top table text output.
1692 SC/03 Collect data on the number of addresses (recipients) as well
1693 as the number of messages in eximstats.
1695 TF/01 Correct an error in the documentation for the redirect router. Exim
1696 does (usually) call initgroups() when daemonizing.
1698 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1699 with consistent privilege compared to when running as a daemon.
1701 TF/03 Note in the spec that $authenticated_id is not set for local
1702 submissions from trusted users.
1704 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1705 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1707 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1708 by adding some example configuration directives to the default
1709 configuration file. A little bit of work is required to uncomment the
1710 directives and define how usernames and passwords are checked, but
1711 there is now a framework to start from.
1713 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1714 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1715 without this. I don't know how relevant this is to other LDAP libraries.
1717 PH/02 Add the verb name to the "unknown ACL verb" error.
1719 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1721 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1723 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1724 directory not expanded when it should be if an expanded home directory
1725 was set for the address (which is overridden by the transport).
1727 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1730 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1731 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1732 because it is too late at that time, and has no effect.
1734 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1735 security issue with \' (bugzilla #107). I could not use the
1736 PQescapeStringConn() function, because it needs a PGconn value as one of
1739 PH/08 When testing addresses using -bt, indicate those final addresses that
1740 are duplicates that would not cause an additional delivery. At least one
1741 person was confused, thinking that -bt output corresponded to deliveries.
1742 (Suppressing duplicates isn't a good idea as you lose the information
1743 about possibly different redirections that led to the duplicates.)
1745 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1746 systems where poll() doesn't work, in particular OS X.
1748 PH/10 Added more information to debugging output for retry time not reached.
1750 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1751 operations in malware.c.
1753 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1756 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1757 syslog_duplication set false, log lines that would normally be written
1758 both the the main log and to the reject log were not written to syslog at
1761 PH/14 In the default configuration, change the use of "message" in ACL warn
1762 statements to "add_header".
1764 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1765 not followed by a command (e.g. "seen endif").
1767 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1768 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1771 PH/17 Added extra conditions to the default value of delay_warning_condition
1775 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1776 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1777 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1780 The Auto-Submitted: and various List- headers are standardised, whereas I
1781 don't think Precedence: ever was.
1783 PH/18 Refactored debugging code in route_finduser() to show more information,
1784 in particular, the error code if getpwnam() issues one.
1786 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1787 This is apparently needed in addition to the PH/07 change above to avoid
1788 any possible encoding problems.
1790 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1791 but not after initializing Perl.
1793 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1794 output them only if debugging. By default they are written stderr,
1795 apparently, which is not desirable.
1797 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1800 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1803 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1805 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1806 authenticated or an ident call has been made. Suppress the default
1807 values for $authenticated_id and $authenticated_sender (but permit -oMai
1808 and -oMas) when testing with -bh.
1810 PH/24 Re-jigged the order of the tests in the default configuration so that the
1811 tests for valid domains and recipients precede the DNS black list and CSA
1812 tests, on the grounds that those ones are more expensive.
1814 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1815 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1816 This bug exists in every version of Exim that I still have, right back to
1819 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1820 However, an attempt to turn on case-sensitivity in a regex key by
1821 including (?-i) didn't work because the subject string was already
1822 lowercased, and the effects were non-intuitive. It turns out that a
1823 one-line patch can be used to allow (?-i) to work as expected.
1829 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1830 other effects) broke the use of negated acl sub-conditions.
1832 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1835 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1836 "Deny" causes Exim to reject the incoming connection with a 554 error.
1837 Unfortunately, if there is a major crisis, such as a disk failure,
1838 tcp-wrappers gives "deny", whereas what one would like would be some
1839 kind of temporary error. A kludge has been added to help with this.
1840 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1841 554 error is used if errno is still zero or contains ENOENT (which occurs
1842 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1845 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1847 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1848 errors. Otherwise a message that provokes a temporary error (when other
1849 messages do not) can cause a whole host to time out.
1851 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1852 addresses were routed directly to files or pipes from a redirect router.
1853 File deliveries just didn't batch; pipe deliveries might have suffered
1856 PH/06 A failure to get a lock for a hints database would erroneously always say
1857 "Failed to get write lock", even when it was really a read lock.
1859 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1860 of 0600. This has been changed to use the value of the lockfile_mode
1861 option (which defaults to 0600).
1863 PH/08 Applied small patch from the Sieve maintainer.
1865 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1866 folder from quota calculations, a direct delivery into this folder messed
1867 up the contents of the maildirsize file. This was because the regex was
1868 used only to exclude .Trash (or whatever) when the size of the mailbox
1869 was calculated. There was no check that a delivery was happening into an
1870 excluded directory. This bug has been fixed by ignoring all quota
1871 processing for deliveries into excluded directories.
1873 PH/10 Added the maildirfolder_create_regex option to appendfile.
1879 PH/01 The code for finding all the local interface addresses on a FreeBSD
1880 system running IPv6 was broken. This may well have applied to all BSD
1881 systems, as well as to others that have similar system calls. The broken
1882 code found IPv4 interfaces correctly, but gave incorrect values for the
1883 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1884 that it would not match correctly against @[] and not recognize the IPv6
1887 PH/02 The ipliteral router was not recognizing addresses of the form user@
1888 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1890 PH/03 Added disable_ipv6.
1892 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1893 decoded form, because it is most often used to construct To: headers
1894 lines in autoreplies, and the decoded form may well be syntactically
1895 invalid. However, $reply_address has leading white space removed, and all
1896 newlines turned into spaces so that the autoreply transport does not
1899 PH/05 If group was specified without a user on a router, and no group or user
1900 was specified on a transport, the group from the router was ignored.
1902 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1903 for visible compile-time settings that can be used to change these
1904 numbers, for those that want even more. Backwards compatibility with old
1905 spool files has been maintained. However, going back to a previous Exim
1906 release will lost any variables that are in spool files.
1908 PH/07 Two small changes when running in the test harness: increase delay when
1909 passing a TCP/IP connection to a new process, in case the original
1910 process has to generate a bounce, and remove special handling of
1911 127.0.0.2 (sic), which is no longer necessary.
1913 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1914 be the same on different OS.
1916 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1919 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1920 whether --show-vars was specified or not
1922 JJ/02 exipick: Added support for new ACL variable spool format introduced
1925 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1926 syntactically invalid From: or Reply-to: line, and a filter used this to
1927 generate an autoreply, and therefore failed to obtain an address for the
1928 autoreply, Exim could try to deliver to a non-existent relative file
1929 name, causing unrelated and misleading errors. What now happens is that
1930 it logs this as a hard delivery error, but does not attempt to create a
1933 PH/11 The exinext utility has a -C option for testing purposes, but although
1934 the given file was scanned by exinext itself; it wasn't being passed on
1935 when Exim was called.
1937 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1938 an end-of-file indication when reading a command response.
1940 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1941 compiled. In many other places in Exim, IPv6 addresses are always
1942 recognized, so I have changed this. It also means that IPv4 domain
1943 literals of the form [IPV4:n.n.n.n] are now always recognized.
1945 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1946 used if the router is not running as root, for example, when verifying at
1947 ACL time, or when using -bh. The debugging output from this situation was
1948 non-existent - all you got was a failure to exec. I have made two
1951 (a) Failures to set uid/gid, the current directory, or a process leader
1952 in a subprocess such as that created by queryprogram now generate
1953 suitable debugging ouput when -d is set.
1955 (b) The queryprogram router detects when it is not running as root,
1956 outputs suitable debugging information if -d is set, and then runs
1957 the subprocess without attempting to change uid/gid.
1959 PH/15 Minor change to Makefile for building test_host (undocumented testing
1962 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1963 additional section of a DNS packet that returns MX or SRV records.
1964 Instead, it always explicitly searches for A/AAAA records. This avoids
1965 major problems that occur when a DNS server includes only records of one
1966 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1967 fixed another bug: if SRV records were looked up and the corresponding
1968 address records were *not* found in the additional section, the port
1969 values from the SRV records were lost.
1971 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1972 using the correct key (the original address) when searching the retry
1973 rules in order to find which one to use for generating the retry hint.
1975 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1976 adding the default one. Similarly, if it contains a Reply-To: header, the
1977 errors_reply_to option, if set, is not used.
1979 PH/19 When calculating a retry time, Exim used to measure the "time since
1980 failure" by looking at the "first failed" field in the retry record. Now
1981 it does not use this if it is later than than the arrival time of the
1982 message. Instead it uses the arrival time. This makes for better
1983 behaviour in cases where some deliveries succeed, thus re-setting the
1984 "first failed" field. An example is a quota failure for a huge message
1985 when small messages continue to be delivered. Without this change, the
1986 "time since failure" will always be short, possible causing more frequent
1987 delivery attempts for the huge message than are intended.
1988 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1990 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1991 $1, $2, $3) because the numerical variables can be reset during some
1992 expansion items (e.g. "match"), thereby losing the authentication data.
1994 PH/21 Make -bV show the size of off_t variables so that the test suite can
1995 decide whether to run tests for quotas > 2G.
1997 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1998 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1999 filecount value is greater than 2G or if a quota value is greater than 2G
2000 on a system where the size of off_t is not greater than 4, a panic error
2003 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2004 never match. The debug and -bh output now contains an explicit error
2005 message indicating a malformed IPv4 address or mask.
2007 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2008 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2009 PH/23 above applies.
2011 PH/25 Do not write to syslog when running in the test harness. The only
2012 occasion when this arises is a failure to open the main or panic logs
2013 (for which there is an explicit test).
2015 PH/26 Added the /no_tell option to "control=freeze".
2017 PH/27 If a host name lookup failed very early in a connection, for example, if
2018 the IP address matched host_lookup and the reverse lookup yielded a name
2019 that did not have a forward lookup, an error message of the form "no IP
2020 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2021 could be logged. Now it outputs the IP address instead of "NULL".
2023 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2024 allows the sender and the authenticated sender to be set when
2025 submitting a message from within Exim. Since child_open_exim() is
2026 documented for local_scan(), the new function should be too.
2028 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2029 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2030 results in an empty string is now treated as unset.
2032 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2034 PH/31 Added sender_verify_fail logging option.
2036 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2037 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2038 tidied the source and removed it altogether.
2040 PH/33 When a queue run was abandoned because the load average was too high, a
2041 log line was always written; now it is written only if the queue_run log
2042 selector is set. In addition, the log line for abandonment now contains
2043 information about the queue run such as the pid. This is always present
2044 in "start" and "stop" lines but was omitted from the "abandon" line.
2046 PH/34 Omit spaces between a header name and the colon in the error message that
2047 is given when verify = headers_syntax fails (if there are lots of them,
2048 the message gets confusing).
2050 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2051 names, as there are now some PTR records that contain slashes. This check
2052 is only to protect against broken name servers that fall over on strange
2053 characters, so the fact that it applies to all lookups doesn't matter.
2055 PH/36 Now that the new test suite is complete, we can remove some of the
2056 special code in Exim that was needed for the old test suite. For example,
2057 sorting DNS records because real resolvers return them in an arbitrary
2058 order. The new test suite's fake resolver always returns records in the
2061 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2062 bounces) except when queue_only is set, to avoid logging races between
2063 the different processes.
2065 PH/38 Panic-die if .include specifies a non-absolute path.
2067 PH/39 A tweak to the "H" retry rule from its user.
2069 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2070 a label. They prevented compilation on older perls.
2072 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2073 a warning to be raised on newish perls.
2075 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2076 on queue. Changes to match documented behaviour of showing count of
2077 messages matching specified criteria.
2079 PH/40 Changed the default ident timeout from 30s to 5s.
2081 PH/41 Added support for the use of login_cap features, on those BSD systems
2082 that have them, for controlling the resources used by pipe deliveries.
2084 PH/42 The content-scanning code uses fopen() to create files in which to put
2085 message data. Previously it was not paying any attention to the mode of
2086 the files. Exim runs with umask(0) because the rest of the code creates
2087 files with open(), and sets the required mode explicitly. Thus, these
2088 files were ending up world-writeable. This was not a big issue, because,
2089 being within the spool directory, they were not world-accessible. I have
2090 created a function called modefopen, which takes an additional mode
2091 argument. It sets umask(777), creates the file, chmods it to the required
2092 mode, then resets the umask. All the relevant calls to fopen() in the
2093 content scanning code have been changed to use this function.
2095 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2096 to 24 hours. This avoids potential overflow problems when processing G
2097 and H retry rules. I suspect nobody ever tinkers with this value.
2099 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2101 PH/45 When the plaintext authenticator is running as a client, the server's
2102 challenges are checked to ensure they are valid base64 strings. By
2103 default, the authentication attempt is cancelled if an invalid string is
2104 received. Setting client_ignore_invalid_base64 true ignores these errors.
2105 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2106 they are received. Thus, the responses can be made to depend on the
2107 challenges. If an invalid string is ignored, an empty string is placed in
2110 PH/46 Messages that are created by the autoreply transport now contains a
2111 References: header, in accordance with RFCs 2822 and 3834.
2113 PH/47 Added authenticated_sender_force to the smtp transport.
2115 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2117 PH/49 Installed latest patch from the Sieve maintainer.
2119 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2120 was also set, the appendfile transport was unnecessarily scanning a
2121 directory of message files (e.g. for maildir delivery) to find the count
2122 of files (along with the size), even though it did not need this
2123 information. It now does the scan only if it needs to find either the
2124 size of the count of files.
2126 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2128 PH/52 Two bugs concerned with error handling when the smtp transport is
2131 (i) Exim was not creating retry information for temporary errors given
2132 for individual recipients after the DATA command when the smtp transport
2133 was used in LMTP mode. This meant that they could be retried too
2134 frequently, and not timed out correctly.
2136 (ii) Exim was setting the flag that allows error details to be returned
2137 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2138 recipients that were returned after the DATA command.
2140 PH/53 This is related to PH/52, but is more general: for any failing address,
2141 when detailed error information was permitted to be returned to the
2142 sender, but the error was temporary, then after the final timeout, only
2143 "retry timeout exceeded" was returned. Now it returns the full error as
2144 well as "retry timeout exceeded".
2146 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2147 do this, and (what is worse) MTAs that accept it.
2149 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2150 will now be deprecated.
2152 PH/56 New os.c-cygwin from the Cygwin maintainer.
2154 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2155 formats (previously only available in exim formats via -bpr, -bpru,
2156 and -bpra. Now also available in native and exiqgrep formats)
2158 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2159 with very large, slow to parse queues
2161 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2163 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2165 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2166 responses to authentication challenges, though it was showing the
2167 challenges; (ii) I've removed the CR characters from the debug output for
2170 PH/58 Allow for the insertion of a newline as well as a space when a string
2171 is turned into more than one encoded-word during RFC 2047 encoding. The
2172 Sieve code now uses this.
2174 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2175 data_4xx, lost_connection, tls_required.
2177 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2178 message was being sent as an SMTP response.
2180 PH/61 Add -l and -k options to exicyclog.
2182 PH/62 When verifying, if an address was redirected to one new address, so that
2183 verification continued, and the new address failed or deferred after
2184 having set something in $address_data, the value of $address_data was not
2185 passed back to the ACL. This was different to the case when no
2186 redirection occurred. The value is now passed back in both cases.
2188 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2189 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2190 use login_cap.h, so on its own it isn't the distinguishing feature. The
2191 new name refers directly to the setclassresources() function.
2193 PH/65 Added configuration files for NetBSD3.
2195 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2197 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2198 is preferred over IPv4.
2200 PH/68 The bounce_return_message and bounce_return_body options were not being
2201 honoured for bounces generated during the reception of non-SMTP messages.
2202 In particular, this applied to messages rejected by the ACL. This bug has
2203 been fixed. However, if bounce_return_message is true and bounce_return_
2204 body is false, the headers that are returned for a non-SMTP message
2205 include only those that have been read before the error was detected.
2206 (In the case of an ACL rejection, they have all been read.)
2208 PH/69 The HTML version of the specification is now built in a directory called
2209 spec_html instead of spec.html, because the latter looks like a path with
2210 a MIME-type, and this confuses some software.
2212 PH/70 Catch two compiler warnings in sieve.c.
2214 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2215 function verify_get_ident() calls ip_connect() to connect a socket, but
2216 if the "connect()" function timed out, ip_connect() used to close the
2217 socket. However, verify_get_ident() also closes the socket later, and in
2218 between Exim writes to the log, which may get opened at this point. When
2219 the socket was closed in ip_connect(), the log could get the same file
2220 descriptor number as the socket. This naturally causes chaos. The fix is
2221 not to close the socket in ip_connect(); the socket should be closed by
2222 the function that creates it. There was only one place in the code where
2223 this was missing, in the iplookup router, which I don't think anybody now
2224 uses, but I've fixed it anyway.
2226 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2227 well as to direct DNS lookups. Otherwise the handling of names in host
2228 lists is inconsistent and therefore confusing.
2234 PH/01 Two changes to the default runtime configuration:
2236 (1) Move the checks for relay_from_hosts and authenticated clients from
2237 after to before the (commented out) DNS black list checks.
2239 (2) Add control=submission to the relay_from_hosts and authenticated
2240 clients checks, on the grounds that messages accepted by these
2241 statements are most likely to be submissions.
2243 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2245 (1) Generate an error if the third argument for the ${prvs expansion is
2248 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2251 (3) Reset the variables that are obtained from the first argument of
2252 ${prvscheck and used in the second argument before leaving the code,
2253 because their memory is reclaimed, so using them afterwards may do
2256 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2257 one (it's much easier than Tom thought :-).
2259 (5) Because of (4), we can now allow for the use of $prvscheck_result
2260 inside the third argument.
2262 PH/03 For some reason, the default setting of PATH when running a command from
2263 a pipe transport was just "/usr/bin". I have changed it to
2266 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2267 anything to be listed in the output from -bV.
2269 PH/05 When a filter generated an autoreply, the entire To: header line was
2270 quoted in the delivery log line, like this:
2272 => >A.N.Other <ano@some.domain> <original@ddress> ...
2274 This has been changed so that it extracts the operative address. There
2275 may be more than one such address. If so, they are comma-separated, like
2278 => >ano@some.domain,ona@other.domain <original@ddress> ...
2280 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2281 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2282 not being looked up in the rDNS to get a host name, Exim was showing the
2283 IP address twice in Received: lines, even though the IP addresses were
2284 identical. For example:
2286 Received: from [1.2.3.4] (helo=[1.2.3.4])
2288 However, if the real host name was known, it was omitting the HELO data
2289 if it matched the actual IP address. This has been tidied up so that it
2290 doesn't show the same IP address twice.
2292 PH/07 When both +timestamp and +memory debugging was on, the value given by
2293 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2294 called by the debug printing, thereby overwriting the timestamp buffer.
2295 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2297 PH/08 When the original message was included in an autoreply transport, it
2298 always said "this is a copy of the message, including all the headers",
2299 even if body_only or headers_only was set. It now gives an appropriate
2302 PH/09 Applied a patch from the Sieve maintainer which:
2304 o fixes some comments
2305 o adds the (disabled) notify extension core
2306 o adds some debug output for the result of if/elsif tests
2307 o points to the current vacation draft in the documentation
2308 and documents the missing references header update
2312 o fixes a bug in processing the envelope test (when testing
2313 multiple envelope elements, the last element determinted the
2316 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2317 Electronic Mail") by including:
2319 Auto-submitted: auto-generated
2321 in the messages that it generates (bounce messages and others, such as
2322 warnings). In the case of bounce messages for non-SMTP mesages, there was
2323 also a typo: it was using "Auto_submitted" (underscore instead of
2324 hyphen). Since every message generated by Exim is necessarily in response
2325 to another message, thes have all been changed to:
2327 Auto-Submitted: auto-replied
2329 in accordance with these statements in the RFC:
2331 The auto-replied keyword:
2333 - SHOULD be used on messages sent in direct response to another
2334 message by an automatic process,
2336 - MUST NOT be used on manually-generated messages,
2338 - MAY be used on Delivery Status Notifications (DSNs) and Message
2339 Disposition Notifications (MDNs),
2341 - MUST NOT be used on messages generated by automatic or periodic
2342 processes, except for messages which are automatic responses to
2345 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2346 to the default Received: header definition.
2348 PH/12 Added log selector acl_warn_skipped (default on).
2350 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2351 variables because (a) they are in the wrong storage pool and (b) even if
2352 they were copied, it wouldn't work properly because of the caching.
2354 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2355 checking when decoding. Apparently there are clients that generate
2356 overlong encoded strings. Why am I not surprised?
2358 PH/15 If the first argument of "${if match_address" was not empty, but did not
2359 contain an "@" character, Exim crashed. Now it writes a panic log message
2360 and treats the condition as false.
2362 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2364 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2365 extension "envelope-auth". The code is finished and in agreement with
2366 other implementations, but there is no documentation so far and in fact,
2367 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2368 not changing the active code.
2370 Print executed "if" and "elsif" statements when debugging is used. This
2371 helps a great deal to understand what a filter does.
2373 Document more things not specified clearly in RFC3028. I had all this
2374 sorted out, when out of a sudden new issues came to my mind. Oops."
2376 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2379 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2380 canonical form (as documented). However, after a host name lookup from
2381 the IP address, check_host() was doing a simple string comparison with
2382 addresses acquired from the DNS when checking that the found name did
2383 have the original IP as one of its addresses. Since any found IPv6
2384 addresses are likely to be in abbreviated form, the comparison could
2385 fail. Luckily, there already exists a function for doing the comparison
2386 by converting both addresses to binary, so now that is used instead of
2387 the text comparison.
2389 PH/20 There was another similar case to PH/19, when a complete host name was
2390 given in a host list; looking up its IP address could give an abbreviated
2391 form, whereas the current host's name might or might not be abbreviated.
2392 The same fix has been applied.
2398 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2399 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2402 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2403 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2405 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2407 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2408 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2409 Also, for "Auto-Submitted", treat anything other than "no" as
2410 non-personal, in accordance with RFC 3834. (Previously it treated
2411 anything starting "auto-" as non-personal.)
2413 TF/01 The control=submission/name=... option had a problem with syntax
2414 errors if the name included a slash character. The /name= option
2415 now slurps the rest of the string, so it can include any characters
2416 but it must come last in the list of options (after /sender_retain
2419 PH/05 Some modifications to the interface to the fake nameserver for the new
2427 TK/01 Added the "success_on_redirect" address verification option. See
2428 NewStuff for rationale and an example.
2430 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2432 PH/02 Patch to exigrep to allow it to work on syslog lines.
2434 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2435 fread() to skip over the body file's header line, because in Cygwin the
2436 header line is locked and is inaccessible.
2438 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2439 co-exist for some time) to make it clear that it is the Exim ID that is
2440 referenced, not the Message-ID: header line.
2442 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2443 string_format() function, because snprintf() does not exist on all
2446 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2447 ${stat: expansion item.
2449 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2450 protocol synchronization error", to keep the pedants happy.
2452 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2453 well as for IRIX systems, when gcc is being used. See the host.c source
2456 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2458 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2461 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2462 transport and to the smtp transport in LMTP mode.
2464 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2466 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2467 run a filter in a subprocess. This could lead to confusion in subsequent
2468 lookups in the parent process. There should also be a search_tidyup() at
2469 the end of the subprocess.
2471 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2472 only if the host matched helo_try_verify_hosts, which caused the
2473 verification to occur when the EHLO/HELO command was issued. The ACL just
2474 tested the remembered result. Now, if a previous verification attempt has
2475 not happened, "verify = helo" does it there and then.
2477 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2479 TK/03 Fix log output including CR from clamd.
2481 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2482 exist provoked a memory error which could cause a segfault.
2484 PH/15 Installed PCRE 6.2
2486 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2488 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2489 of the problem. Specifically, suggested +O2 rather than +O1 for the
2492 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2494 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2495 record, the port was not correctly logged when the outgoing_port log
2496 selector was set (it logged the transort's default port).
2498 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2499 fallback_hosts, and "hosts" in the smtp transport.
2501 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2502 host errors such as "Connection refused".
2504 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2505 authentication with radiusclient 0.4.9:
2507 - Error returned from rc_read_config was caught wrongly
2508 - Username/password not passed on to radius server due to wrong length.
2510 The presumption is that some radiusclient API changes for 4.51/PH/17
2511 were not taken care of correctly. The code is still untested by me (my
2512 Linux distribution still has 0.3.2 of radiusclient), but it was
2513 contributed by a Radius user.
2515 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2516 expanding the "port" option of the smtp transport.
2518 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2519 while reading a MIME header. Thanks to Tom Hughes for a patch.
2521 PH/24 Include config.h inside local_scan.h so that configuration settings are
2524 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2525 that in an ACL for RCPT (for example), you can examine exactly what was
2528 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2529 commands, but it was not correctly comparing the address with the actual
2530 client host address. Thus, it would show the EHLO address in Received:
2531 header lines when this was not necessary.
2533 PH/27 Added the % operator to ${eval:}.
2535 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2536 it should be ignoring failures (because with -C, for example, it has lost
2537 privilege). It wasn't ignoring creation failures other than "already
2540 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2543 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2544 because an input line was too long, either on its own, or by virtue of
2545 too many continuations, the temporary file was not being removed, and the
2546 return code was incorrect.
2548 PH/31 Missing "BOOL" in function definition in filtertest.c.
2550 PH/32 Applied Sieve patches from the maintainer.
2552 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2554 PH/33 Added "verify = not_blind".
2556 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2557 Local/Makefile (with some defaults set). These are used in built scripts
2558 such as exicyclog, but they have never been used in the exim_install
2559 script (though there are many overriding facilities there). I have
2560 arranged that the exim_install script now takes note of these two
2563 PH/35 Installed configuration files for Dragonfly.
2565 PH/36 When a locally submitted message by a trusted user did not contain a
2566 From: header, and the sender address was obtained from -f or from an SMTP
2567 MAIL command, and the trusted user did not use -F to supply a sender
2568 name, $originator_name was incorrectly used when constructing a From:
2569 header. Furthermore, $originator_name was used for submission mode
2570 messages from external hosts without From: headers in a similar way,
2571 which is clearly wrong.
2573 PH/37 Added control=suppress_local_fixups.
2575 PH/38 When log_selector = +received_sender was set, and the addition of the
2576 sender made the log line's construction buffer exactly full, or one byte
2577 less than full, an overflow happened when the terminating "\n" was
2580 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2581 when the result of a list match is failure because a DNS lookup failed.
2583 PH/40 RM_COMMAND is now used in the building process.
2585 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2586 the "build-* directories that it finds.
2588 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2589 address in a domain literal was a prefix of an interface address.
2591 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2592 when verifying a sender address, unless rewrite_headers is false.
2594 PH/44 Wrote a long comment about why errors_to addresses are verified as
2595 recipients, not senders.
2597 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2598 the ratelimit ACL was added.
2600 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2602 PH/46 Added extra information about PostgreSQL errors to the error string.
2604 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2605 avoiding the need to install special zones in a real server. This is
2606 backwards compatible; if it can't find the fake resolver, it drops back.
2607 Thus, both old and new test suites can be run.
2609 TF/02 Added util/ratelimit.pl
2611 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2612 clock is set back in time.
2614 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2615 Candler <B.Candler@pobox.com>.
2617 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2618 OK for addresses that are the result of redirections.
2620 PH/48 A number of further additions for the benefit of the new test suite,
2621 including a fake gethostbyname() that interfaces to the fake DNS resolver
2624 TF/06 The fix for widen_domains has also been applied to qualify_single and
2625 search_parents which are the other dnslookup options that can cause
2628 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2631 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2633 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2634 service" if it encountered an MX record with an empty target hostname.
2635 The message is now "an MX or SRV record indicated no SMTP service".
2637 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2638 if the DNS of the sending site is misconfigured. This is quite a
2639 common situation. This change restores the behaviour of treating a
2640 helo verification defer as a failure.
2642 PH/51 If self=fail was set on a router, the bounce message did not include the
2643 actual error message.
2649 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2651 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2652 command itself ended in error, the underlying message about the transport
2653 filter timeout was being overwritten with the pipe command error. Now the
2654 underlying error message should be appended to the second error message.
2656 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2658 PH/02 Reduce the amount of output that "make" produces by default. Full output
2659 can still be requested.
2661 PH/03 The warning log line about a condition test deferring for a "warn" verb
2662 was being output only once per connection, rather than after each
2663 occurrence (because it was using the same function as for successful
2664 "warn" verbs). This seems wrong, so I have changed it.
2666 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2667 it should not have, which might have caused a crash in the right
2668 circumstances, but probably never did.
2670 PH/04 Installed a modified version of Tony Finch's patch to make submission
2671 mode fix the return path as well as the Sender: header line, and to
2672 add a /name= option so that you can make the user's friendly name appear
2675 TF/03 Added the control = fakedefer ACL modifier.
2677 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2678 Mark Lowes for thorough testing.
2680 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2682 TK/03 Merged latest SRS patch from Miles Wilton.
2684 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2685 with the definition in sysexits.h (which is #included earlier).
2686 Fortunately, Exim does not actually use EX_OK. The code used to try to
2687 preserve the sysexits.h value, by assumimg that macro definitions were
2688 scanned for macro replacements. I have been disabused of this notion,
2689 so now the code just undefines EX_OK before #including unistd.h.
2691 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2692 in the smtp transport. When a block could not be written in a single
2693 write() function, the timeout was being re-applied to each part-write.
2694 This seems wrong - if the receiver was accepting one byte at a time it
2695 would take for ever. The timeout is now adjusted when this happens. It
2696 doesn't have to be particularly precise.
2698 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2699 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2701 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2702 without a domain if the check to <postmaster@domain> fails.
2704 SC/01 Eximstats: added -xls and the ability to specify output files
2705 (patch written by Frank Heydlauf).
2707 SC/02 Eximstats: use FileHandles for outputing results.
2709 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2711 SC/04 Eximstats: fixed display of large numbers with -nvr option
2713 SC/05 Eximstats: fixed merging of reports with empty tables.
2715 SC/06 Eximstats: added the -include_original_destination flag
2717 SC/07 Eximstats: removed tabs and trailing whitespace.
2719 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2721 TK/06 MBOX spool code: Add real "From " MBOX separator line
2722 so the .eml file is really in mbox format (even though
2723 most programs do not really care). Patch from Alex Miller.
2725 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2726 The latter is generated from $received_to and is only set if the
2727 message has one envelope recipient. SA can use these headers,
2728 obviously out-of-the-box. Patch from Alex Miller.
2730 PH/08 The ${def test on a variable was returning false if the variable's
2731 value was "0", contrary to what the specification has always said!
2732 The result should be true unless the variable is empty.
2734 PH/09 The syntax error of a character other than { following "${if
2735 def:variable_name" (after optional whitespace) was not being diagnosed.
2736 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2737 accidental colon was present, for example, could give incorrect results.
2739 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2740 result is used (not including appendfile, where other changes are about
2743 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2744 This involved changing a lot of size variables from int to off_t. It
2745 should work with maildirs and everything.
2747 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2748 spamd dying while we are connected to it.
2750 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2753 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2754 function for BDB 4.3.
2756 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2758 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2759 the sources for PCRE have been re-arranged and more files are now
2762 PH/15 The code I had for printing potentially long long variables in PH/11
2763 above was not the best (it lost precision). The length of off_t variables
2764 is now inspected at build time, and an appropriate printing format (%ld
2765 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2766 to be "long long int" or "long int". This is needed for the internal
2767 formatting function string_vformat().
2769 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2770 the configuration file to be ":syslog", then the script "guesses" where
2771 the logs files are, rather than using the compiled in default. In our
2772 case the guess is not the same as the compiled default, so the script
2773 suddenly stopped working when I started to use syslog. The patch checks
2774 to see if log_file_path is "". If so, it attempts to read it from exim
2775 with no configuration file to get the compiled in version, before it
2776 falls back to the previous guessing code."
2778 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2779 implementing BATV in an Exim configuration. See NewStuff for the gory
2782 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2783 Makefile that are specific to HP-UX.
2785 PH/18 If the "use_postmaster" option was set for a recipient callout together
2786 with the "random" option, the postmaster address was used as the MAIL
2787 FROM address for the random test, but not for the subsequent recipient
2788 test. It is now used for both.
2790 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2791 patch removes a few documentation additions to RFC 3028, because the
2792 latest draft now contains them. It adds the new en;ascii-case comparator
2793 and a new error check for 8bit text in MIME parts. Comparator and
2794 require names are now matched exactly. I enabled the subaddress
2795 extension, but it is not well tested yet (read: it works for me)."
2797 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2798 rework some of the code of TK/09 above to avoid the hardwired use of
2799 "%lld" and "long long". Replaced the call to snprintf() with a call to
2802 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2803 records point to non-existent hosts", "retry timeout exceeded", and
2804 "retry time not reached for any host after a long failure period".
2806 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2807 experimental DomainKeys support:
2809 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2810 (2) On an error such as an illegally used "control", the wrong name for
2811 the control was given.
2813 These problems did NOT occur unless DomainKeys support was compiled.
2815 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2817 PH/24 Added ${if match_ip condition.
2819 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2820 databases so that it will be absolutely obvious if a crash occurs in the
2821 DB library. This is a regular occurrence (often caused by mis-matched
2824 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2825 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2826 picked up on a user's system that detects such things. There doesn't seem
2827 to be a gcc warning option for this - only an attribute that has to be
2828 put on the function's prototype. It seems that in Fedora Core 4 they have
2829 set this on a number of new functions. No doubt there will be more in due
2832 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2833 specify a transport. However, if an address that was verified by such a
2834 router was the subject of a callout, Exim crashed because it tried to
2835 read the rcpt_include_affixes from the non-existent transport. Now it
2836 just assumes that the setting of that option is false. This bug was
2837 introduced by 4.51/PH/31.
2839 PH/28 Changed -d+all to exclude +memory, because that information is very
2840 rarely of interest, but it makes the output a lot bigger. People tend to
2841 do -d+all out of habit.
2843 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2844 code in os-type was giving problems when libc.so lives in lib64, like on
2847 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2848 aren't the modern standard, and it seems that some systems' include files
2849 don't always have them. Exim was already checking for some of the newer
2850 ones like T_AAAA, and defining it itself. I've added checks for all the
2851 record types that Exim uses.
2853 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2854 not automatically generating a new one, as it is supposed to. This
2855 prevented TLS from working. If the file did exist, but contained invalid
2856 data, a new version was generated, as expected. It was only the case of a
2857 non-existent file that was broken.
2859 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2860 with a change in libdomainkeys > 0.64.
2862 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2863 from DNS. If the selector record carries the flag, it now has
2864 precedence over the domain-wide flag.
2866 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2868 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2869 the use of a port name that isn't defined in /etc/services) occurred, the
2870 message was deferred as in a normal delivery, and thus remained on the
2871 spool, instead of being failed because of the mua_wrapper setting. This
2872 is now fixed, and I tidied up some of the mua_wrapper messages at the
2875 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2876 the messages in a hash of arrays rather than using individual hashes.
2877 This is a bit cleaner and results in dramatic memory savings, albeit
2878 at a slight CPU cost.
2880 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2881 as requested by Marc Sherman.
2883 SC/10 Eximstats: added histograms for user specified patterns as requested
2886 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2888 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2889 fopen() in the content-scanning modules that did not already have it.
2895 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2896 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2898 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2900 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2902 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2903 to test Sieve filters that use "vacation".
2905 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2906 that changes the way the GnuTLS parameters are stored in the cache file.
2907 The new format can be generated externally. For backward compatibility,
2908 if the data in the cache doesn't make sense, Exim assumes it has read an
2909 old-format file, and it generates new data and writes a new file. This
2910 means that you can't go back to an older release without removing the
2913 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2914 work if there are any delivery delays because "one_time" forces the
2915 parent to be marked "delivered", so its unseen clone is never tried
2916 again. For this reason, Exim now forbids the simultaneous setting of
2919 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2920 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2921 entry if you want to know the details. The fix, however, neglected to
2922 consider the case when local delivery batching is involved. The test for
2923 "previously delivered" was not happening when checking to see if an
2924 address could be batched with a previous (undelivered) one; under
2925 certain circumstances this could lead to multiple deliveries to the same
2928 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2929 in its include files, and this causes problems building Exim.
2931 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2932 header_syntax) but Exim was just ignoring anything given after a slash.
2933 In particular, this caused confusion with an attempt to use "verify =
2934 reverse_host_lookup/defer_ok". An error is now given when options are
2935 supplied for verify items that do not have them. (Maybe reverse_host_
2936 lookup should have a defer_ok option, but that's a different point.)
2938 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2939 defined by RFC 821) to 2048, because there were problems with some AUTH
2940 commands, and RFC 1869 says the size should be increased for extended
2941 SMTP commands that take arguments.
2943 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2946 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2947 "unknown" error; now it says that the functionality isn't in the binary.
2949 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2950 an address' error message when a string expansion fails (syntax or
2951 whatever). Otherwise the password may appear in the log. Following change
2952 PH/42 below, there is no longer a chance of it appearing in a bounce
2955 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2957 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2958 of its addresses was ever tried. (Bugzilla bug #2.)
2960 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2961 the result incorrectly in the debug output. (It correctly added a newline
2962 to what was transported.)
2964 TF/01 Added $received_time.
2966 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2967 commented out examples of how to interface to a virus scanner and to
2968 SpamAssassin. Also added commented examples of av_scanner and
2969 spamd_address settings.
2971 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2972 and controls are allowed in which ACLs. There were a couple of minor
2973 errors. Some of the entries in the conditions table (which is a table of
2974 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2975 negation of where the condition IS allowed.
2977 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2979 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2980 header file does not have a version number, so I've had to invent a new
2981 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2982 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2983 radiusclient), but it was contributed by a Radius user.
2985 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2986 files or directories, for OpenSSL.
2988 PH/19 When an Exim process that is running as root has to create an Exim log
2989 file, it does so in a subprocess that runs as exim:exim so as to get the
2990 ownership right at creation (otherwise, other Exim processes might see
2991 the file with the wrong ownership). There was no test for failure of this
2992 fork() call, which would lead to the process getting stuck as it waited
2993 for a non-existent subprocess. Forks do occasionally fail when resources
2994 run out. I reviewed all the other calls to fork(); they all seem to check
2997 PH/20 When checking for unexpected SMTP input at connect time (before writing
2998 the banner), Exim was not dealing correctly with a non-positive return
2999 from the read() function. If the client had disconnected by this time,
3000 the result was a log entry for a synchronization error with an empty
3001 string after "input=" when read() returned zero. If read() returned -1
3002 (an event I could not check), uninitialized data bytes were printed.
3003 There were reports of junk text (parts of files, etc) appearing after
3006 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3008 PH/22 Added support for macro redefinition, and (re)definition in between
3009 driver and ACL definitions.
3011 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3012 forgetting to use the resulting value; it was using the unexpanded value.
3014 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3015 hadn't been configured. The fix is from Juergen Kreileder, who
3016 understands it better than I do:
3018 "Here's what I see happening with three configured cyrus_sasl
3019 authenticators configured (plain, login, cram-md5):
3021 On startup auth_cyrus_sasl_init() gets called for each of these.
3022 This means three calls to sasl_listmech() without a specified mech_list.
3023 => SASL tests which mechs of all available mechs actually work
3024 => three warnings about OTP not working
3025 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3027 With the patch, sasl_listmech() also gets called three times. But now
3028 SASL's mech_list option is set to the server_mech specified in the the
3029 authenticator. Or in other words, the answer from sasl_listmech()
3030 gets limited to just the mech you're testing for (which is different
3032 => the return list contains just 'plain' or 'login', 'cram-md5' or
3033 nothing depending on the value of ob->server_mech.
3035 I've just tested the patch: Authentication still works fine,
3036 unavailable mechs specified in the exim configuration are still
3037 caught, and the auth.log warnings about OTP are gone."
3039 PH/25 When debugging is enabled, the contents of the command line are added
3040 to the debugging output, even when log_selector=+arguments is not
3043 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3044 answer is "GNU", and only if the return is "GNU/something" is the answer
3047 PH/27 $acl_verify_message is now set immediately after the failure of a
3048 verification in an ACL, and so is available in subsequent modifiers. In
3049 particular, the message can be preserved by coding like this:
3051 warn !verify = sender
3052 set acl_m0 = $acl_verify_message
3054 Previously, $acl_verify_message was set only while expanding "message"
3055 and "log_message" when a very denied access.
3057 PH/28 Modified OS/os.c-Linux with
3059 -#ifndef OS_LOAD_AVERAGE
3060 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3062 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3063 nomenclature these days.)
3065 PH/29 Installed patch from the Sieve maintainer that adds the options
3066 sieve_useraddress and sieve_subaddress to the redirect router.
3068 PH/30 In these circumstances:
3069 . Two addresses routed to the same list of hosts;
3070 . First host does not offer TLS;
3071 . First host accepts first address;
3072 . First host gives temporary error to second address;
3073 . Second host offers TLS and a TLS session is established;
3074 . Second host accepts second address.
3075 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3076 and peerdn, if requested) that were in fact used only for the second
3079 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3080 attention to any local part prefix or suffix that was matched by the
3081 router that accepted the address. It now behaves in the same way as it
3082 does for delivery: the affixes are removed from the local part unless
3083 rcpt_include_affixes is set on the transport.
3085 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3086 timeout during the DATA phase of an incoming message.
3088 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3089 applied a patch sanctioned by the Sieve maintainer.
3091 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3092 the uid or gid is negative. A case of a negative gid caused this to be
3093 noticed. The fix allows for either to be negative.
3095 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3096 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3099 PH/36 The change PH/12 above was broken. Fixed it.
3101 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3102 the grounds that routing the same address twice would always produce the
3103 same answer. This might have been true once, but it is certainly no
3104 longer true now. Routing a child address may depend on the previous
3105 routing that produced that child. Some complicated redirection strategies
3106 went wrong when messages had multiple recipients, and made Exim's
3107 behaviour dependent on the order in which the addresses were given.
3109 I have moved the duplicate checking until after the routing is complete.
3110 Exim scans the addresses that are assigned to local and remote
3111 transports, and removes any duplicates. This means that more work will be
3112 done, as duplicates will always all be routed, but duplicates are
3113 presumably rare, so I don't expect this is of any significance.
3115 For deliveries to pipes, files, and autoreplies, the duplicate checking
3116 still happens during the routing process, since they are not going to be
3119 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3120 It corrects a timeout issue with spamd. This is Ian's comment: "The
3121 background is that sometimes spamd either never reads data from a
3122 connection it has accepted, or it never writes response data. The exiscan
3123 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3124 blindly assumes that writes won't block so it may never time out."
3126 PH/39 Allow G after quota size as well as K and M.
3128 PH/40 The value set for $authenticated_id in an authenticator may not contain
3129 binary zeroes or newlines because the value is written to log lines and
3130 to spool files. There was no check on this. Now the value is run through
3131 the string_printing() function so that such characters are converted to
3132 printable escape sequences.
3134 PH/41 $message_linecount is a new variable that contains the total number of
3135 lines in the message. Compare $body_linecount, which is the count for the
3138 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3139 bounce and delay warning messages, except in certain special cases, which
3142 (a) An SMTP error message from a remote host;
3143 (b) A message specified in a :fail: redirection;
3144 (c) A message specified in a "fail" command in a system filter;
3145 (d) A message specified in a FAIL return from the queryprogram router;
3146 (e) A message specified by the cannot_route_message router option.
3148 In these cases only, Exim does include the error details in bounce and
3149 warning messages. There are also a few cases where bland messages such
3150 as "unrouteable address" or "local delivery error" are given.
3152 PH/43 $value is now also set for the "else" part of a ${run expansion.
3154 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3155 being worked on, but at least Exim now implements the latest version to
3158 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3159 process to complete was treated as a delivery failure, a timeout while
3160 writing the message to the pipe was logged, but erroneously treated as a
3161 successful delivery. Such timeouts include transport filter timeouts. For
3162 consistency with the overall process timeout, these timeouts are now
3163 treated as errors, giving rise to delivery failures by default. However,
3164 there is now a new Boolean option for the pipe transport called
3165 timeout_defer, which, if set TRUE, converts the failures into defers for
3166 both kinds of timeout. A transport filter timeout is now identified in
3169 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3170 systems where "make" and "gmake" are different, calling "gmake" at top
3171 level broke things. I've arranged for the value of $(MAKE) to be passed
3172 from the Makefile to this script so that it can call the same version of
3176 A note about Exim versions 4.44 and 4.50
3177 ----------------------------------------
3179 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3180 changes of various kinds. As a consequence, a big documentation update was
3181 needed. This delayed the release for rather longer than seemed good, especially
3182 in the light of a couple of (minor) security issues. Therefore, the changes
3183 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3184 release. So 4.44 and 4.50 are in effect two different branches that both start
3187 I have left the 4.50 change log unchanged; it contains all the changes since
3188 4.43. The change log for 4.44 is below; many of its items are identical to
3189 those for 4.50. This seems to be the most sensible way to preserve the
3190 historical information.
3196 1. Minor wording change to the doc/README.SIEVE file.
3198 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3199 computation of the current number of files was incorrect.
3201 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3202 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3205 4. Give more explanation in the error message when the command for a transport
3206 filter fails to execute.
3208 5. There are several places where Exim runs a non-Exim command in a
3209 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3210 was being done only for the command run by the queryprogram router. It is
3211 now done for all such subprocesses. The other cases are: ${run, transport
3212 filters, and the commands run by the lmtp and pipe transports.
3214 6. Added CONFIGURE_GROUP build-time option.
3216 7. Some older OS have a limit of 256 on the maximum number of file
3217 descriptors. Exim was using setrlimit() to set 1000 as a large value
3218 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3219 systems. I've change it so that if it can't get 1000, it tries for 256.
3221 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3222 was an oversight, and furthermore, ever since the addition of extra
3223 controls (e.g. 4.43/32), the checks on when to allow different forms of
3224 "control" were broken. There should now be diagnostics for all cases when a
3225 control that does not make sense is encountered.
3227 9. Added the /retain_sender option to "control=submission".
3229 10. $recipients is now available in the predata ACL (oversight).
3231 11. Tidy the search cache before the fork to do a delivery from a message
3232 received from the command line. Otherwise the child will trigger a lookup
3233 failure and thereby defer the delivery if it tries to use (for example) a
3234 cached ldap connection that the parent has called unbind on.
3236 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3237 of $address_data from the recipient verification was clobbered by the
3238 sender verification.
3240 13. The value of address_data from a sender verification is now available in
3241 $sender_address_data in subsequent conditions in the ACL statement.
3243 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3245 15. Added a new option "connect=<time>" to callout options, to set a different
3248 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3249 was its contents. (It was OK if the option was not defined at all.)
3251 17. A "Completed" log line is now written for messages that are removed from
3252 the spool by the -Mrm option.
3254 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3255 information about exactly what failed.
3257 19. Added -dd to debug only the daemon process.
3259 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3260 handles timeouts, both on the server side and network timeouts. Renamed the
3261 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3263 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3264 It is now set to "smtps".
3266 22. $host_address is now set to the target address during the checking of
3267 ignore_target_hosts.
3269 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3270 being passed; this would have caused $sender_host_name to have been used if
3271 matching the list had actually called for a host name (not very likely,
3272 since this list is usually IP addresses). A host name is now passed as
3275 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3276 code that specifies a non-restarting handler (typically sigaction() in
3277 modern systems) in an attempt to fix a rare and obscure crash bug.
3279 25. Narrowed the window for a race in the daemon that could cause it to ignore
3280 SIGCHLD signals. This is not a major problem, because they are used only to
3281 wake it up if nothing else does.
3283 26. A malformed maildirsize file could cause Exim to calculate negative values
3284 for the mailbox size or file count. Odd effects could occur as a result.
3285 The maildirsize information is now recalculated if the size or filecount
3288 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3289 support for a long time. Removed HAVE_SYS_VFS_H.
3291 28. Installed the latest version of exipick from John Jetmore.
3293 29. In an address list, if the pattern was not a regular expression, an empty
3294 subject address (from a bounce message) matched only if the pattern was an
3295 empty string. Non-empty patterns were not even tested. This was the wrong
3296 because it is perfectly reasonable to use an empty address as part of a
3297 database query. An empty address is now tested by patterns that are
3298 lookups. However, all the other forms of pattern expect the subject to
3299 contain a local part and a domain, and therefore, for them, an empty
3300 address still always fails if the pattern is not itself empty.
3302 30. Exim went into a mad DNS loop when attempting to do a callout where the
3303 host was specified on an smtp transport, and looking it up yielded more
3304 than one IP address.
3306 31. Re-factored the code for checking spool and log partition space into a
3307 function that finds that data and another that does the check. The former
3308 is then used to implement four new variables: $spool_space, $log_space,
3309 $spool_inodes, and $log_inodes.
3311 32. The RFC2047 encoding function was originally intended for short strings
3312 such as real names; it was not keeping to the 75-character limit for
3313 encoded words that the RFC imposes. It now respects the limit, and
3314 generates multiple encoded words if necessary. To be on the safe side, I
3315 have increased the buffer size for the ${rfc2047: expansion operator from
3318 33. It is now permitted to omit both strings after an "if" condition; if the
3319 condition is true, the result is "true". As before, when the second string
3320 is omitted, a false condition yields an empty string. This makes it less
3321 cumbersome to write custom ACL and router conditions.
3323 34. Failure to deliver a bounce message always caused it to be frozen, even if
3324 there was an errors_to setting on the router. The errors_to setting is now
3327 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3328 canonical form (fully expanded) before being placed in
3329 $sender_host_address.
3331 36. The table in the code that translates DNS record types into text (T_A to
3332 "A" for instance) was missing entries for NS and CNAME. It is just possible
3333 that this could have caused confusion if both these types were looked up
3334 for the same domain, because the text type is used as part of Exim's
3335 per-process caching. But the chance of anyone hitting this buglet seems
3338 37. The dnsdb lookup has been extended in a number of ways.
3340 (1) There is a new type, "zns", which walks up the domain tree until it
3341 finds some nameserver records. It should be used with care.
3343 (2) There is a new type, "mxh", which is like "mx" except that it returns
3344 just the host names, not the priorities.
3346 (3) It is now possible to give a list of domains (or IP addresses) to be
3347 looked up. The behaviour when one of the lookups defers can be
3348 controlled by a keyword.
3350 (4) It is now possible to specify the separator character for use when
3351 multiple records are returned.
3353 38. The dnslists ACL condition has been extended: it is now possible to supply
3354 a list of IP addresses and/or domains to be looked up in a particular DNS
3357 39. Added log_selector=+queue_time_overall.
3359 40. When running the queue in the test harness, wait just a tad after forking a
3360 delivery process, to get repeatability of debugging output.
3362 41. Include certificate and key file names in error message when GnuTLS fails
3363 to set them up, because the GnuTLS error message doesn't include the name
3364 of the failing file when there is a problem reading it.
3366 42. Allow both -bf and -bF in the same test run.
3368 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3370 44. The "Exiscan patch" is now merged into the mainline Exim source.
3372 45. Sometimes the final signoff response after QUIT could fail to get
3373 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3374 < 0 before doing a fflush(). This bug looks as though it goes back to the
3375 introduction of TLS in release 3.20, but "sometimes" must have been rare
3376 because the tests only now provoked it.
3378 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3379 (this can affect the format of dates).
3381 47. exim_tidydb, when checking for the continued existence of a message for
3382 which it has found a message-specific retry record, was not finding
3383 messages that were in split spool directories. Consequently, it was
3384 deleting retry records that should have stayed in existence.
3386 48. Steve fixed some bugs in eximstats.
3388 49. The SPA authentication driver was not abandoning authentication and moving
3389 on to the next authenticator when an expansion was forced to fail,
3390 contradicting the general specification for all authenticators. Instead it
3391 was generating a temporary error. It now behaves as specified.
3393 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3394 (the order specifies the preference for clients). The order is now AES256,
3395 AES128, 3DES, ARCFOUR128.
3397 51. Small patch to Sieve code - explicitly set From: when generating an
3400 52. Exim crashed if a remote delivery caused a very long error message to be
3401 recorded - for instance if somebody sent an entire SpamAssassin report back
3402 as a large number of 550 error lines. This bug was coincidentally fixed by
3403 increasing the size of one of Exim's internal buffers (big_buffer) that
3404 happened as part of the Exiscan merge. However, to be on the safe side, I
3405 have made the code more robust (and fixed the comments that describe what
3408 53. Now that there can be additional text after "Completed" in log lines (if
3409 the queue_time_overall log selector is set), a one-byte patch to exigrep
3410 was needed to allow it to recognize "Completed" as not the last thing in
3413 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3414 patch that reportedly fixes this has been added. I am not expert enough to
3415 create a test for it. This is what the patch creator wrote:
3417 "I found a little strange behaviour of ldap code when working with
3418 Windows 2003 AD Domain, where users was placed in more than one
3419 Organization Units. When I tried to give exim partial DN, the exit code
3420 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3421 But simultaneously result of request was absolutely normal ldap result,
3422 so I produce this patch..."
3424 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3425 REFERENCE, so I have modified the code to exclude the patch when that macro
3428 55. Some experimental protocols are using DNS PTR records for new purposes. The
3429 keys for these records are domain names, not reversed IP addresses. The
3430 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3431 leaves it alone. Component reversal etc. now happens only for IP addresses.
3434 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3436 57. Double the size of the debug message buffer (to 2048) so that more of very
3437 long debug lines gets shown.
3439 58. The exicyclog utility now does better if the number of log files to keep
3440 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3442 59. Two changes related to the smtp_active_hostname option:
3444 (1) $smtp_active_hostname is now available as a variable.
3445 (2) The default for smtp_banner uses $smtp_active_hostname instead
3446 of $primary_hostname.
3448 60. The host_aton() function is supposed to be passed a string that is known
3449 to be a valid IP address. However, in the case of IPv6 addresses, it was
3450 not checking this. This is a hostage to fortune. Exim now panics and dies
3451 if the condition is not met. A case was found where this could be provoked
3452 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3453 components; fortuitously, this particular loophole had already been fixed
3454 by change 4.50/55 above.
3456 If there are any other similar loopholes, the new check in host_aton()
3457 itself should stop them being exploited. The report I received stated that
3458 data on the command line could provoke the exploit when Exim was running as
3459 exim, but did not say which command line option was involved. All I could
3460 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3461 running as the user.
3464 61. There was a buffer overflow vulnerability in the SPA authentication code
3465 (which came originally from the Samba project). I have added a test to the
3466 spa_base64_to_bits() function which I hope fixes it.
3469 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3470 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3472 63. The daemon start-up calls getloadavg() while still root for those OS that
3473 need the first call to be done as root, but it missed one case: when
3474 deliver_queue_load_max is set with deliver_drop_privilege. This is
3475 necessary for the benefit of the queue runner, because there is no re-exec
3476 when deliver_drop_privilege is set.
3478 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3479 This has been fixed.
3481 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3482 was in use, was not putting the data itself into the right store pool;
3483 consequently, it could be overwritten for a subsequent message in the same
3484 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3487 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3489 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3490 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3491 the function were treating the return as a boolean value, which happened to
3492 work because 0=false and not-0=true, but is not correct code.
3494 68. The host_aton() function was not handling scoped IPv6 addresses (those
3495 with, for example, "%eth0" on the end) correctly.
3497 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3498 negated items (that is, ~something) in unsigned ints. Some compilers
3499 apparently mutter when there is no cast.
3501 70. If an address verification called from an ACL failed, and did not produce a
3502 user-specific message (i.e. there was only a "system" message), nothing was
3503 put in $acl_verify_message. In this situation, it now puts the system
3506 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3507 session; change 4.31/43 added the unwanted input to the log line - except
3508 that it did not do this in the start of session case. It now does.
3510 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3511 This is wrong and can cause the other end to generate a synchronization
3512 error if it is another Exim or anything else that does the synchronization
3513 check. A QUIT command is no longer sent after a timeout.
3515 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3516 during host lookups.
3518 74. The defer_ok option of callout verification was not working if it was used
3519 when verifying addresses in header lines, that is, for this case:
3521 verify = header_sender/callout=defer_ok
3523 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3524 those file descriptors could be used for SMTP connections. If anything
3525 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3526 could be sent to the SMTP client, causing chaos. The daemon now opens
3527 stdin, stdout, and stderr to /dev/null when it puts itself into the
3530 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3531 log by default. The user can override this with suitable Perl magic.
3533 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3534 the log message when discard triggers, was not working for the DATA ACL or
3535 for the non-SMTP ACL.
3537 78. Error message wording change in sieve.c.
3539 79. If smtp_accept_max_per_host was set, the number of connections could be
3540 restricted to fewer than expected, because the daemon was trying to set up
3541 a new connection before checking whether the processes handling previous
3542 connections had finished. The check for completed processes is now done
3543 earlier. On busy systems, this bug wouldn't be noticed because something
3544 else would have woken the daemon, and it would have reaped the completed
3547 80. If a message was submitted locally by a user whose login name contained one
3548 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3549 It caused a spool format error. I have fixed the spool reading code. A
3550 related problem was that the "from" clause in the Received: line became
3551 illegal because of the space(s). It is now covered by ${quote_local_part.
3553 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3556 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3558 83. Patch from Sieve maintainer to fix unterminated string problem in
3559 "vacation" handling.
3561 84. Some minor changes to the Linux configuration files to help with other
3562 OS variants using glibc.
3564 85. One more patch for Sieve to update vacation handling to latest spec.
3567 ----------------------------------------------------
3568 See the note above about the 4.44 and 4.50 releases.
3569 ----------------------------------------------------
3575 1. Change 4.43/35 introduced a bug that caused file counts to be
3576 incorrectly computed when quota_filecount was set in an appendfile
3579 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3580 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3583 3. Give more explanation in the error message when the command for a transport
3584 filter fails to execute.
3586 4. There are several places where Exim runs a non-Exim command in a
3587 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3588 was being done only for the command run by the queryprogram router. It is
3589 now done for all such subprocesses. The other cases are: ${run, transport
3590 filters, and the commands run by the lmtp and pipe transports.
3592 5. Some older OS have a limit of 256 on the maximum number of file
3593 descriptors. Exim was using setrlimit() to set 1000 as a large value
3594 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3595 systems. I've change it so that if it can't get 1000, it tries for 256.
3597 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3598 was an oversight, and furthermore, ever since the addition of extra
3599 controls (e.g. 4.43/32), the checks on when to allow different forms of
3600 "control" were broken. There should now be diagnostics for all cases when a
3601 control that does not make sense is encountered.
3603 7. $recipients is now available in the predata ACL (oversight).
3605 8. Tidy the search cache before the fork to do a delivery from a message
3606 received from the command line. Otherwise the child will trigger a lookup
3607 failure and thereby defer the delivery if it tries to use (for example) a
3608 cached ldap connection that the parent has called unbind on.
3610 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3611 of $address_data from the recipient verification was clobbered by the
3612 sender verification.
3614 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3615 was its contents. (It was OK if the option was not defined at all.)
3617 11. A "Completed" log line is now written for messages that are removed from
3618 the spool by the -Mrm option.
3620 12. $host_address is now set to the target address during the checking of
3621 ignore_target_hosts.
3623 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3624 being passed; this would have caused $sender_host_name to have been used if
3625 matching the list had actually called for a host name (not very likely,
3626 since this list is usually IP addresses). A host name is now passed as
3629 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3630 code that specifies a non-restarting handler (typically sigaction() in
3631 modern systems) in an attempt to fix a rare and obscure crash bug.
3633 15. Narrowed the window for a race in the daemon that could cause it to ignore
3634 SIGCHLD signals. This is not a major problem, because they are used only to
3635 wake it up if nothing else does.
3637 16. A malformed maildirsize file could cause Exim to calculate negative values
3638 for the mailbox size or file count. Odd effects could occur as a result.
3639 The maildirsize information is now recalculated if the size or filecount
3642 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3643 support for a long time. Removed HAVE_SYS_VFS_H.
3645 18. Updated exipick to current release from John Jetmore.
3647 19. Allow an empty sender to be matched against a lookup in an address list.
3648 Previously the only cases considered were a regular expression, or an
3651 20. Exim went into a mad DNS lookup loop when doing a callout where the
3652 host was specified on the transport, if the DNS lookup yielded more than
3655 21. The RFC2047 encoding function was originally intended for short strings
3656 such as real names; it was not keeping to the 75-character limit for
3657 encoded words that the RFC imposes. It now respects the limit, and
3658 generates multiple encoded words if necessary. To be on the safe side, I
3659 have increased the buffer size for the ${rfc2047: expansion operator from
3662 22. Failure to deliver a bounce message always caused it to be frozen, even if
3663 there was an errors_to setting on the router. The errors_to setting is now
3666 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3667 canonical form (fully expanded) before being placed in
3668 $sender_host_address.
3670 24. Updated eximstats to version 1.33
3672 25. Include certificate and key file names in error message when GnuTLS fails
3673 to set them up, because the GnuTLS error message doesn't include the name
3674 of the failing file when there is a problem reading it.
3676 26. Expand error message when OpenSSL has problems setting up cert/key files.
3679 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3680 (this can affect the format of dates).
3682 28. exim_tidydb, when checking for the continued existence of a message for
3683 which it has found a message-specific retry record, was not finding
3684 messages that were in split spool directories. Consequently, it was
3685 deleting retry records that should have stayed in existence.
3687 29. eximstats updated to version 1.35
3688 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3689 1.35 - bugfix such that pie charts by volume are generated correctly
3691 30. The SPA authentication driver was not abandoning authentication and moving
3692 on to the next authenticator when an expansion was forced to fail,
3693 contradicting the general specification for all authenticators. Instead it
3694 was generating a temporary error. It now behaves as specified.
3696 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3697 (the order specifies the preference for clients). The order is now AES256,
3698 AES128, 3DES, ARCFOUR128.
3700 31. Small patch to Sieve code - explicitly set From: when generating an
3703 32. Exim crashed if a remote delivery caused a very long error message to be
3704 recorded - for instance if somebody sent an entire SpamAssassin report back
3705 as a large number of 550 error lines. This bug was coincidentally fixed by
3706 increasing the size of one of Exim's internal buffers (big_buffer) that
3707 happened as part of the Exiscan merge. However, to be on the safe side, I
3708 have made the code more robust (and fixed the comments that describe what
3711 33. Some experimental protocols are using DNS PTR records for new purposes. The
3712 keys for these records are domain names, not reversed IP addresses. The
3713 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3714 leaves it alone. Component reversal etc. now happens only for IP addresses.
3717 34. The host_aton() function is supposed to be passed a string that is known
3718 to be a valid IP address. However, in the case of IPv6 addresses, it was
3719 not checking this. This is a hostage to fortune. Exim now panics and dies
3720 if the condition is not met. A case was found where this could be provoked
3721 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3722 components; fortuitously, this particular loophole had already been fixed
3723 by change 4.50/55 or 4.44/33 above.
3725 If there are any other similar loopholes, the new check in host_aton()
3726 itself should stop them being exploited. The report I received stated that
3727 data on the command line could provoke the exploit when Exim was running as
3728 exim, but did not say which command line option was involved. All I could
3729 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3730 running as the user.
3733 35. There was a buffer overflow vulnerability in the SPA authentication code
3734 (which came originally from the Samba project). I have added a test to the
3735 spa_base64_to_bits() function which I hope fixes it.
3738 36. The daemon start-up calls getloadavg() while still root for those OS that
3739 need the first call to be done as root, but it missed one case: when
3740 deliver_queue_load_max is set with deliver_drop_privilege. This is
3741 necessary for the benefit of the queue runner, because there is no re-exec
3742 when deliver_drop_privilege is set.
3744 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3745 was in use, was not putting the data itself into the right store pool;
3746 consequently, it could be overwritten for a subsequent message in the same
3747 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3750 38. Sometimes the final signoff response after QUIT could fail to get
3751 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3752 < 0 before doing a fflush(). This bug looks as though it goes back to the
3753 introduction of TLS in release 3.20, but "sometimes" must have been rare
3754 because the tests only now provoked it.
3760 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3761 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3762 Now it returns an integer. A number of calls were still expecting a T/F
3763 return. Fortuitously, in all cases, the tests worked in OK situations,
3764 which is the norm. However, things would have gone wrong on any write
3765 failures on the smtp file descriptor. This function is used when sending
3766 messages over SMTP and also when doing verify callouts.
3768 2. When Exim is called to do synchronous delivery of a locally submitted
3769 message (the -odf or -odi options), it no longer closes stderr before doing
3772 3. Implemented the mua_wrapper option.
3774 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3776 5. Implemented the functions header_remove(), header_testname(),
3777 header_add_at_position(), and receive_remove_recipient(), and exported them
3780 6. If an ACL "warn" statement specified the addition of headers, Exim already
3781 inserted X-ACL-Warn: at the start if there was no header name. However, it
3782 was not making this test for the second and subsequent header lines if
3783 there were newlines in the string. This meant that an invalid header could
3784 be inserted if Exim was badly configured.
3786 7. Allow an ACL "warn" statement to add header lines at the start or after all
3787 the Received: headers, as well as at the end.
3789 8. Added the rcpt_4xx retry error code.
3791 9. Added postmaster_mailfrom=xxx to callout verification option.
3793 10. Added mailfrom=xxxx to the callout verification option, for verify=
3796 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3797 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3799 12. Inserted some casts to stop certain compilers warning when using pointer
3800 differences as field lengths or precisions in printf-type calls (mostly
3801 affecting debugging statements).
3803 13. Added optional readline() support for -be (dynamically loaded).
3805 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3806 same clock tick as a message's arrival, so that its received time was the
3807 same as the "first fail" time on the retry record, and that message
3808 remained on the queue past the ultimate address timeout, every queue runner
3809 would try a delivery (because it was past the ultimate address timeout) but
3810 after another failure, the ultimate address timeout, which should have then
3811 bounced the address, did not kick in. This was a "< instead of <=" error;
3812 in most cases the first failure would have been in the next clock tick
3813 after the received time, and all would be well.
3815 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3816 being recognized when the domain list was tested by the match_domain
3817 condition in an expansion string.
3819 16. Added the ${str2b64: operator.
3821 17. Exim was always calling setrlimit() to set a large limit for the number of
3822 processes, without checking whether the existing limit was already
3823 adequate. (It did check for the limit on file descriptors.) Furthermore,
3824 errors from getrlimit() and setrlimit() were being ignored. Now they are
3825 logged to the main and panic logs, but Exim does carry on, to try to do its
3826 job under whatever limits there are.
3828 18. Imported PCRE 5.0.
3830 19. Trivial typo in log message " temporarily refused connection" (the leading
3833 20. If the log selector return_path_on_delivery was set and an address was
3834 redirected to /dev/null, the delivery process crashed because it assumed
3835 that a return path would always be set for a "successful" delivery. In this
3836 case, the whole delivery is bypassed as an optimization, and therefore no
3839 21. Internal re-arrangement: the function for sending a challenge and reading
3840 a response while authentication was assuming a zero-terminated challenge
3841 string. It's now changed to take a pointer and a length, to allow for
3842 binary data in such strings.
3844 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3846 23. Exim was not respecting finduser_retries when seeking the login of the
3847 uid under which it was called; it was always trying 10 times. (The default
3848 setting of finduser_retries is zero.) Also, it was sleeping after the final
3849 failure, which is pointless.
3851 24. Implemented tls_on_connect_ports.
3853 25. Implemented acl_smtp_predata.
3855 26. If the domain in control=submission is set empty, Exim assumes that the
3856 authenticated id is a complete email address when it generates From: or
3857 Sender: header lines.
3859 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3860 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3861 chown and chgrp in /bin and hostname in /usr/bin.
3863 28. Exim was keeping the "process log" file open after each use, just as it
3864 does for the main log. This opens the possibility of it remaining open for
3865 long periods when the USR1 signal hits a daemon. Occasional processlog
3866 errors were reported, that could have been caused by this. Anyway, it seems
3867 much more sensible not to leave this file open at all, so that is what now
3870 29. The long-running daemon process does not normally write to the log once it
3871 has entered its main loop, and it closes the log before doing so. This is
3872 so that log files can straightforwardly be renamed and moved. However,
3873 there are a couple of unusual error situations where the daemon does write
3874 log entries, and I had neglected to close the log afterwards.
3876 30. The text of an SMTP error response that was received during a remote
3877 delivery was being truncated at 512 bytes. This is too short for some of
3878 the long messages that one sometimes sees. I've increased the limit to
3881 31. It is now possible to make retry rules that apply only when a message has a
3882 specific sender, in particular, an empty sender.
3884 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3885 it possible to be selective about when SMTP synchronization is enforced.
3887 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3889 32. Implemented hosts_connection_nolog.
3891 33. Added an ACL for QUIT.
3893 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3896 35. Added mailbox_size and mailbox_filecount to appendfile.
3898 36. Added control = no_multiline_responses to ACLs.
3900 37. There was a bug in the logic of the code that waits for the clock to tick
3901 in the case where the clock went backwards by a substantial amount such
3902 that the microsecond fraction of "now" was more than the microsecond
3903 fraction of "then" (but the whole seconds number was less).
3905 38. Added support for the libradius Radius client library this is found on
3906 FreeBSD (previously only the radiusclient library was supported).
3912 1. When certain lookups returned multiple values in the form name=value, the
3913 quoting of the values was not always being done properly. Specifically:
3914 (a) If the value started with a double quote, but contained no whitespace,
3916 (b) If the value contained whitespace other than a space character (i.e.
3917 tabs or newlines or carriage returns) it was not quoted.
3918 This fix has been applied to the mysql and pgsql lookups by writing a
3919 separate quoting function and calling it from the lookup code. The fix
3920 should probably also be applied to nisplus, ibase and oracle lookups, but
3921 since I cannot test any of those, I have not disturbed their existing code.
3923 2. A hit in the callout cache for a specific address caused a log line with no
3924 reason for rejecting RCPT. Now it says "Previous (cached) callout
3925 verification failure".
3927 3. There was an off-by-one bug in the queryprogram router. An over-long
3928 return line was truncated at 256 instead of 255 characters, thereby
3929 overflowing its buffer with the terminating zero. As well as fixing this, I
3930 have increased the buffer size to 1024 (and made a note to document this).
3932 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3933 when Exim is waiting for an SMTP response from a remote server, Exim
3934 restarts its select() call on the socket, thereby resetting its timeout.
3935 This is not a problem when such interrupts are rare. Somebody set up a cron
3936 job to run exiwhat every 2 minutes, which is less than the normal select()
3937 timeout (5 or 10 minutes). This meant that the select() timeout never
3938 kicked in because it was always reset. I have fixed this by comparing the
3939 time when an interrupt arrives with the time at the start of the first call
3940 to select(). If more time than the timeout has elapsed, the interrupt is
3941 treated as a timeout.
3943 5. Some internal re-factoring in preparation for the addition of Sieve
3944 extensions (by MH). In particular, the "personal" test is moved to a
3945 separate function, and given an option for scanning Cc: and Bcc: (which is
3946 not set for Exim filters).
3948 6. When Exim created an email address using the login of the caller as the
3949 local part (e.g. when creating a From: or Sender: header line), it was not
3950 quoting the local part when it contained special characters such as @.
3952 7. Installed new OpenBSD configuration files.
3954 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3955 try to make them clearer.
3957 9. Callout options, other than the timeout value, were being ignored when
3958 verifying sender addresses in header lines. For example, when using
3960 verify = header_sender/callout=no_cache
3962 the cache was (incorrectly) being used.
3964 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3965 only the Cygwin environment.
3967 11. When return_path_on_delivery was set as a log selector, if different remote
3968 addresses in the same message used different return paths and parallel
3969 remote delivery occurred, the wrong values would sometimes be logged.
3970 (Whenever a remote delivery process finished, the return path value from
3971 the most recently started remote delivery process was logged.)
3973 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3974 lines when AUTH and/or TLS are in use. This is the "received protocol"
3975 field. Exim used to use "asmtp" for authenticated SMTP, without any
3976 indication (in the protocol name) for TLS use. Now it follows the RFC and
3977 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3978 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3979 names appear in log lines as well as in Received: header lines.
3981 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3982 extensions, and comparison tests, and to fix some bugs.
3984 14. Changes to the "personal" filter test:
3986 (1) The test was buggy in that it was just doing the equivalent of
3987 "contains" tests on header lines. For example, if a user's address was
3988 anne@some.where, the "personal" test would incorrectly be true for
3990 To: susanne@some.where
3992 This test is now done by extracting each address from the header in turn,
3993 and checking the entire address. Other tests that are part of "personal"
3994 are now done using regular expressions (for example, to check local parts
3995 of addresses in From: header lines).
3997 (2) The list of non-personal local parts in From: addresses has been
3998 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3999 taken from the Sieve specification recommendations.
4001 (3) If the message contains any header line starting with "List-" it is
4002 treated as non-personal.
4004 (4) The test for "circular" in the Subject: header line has been removed
4005 because it now seems ill-conceived.
4007 15. Minor typos in src/EDITME comments corrected.
4009 16. Installed latest exipick from John Jetmore.
4011 17. If headers_add on a router specified a text string that was too long for
4012 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4013 of string_sprintf() is now avoided.
4015 18. $message_body_size was not set (it was always zero) when running the DATA
4016 ACL and the local_scan() function.
4018 19. For the "mail" command in an Exim filter, no default was being set for
4019 the once_repeat time, causing a random time value to be used if "once" was
4020 specified. (If the value happened to be <= 0, no repeat happened.) The
4021 default is now 0s, meaning "never repeat". The "vacation" command was OK
4022 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4023 (I found it when inspecting the code).
4025 20. There is now an overall timeout for performing a callout verification. It
4026 defaults to 4 times the callout timeout, which applies to individual SMTP
4027 commands during the callout. The overall timeout applies when there is more
4028 than one host that can be tried. The timeout is checked before trying the
4029 next host. This prevents very long delays if there are a large number of
4030 hosts and all are timing out (e.g. when the network connections are timing
4031 out). The value of the overall timeout can be changed by specifying an
4032 additional sub-option for "callout", called "maxwait". For example:
4034 verify = sender/callout=5s,maxwait=20s
4036 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4037 to the end before writing, but this should make it even safer).
4039 22. Exim was forgetting that it had advertised PIPELINING for the second and
4040 subsequent messages on an SMTP connection. It was also not resetting its
4041 memory on STARTTLS and an internal HELO.
4043 23. When Exim logs an SMTP synchronization error within a session, it now
4044 records whether PIPELINING has been advertised or not.
4046 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4047 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4048 rather than long int.
4050 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4052 26. Added the never_mail option to autoreply.
4058 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4059 crash if the getsockname() call failed; this can happen if a connection is
4060 closed very soon after it is established. The problem was simply in the
4061 order in which certain operations were done, causing Exim to try to write
4062 to the SMTP stream before it had set up the file descriptor. The bug has
4063 been fixed by making things happen in the correct order.
4069 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4070 before the connection was closed, thus losing the rejection response.
4072 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4073 some early Solaris releases, but causes trouble in current releases where
4074 socklen_t is defined.
4076 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4079 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4082 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4083 that releasing the top of it at the end releases what was used for sub-
4084 expansions (unless the block got too big). However, discard this block if
4085 the first thing is a variable or header, so that we can use its block when
4086 it is dynamic (useful for very large $message_headers, for example).
4088 6. Lookups now cache *every* query, not just the most recent. A new, separate
4089 store pool is used for this. It can be recovered when all lookup caches are
4090 flushed. Lookups now release memory at the end of their result strings.
4091 This has involved some general refactoring of the lookup sources.
4093 7. Some code has been added to the store_xxx() functions to reduce the amount
4094 of flapping under certain conditions.
4096 8. log_incoming_interface used to affect only the <= reception log lines. Now
4097 it causes the local interface and port to be added to several more SMTP log
4098 lines, for example "SMTP connection from", and rejection lines.
4100 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4102 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4104 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4105 could be overwritten at the end of the current message (or the start of a
4106 new message if it was set in a HELO ACL). The value is now preserved for
4107 the duration of the SMTP connection.
4109 12. If a transport had a headers_rewrite setting, and a matching header line
4110 contained an unqualified address, that address was qualified, even if it
4111 did not match any rewriting rules. The underlying bug was that the values
4112 of the flags that permit the existence of unqualified sender and recipient
4113 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4114 non-local messages, and by -bnq for local messages) were not being
4115 preserved with the message after it was received.
4117 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4118 "next input=" as part of the text comprising the host identity instead of
4119 the correct text. The code was using the same buffer for two different
4120 strings. However, depending on which order the printing function evaluated
4121 its arguments, the bug did not always show up. Under Linux, for example, my
4122 test suite worked just fine.
4124 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4125 doesn't work with some older versions of Perl. It has been changed to "my",
4126 which in any case is probably the better facility to use.
4128 15. A really picky compiler found some instances of statements for creating
4129 error messages that either had too many or two few arguments for the format
4132 16. The size of the buffer for calls to the DNS resolver has been increased
4133 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4134 for addresses that have a lot of PTR records. This alleviates a problem; it
4135 does not fully solve it.
4137 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4138 buffer now truncates the list and logs the incident, which is the same
4139 action as happens when Exim is looking up a host name and its aliases.
4140 Previously in this situation something unpredictable would happen;
4141 sometimes it was "internal error: store_reset failed".
4143 18. If a server dropped the connection unexpectedly when an Exim client was
4144 using GnuTLS and trying to read a response, the client delivery process
4145 crashed while trying to generate an error log message.
4147 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4148 string, for example:
4150 warn message = H1: something\nH2: something
4152 the text was added as a single header line from Exim's point of view
4153 though it ended up OK in the delivered message. However, searching for the
4154 second and subsequent header lines using $h_h2: did not work. This has been
4155 fixed. Similarly, if a system filter added multiple headers in this way,
4156 the routers could not see them.
4158 20. Expanded the error message when iplsearch is called with an invalid key to
4159 suggest using net-iplsearch in a host list.
4161 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4162 ACLs are no longer actually imposed (and a message to that effect is
4165 22. If a "gecos" field in a passwd entry contained escaped characters, in
4166 particular, if it contained a \" sequence, Exim got it wrong when building
4167 a From: or a Sender: header from that name. A second bug also caused
4168 incorrect handling when an unquoted " was present following a character
4169 that needed quoting.
4171 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4172 was not being matched caselessly.
4174 24. Arranged for all hyphens in the exim.8 source to be escaped with
4177 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4178 properly. Recipient callout cache records were still being keyed to include
4179 the sender, even when use_sender was set false. This led to far more
4180 callouts that were necessary. The sender is no longer included in the key
4181 when use_sender is false.
4183 26. Added "control = submission" modifier to ACLs.
4185 27. Added the ${base62d: operator to decode base 62 numbers.
4187 28. dnsdb lookups can now access SRV records.
4189 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4190 the configuration file.
4192 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4193 (-v) mode. This makes the output for a verbose queue run more intelligible.
4195 31. Added a use_postmaster feature to recipient callouts.
4197 32. Added the $body_zerocount variable, containing the number of binary zero
4198 bytes in the message body.
4200 33. The time of last modification of the "new" subdirectory is now used as the
4201 "mailbox time last read" when there is a quota error for a maildir
4204 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4206 35. Added +ignore_unknown as a special item in host lists.
4208 36. Code for decoding IPv6 addresses in host lists is now included, even if
4209 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4210 address was recognized as an IP address, but was then not correctly decoded
4211 into binary, causing unexpected and incorrect effects when compared with
4218 1. Very minor rewording of debugging text in manualroute to say "list of
4219 hosts" instead of "hostlist".
4221 2. If verify=header_syntax was set, and a header line with an unqualified
4222 address (no domain) and a large number of spaces between the end of the
4223 name and the colon was received, the reception process suffered a buffer
4224 overflow, and (when I tested it) crashed. This was caused by some obsolete
4225 code that should have been removed. The fix is to remove it!
4227 3. When running in the test harness, delay a bit after writing a bounce
4228 message to get a bit more predictability in the log output.
4230 4. Added a call to search_tidyup() just before forking a reception process. In
4231 theory, someone could use a lookup in the expansion of smtp_accept_max_
4232 per_host which, without the tidyup, could leave open a database connection.
4234 5. Added the variables $recipient_data and $sender_data which get set from a
4235 lookup success in an ACL "recipients" or "senders" condition, or a router
4236 "senders" option, similar to $domain_data and $local_part_data.
4238 6. Moved the writing of debug_print from before to after the "senders" test
4241 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4242 problems for message scanning, either using a data ACL, or using
4243 local_scan() because the Received: header was not generated till after they
4244 were called (in order to set the time as the time of reception completion).
4245 I have revised the way this works. The header is now generated after the
4246 body is received, but before the ACL or local_scan() are called. After they
4247 are run, the timestamp in the header is updated.
4253 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4254 before starting a queue runner without re-exec. This happened only when
4255 deliver_drop_privilege was set or when the Exim user was set to root. The
4256 effect of the bug was that timeouts during subsequent deliveries caused
4257 crashes instead of being properly handled. The handler is now left at its
4258 default (and expected) setting.
4260 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4261 message, again when deliver_drop_privilege is set or Exim is run as root.
4262 The bug described in (1) was not present in this case, but the tidying up
4263 of the other signals was missing. I have made the two cases consistent.
4265 3. The ignore_target_hosts setting on a manualroute router was being ignored
4266 for hosts that were looked up using the /MX notation.
4268 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4271 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4272 operated on the sender address. After changing the $sender_address to <>
4273 for the sender address verify, Exim was re-instated it as the original
4274 (before rewriting) address, but remembering that it had rewritten it, so it
4275 wasn't rewriting it again. This bug also had the effect of breaking the
4276 sender address verification caching when the sender address was rewritten.
4278 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4279 This has been changed so that if the ip literal address matches
4280 ignore_target_hosts, the router declines.
4282 7. Added expansion conditions match_domain, match_address, and match_local_
4283 part (NOT match_host).
4285 8. The placeholder for the Received: header didn't have a length field set.
4287 9. Added code to Exim itself and to exim_lock to test for a specific race
4288 condition that could lead to file corruption when using MBX delivery. The
4289 issue is with the lockfile that is created in /tmp. If this file is removed
4290 after a process has opened it but before that process has acquired a lock,
4291 there is the potential for a second process to recreate the file and also
4292 acquire a lock. This could lead to two Exim processes writing to the file
4293 at the same time. The added code performs the same test as UW imapd; it
4294 checks after acquiring the lock that its file descriptor still refers to
4295 the same named file.
4297 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4298 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4301 11. Added the smtp_active_hostname option. If used, this will typically be made
4302 to depend on the incoming interface address. Because $interface_address is
4303 not set up until the daemon has forked a reception process, error responses
4304 that can happen earlier (such as "too many connections") no longer contain
4307 12. If an expansion in a condition on a "warn" statement fails because a lookup
4308 defers, the "warn" statement is abandoned, and the next ACL statement is
4309 processed. Previously this caused the whole ACL to be aborted.
4311 13. Added the iplsearch lookup type.
4313 14. Added ident_timeout as a log selector.
4315 15. Added tls_certificate_verified as a log selector.
4317 16. Added a global option tls_require_ciphers (compare the smtp transport
4318 option of the same name). This controls incoming TLS connections.
4320 17. I finally figured out how to make tls_require_ciphers do a similar thing
4321 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4322 before starting the TLS session.
4324 18. Tabs are now shown as \t in -bP output.
4326 19. If the log selector return_path_on_delivery was set, Exim crashed when
4327 bouncing a message because it had too many Received: header lines.
4329 20. If two routers both had headers_remove settings, and the first one included
4330 a superfluous trailing colon, the final name in the first list and the
4331 first name in the second list were incorrectly joined into one item (with a
4332 colon in the middle).
4338 1. Added -C and -D options to the exinext utility, mainly to make it easier
4339 to include in the automated testing, but these could be helpful when
4340 multiple configurations are in use.
4342 2. The exinext utility was not formatting the output nicely when there was
4343 an alternate port involved in the retry record key, nor when there was a
4344 message id as well (for retries that were specific to a specific message
4345 and a specific host). It was also confused by IPv6 addresses, because of
4346 the additional colons they contain. I have fixed the IPv4 problem, and
4347 patched it up to do a reasonable job for IPv6.
4349 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4350 delivery, the log line now contains "pipelined" if PIPELINING was used.
4352 4. An SMTP transport process used to panic and die if the bind() call to set
4353 an explicit outgoing interface failed. This has been changed; it is now
4354 treated in the same way as a connect() failure.
4356 5. A reference to $sender_host_name in the part of a conditional expansion
4357 that was being skipped was still causing a DNS lookup. This no longer
4360 6. The def: expansion condition was not recognizing references to header lines
4361 that used bh_ and bheader_.
4363 7. Added the _cache feature to named lists.
4365 8. The code for checking quota_filecount in the appendfile transport was
4366 allowing one more file than it should have been.
4368 9. For compatibility with Sendmail, the command line option
4376 and sets the incoming protocol and host name (for trusted callers). The
4377 host name and its colon can be omitted when only the protocol is to be set.
4378 Note the Exim already has two private options, -pd and -ps, that refer to
4379 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4380 "s", but I don't think that's a major issue.
4382 10. A number of refactoring changes to the code, none of which should affect
4385 (a) The number of logging options was getting close to filling up the
4386 32-bit word that was used as a bit map. I have split them into two classes:
4387 those that are passed in the argument to log_write(), and those that are
4388 only ever tested independently outside of that function. These are now in
4389 separate 32-bit words, so there is plenty of room for expansion again.
4390 There is no change in the user interface or the logging behaviour.
4392 (b) When building, for example, log lines, the code previously used a
4393 macro that called string_cat() twice, in order to add two strings. This is
4394 not really sufficiently general. Furthermore, there was one instance where
4395 it was actually wrong because one of the argument was used twice, and in
4396 one call a function was used. (As it happened, calling the function twice
4397 did not affect the overall behaviour.) The macro has been replaced by a
4398 function that can join an arbitrary number of extra strings onto a growing
4401 (c) The code for expansion conditions now uses a table and a binary chop
4402 instead of a serial search (which was left over from when there were very
4403 few conditions). Also, it now recognizes conditions like "pam" even when
4404 the relevant support is not compiled in: a suitably worded error message is
4405 given if an attempt is made to use such a condition.
4407 11. Added ${time_interval:xxxxx}.
4409 12. A bug was causing one of the ddress fields not to be passed back correctly
4410 from remote delivery subprocesses. The field in question was not being
4411 subsequently used, so this caused to problems in practice.
4413 13. Added new log selectors queue_time and deliver_time.
4415 14. Might have fixed a bug in maildirsizefile handling that threw up
4416 "unexpected character" debug warnings, and recalculated the data
4417 unnecessarily. In any case, I expanded the warning message to give more
4420 15. Added the message "Restricted characters in address" to the statements in
4421 the default ACL that block characters like @ and % in local parts.
4423 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4424 Three changes have been made:
4426 (a) There was a serious bug; a negative response to MAIL caused the whole
4427 recipient domain to be cached as invalid, thereby blocking all messages
4428 to all local parts at the same domain, from all senders. This bug has
4429 been fixed. The domain is no longer cached after a negative response to
4430 MAIL if the sender used is not empty.
4432 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4435 (c) A new callout option, "use_sender" has been added for people who want
4436 the modified behaviour.
4442 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4445 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4446 indeed breaks things for older releases.
4448 3. Added additional logging to the case where there is a problem reading data
4449 from a filter that is running in a subprocess using a pipe, in order to
4450 try to track down a specific problem.
4452 4. Testing facility fudge: when running in the test harness and attempting
4453 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4454 getting "No route to host". Convert this to a timeout.
4456 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4459 6. Some OS don't have socklen_t but use size_t instead. This affects the
4460 fifth argument of getsockopt() amongst other things. This is now
4461 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4462 can be set for individual OS. I have set it for SunOS5, OSF1, and
4463 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4464 some earlier ones do not.
4466 7. Change 4.30/15 was not doing the test caselessly.
4468 8. The standard form for an IPv6 address literal was being rejected by address
4469 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4470 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4471 this, as well as the form without the "IPv6" on the front (but only when
4472 address literals are enabled, of course).
4474 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4476 10. Exim crashed if a message with an empty sender address specified by -f
4477 encountered a router with an errors_to setting. This could be provoked only
4478 by a command such as
4482 where an empty string was supplied; "<>" did not hit this bug.
4484 11. Installed PCRE release 4.5.
4486 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4487 remained set. It is now erased.
4489 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4490 times from message ids (which are base 36 rather than the normal 62).
4492 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4493 were being counted as actual protocol errors, and logged if the log
4494 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4495 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4496 and DATA following a set of rejected RCPTs do not count as protocol errors.
4497 In other words, Exim assumes they were pipelined, though this may not
4498 actually be the case. Of course, in all cases the client gets an
4499 appropriate error code.
4501 15. If a lookup fails in an ACL condition, a message about the failure may
4502 be available; it is used if testing the ACL cannot continue, because most
4503 such messages specify what the cause of the deferral is. However, some
4504 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4505 that caused an old message to be retained and used if a later statement
4506 caused a defer, replacing the real cause of the deferral.
4508 16. If an IP address had so many PTR records that the DNS lookup buffer
4509 was not large enough to hold them, Exim could crash while trying to process
4510 the truncated data. It now detects and logs this case.
4512 17. Further to 4.21/58, another change has been made: if (and only if) the
4513 first line of a message (the first header line) ends with CRLF, a bare LF
4514 in a subsequent header line has a space inserted after it, so as not to
4515 terminate the header.
4517 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4518 unnecessarily, used atoi() instead of strtol(), and didn't check the
4519 termination when getting file sizes from file names by regex.
4521 19. Completely re-implemented the support for maildirsize files, in the light
4522 of a number of problems with the previous contributed implementation
4523 (4.30/29). In particular:
4525 . If the quota is zero, the maildirsize file is maintained, but no quota is
4528 . If the maildir directory does not exist, it is created before any attempt
4529 to write a maildirsize file.
4531 . The quota value in the file is just a cache; if the quota is changed in
4532 the transport, the new value overrides.
4534 . A regular expression is available for excluding directories from the
4537 20. The autoreply transport checks the characters in options that define the
4538 message's headers; it allows continued headers, but it was checking with
4539 isspace() after an embedded newline instead of explicitly looking for a
4542 21. If all the "regular" hosts to which an address was routed had passed their
4543 expiry times, and had not reached their retry times, the address was
4544 bounced, even if fallback hosts were defined. Now Exim should go on to try
4547 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4548 equivalent code in the SMTP transport. Some hosts send humungous responses
4549 to HELO/EHLO, more than 1024 it seems.
4551 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4552 gives compiler warnings in some environments. I've now done it "properly",
4555 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4556 (because of problems with the built-in one) was declared to return uschar *
4557 instead of char *, causing compiler failure.
4559 25. Fixed a file descriptor leak when processing alias/forward files.
4561 26. Fixed a minor format string issue in dbfn.c.
4563 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4565 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4566 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4569 29. When a sender address is verified, it is cached, to save repeating the test
4570 when there is more than one recipient in a message. However, when the
4571 verification involves a callout, it is possible for different callout
4572 options to be set for different recipients. It is too complicated to keep
4573 track of this in the cache, so now Exim always runs a verification when a
4574 callout is required, relying on the callout cache for the optimization.
4575 The overhead is duplication of the address routing, but this should not be
4578 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4579 to be verified with callout=postmaster, and the main callout worked but the
4580 postmaster check failed, the verification correctly failed. However, if a
4581 subsequent RCPT command asked for sender verification *without* the
4582 postmaster check, incorrect caching caused this verification also to fail,
4585 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4586 it was not caching the DNS options (qualify_single, search_parents) that
4587 were used when the lookup failed. A subsequent lookup with different
4588 options therefore always gave the same answer, though there were cases
4589 where it should not have. (Example: a "domains = !$mx_any" option on a
4590 dnslookup router: the "domains" option is always processed without any
4591 widening, but the router might have qualify_single set.) Now Exim uses the
4592 cached value only when the same options are set.
4594 32. Added John Jetmore's "exipick" utility to the distribution.
4596 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4597 than a timeout (e.g. a certificate is required, and is not provided), an
4598 Exim server now closes the connection immediately. Previously it waited for
4599 the client to close - but if the client is SSL, it seems that they each
4600 wait for each other, leading to a delay before one of them times out.
4602 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4603 maintained 0.8.x compatibility because I don't think many are using it, and
4604 it is clearly obsolete.
4606 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4609 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4610 client certificate was expired. A simple patch fixes this, though I don't
4611 understand the full logic of why the verify callback is called multiple
4614 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4615 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4616 which causes problems with some clients (such as the Certicom SSL Plus
4617 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4618 disables the coutermeasure allowing Eudora to connect."
4620 38. Exim was not checking that a write() to a log file succeeded. This could
4621 lead to Bad Things if a log got too big, in particular if it hit a file
4622 size limit. Exim now panics and dies if it cannot write to a log file, just
4623 as it does if it cannot open a log file.
4625 39. Modified OS/Makefile-Linux so that it now contains
4627 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4629 The two -D definitions ensure that Exim is compiled with large file
4630 support, which makes it possible to handle log files that are bigger than
4633 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4634 instance) a domain was checked against a named list that involved a lookup,
4635 causing $domain_data to be set, then another domain was checked against the
4636 same list, then the first domain was re-checked, the value of $domain_data
4637 after the final check could be wrong. In particular, if the second check
4638 failed, it could be set empty. This bug probably also applied to
4641 41. The strip_trailing_dot option was not being applied to the address given
4642 with the -f command-line option.
4644 42. The code for reading a message's header from the spool was incrementing
4645 $received_count, but never initializing it. This meant that the value was
4646 incorrect (doubled) while delivering a message in the same process in which
4647 it was received. In the most common configuration of Exim, this never
4648 happens - a fresh exec is done - but it can happen when
4649 deliver_drop_privilege is set.
4651 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4652 it now includes up to 150 characters of the unexpected data in the log
4655 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4656 and building data strings. The size of both of these buffers was 10 000
4657 bytes - far larger than anybody would *ever* want, thought I. Needless to
4658 say, somebody hit the limit. I have increased the maximum line length to
4659 20 000 and the maximum data length of concatenated lines to 100 000. I have
4660 also fixed two bugs, because there was no checking on these buffers. Tsk,
4661 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4662 buffer is too small.
4664 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4665 lsearch lookups. Now it does.
4667 46. When parsing a route_list item in a manualroute router, a fixed-length
4668 buffer was used for the list of hosts. I made this 1024 bytes long,
4669 thinking that nobody would ever have a list of hosts that long. Wrong.
4670 Somebody had a whole pile of complicated expansion conditions, and the
4671 string was silently truncated, leading to an expansion error. It turns out
4672 that it is easier to change to an unlimited length (owing to other changes
4673 that have happened since this code was originally written) than to build
4674 structure for giving a limitation error. The length of the item that
4675 expands into the list of hosts is now unlimited.
4677 47. The lsearch lookup could not handle data where the length of text line was
4678 more than 4095 characters. Such lines were truncated, leading to shortened
4679 data being returned. It should now handle lines of any length.
4681 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4682 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4685 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4686 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4687 no longer happens in comment lines. A list of replacements is now placed
4688 at the head of all of the source files, except those whose only change is
4689 to replace PERL_COMMAND in the very first #! line.
4691 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4692 messages on the queue, with a bottom-up merge sort, using code contributed
4693 by Michael Haardt. This should make operations like -bp somewhat faster on
4694 large queues. It won't affect queue runners, except when queue_run_in_order
4697 51. Installed eximstats 1.31 in the distribution.
4699 52. Added support for SRV lookups to the dnslookup router.
4701 53. If an ACL referred to $message_body or $message_body_end, the value was not
4702 reset for any messages that followed in the same SMTP session.
4704 54. The store-handling optimization for building very long strings was not
4705 differentiating between the different store pools. I don't think this
4706 actually made any difference in practice, but I've tidied it.
4708 55. While running the routers to verify a sender address, $sender_address
4709 was still set to the sender address. This is wrong, because when routing to
4710 send a bounce to the sender, it would be empty. Therefore, I have changed
4711 it so that, while verifying a sender address, $sender_address is set to <>.
4712 (There is no change to what happens when verifying a recipient address.)
4714 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4715 target A or AAAA records (if not already returned) without resetting the
4716 qualify_single or search_parents options of the DNS resolver. These are
4717 inappropriate in this case because the targets of MX and SRV records must
4718 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4719 target that, when qualified, matched something in the local domain. These
4720 two options are now turned off when doing these lookups.
4722 57. It seems that at least some releases of Reiserfs (which does not have the
4723 concept of a fixed number of inodes) returns zero and not -1 for the
4724 number of available inodes. This interacted badly with check_spool_inodes,
4725 which assumed that -1 was the "no such thing" setting. What I have done is
4726 to check that the total number of inodes is greater than zero before doing
4727 the test of how many are available.
4729 58. When a "warn" ACL statement has a log_message modifier, the message is
4730 remembered, and not repeated. This is to avoid a lot of repetition when a
4731 message has many recipients that cause the same warning to be written.
4732 Howewer, Exim was preserving the list of already written lines for an
4733 entire SMTP session, which doesn't seem right. The memory is now reset if a
4734 new message is started.
4736 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4737 debugging output unless log_rewrite was also set.
4739 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4740 of (int)(handle) when we know that handle contains (void *)(-1).
4742 61. The Exim daemon panic-logs an error return when it closes the incoming
4743 connection. However "connection reset by peer" seems to be common, and
4744 isn't really an error worthy of noting specially, so that particular error
4747 62. When Exim is trying to find all the local interfaces, it used to panic and
4748 die if the ioctl to get the interface flags failed. However, it seems that
4749 on at least one OS (Solaris 9) it is possible to have an interface that is
4750 included in the list of interfaces, but for which you get a failure error
4751 for this call. This happens when the interface is not "plumbed" into a
4752 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4753 failure of the "get flags" call assumes that the interface is down.
4755 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4756 makes life easier for people who are doing arithmetic on fields extracted
4757 from dates, where you often get leading zeros that should not be
4758 interpreted as octal.
4760 64. Added qualify_domain to the redirect router, to override the global
4763 65. If a pathologically long header line contained very many addresses (the
4764 report of this problem mentioned 10 000) and each of them was rewritten,
4765 Exim could use up a very large amount of memory. (It kept on making new
4766 copies of the header line as it rewrote, and never released the old ones.)
4767 At the expense of a bit more processing, the header rewriting function has
4768 been changed so that it no longer eats memory in this way.
4770 66. The generation of the Received: header has been moved from the time that a
4771 message starts to be received, to the time that it finishes. The timestamp
4772 in the Received: header should now be very close to that of the <= log
4773 line. There are two side-effects of this change:
4775 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4776 logged header lines no longer include the local Received: line, because
4777 it has not yet been created. The same applies to a copy of the message
4778 that is returned to a non-SMTP sender when a message is rejected.
4780 (b) When a filter file is tested using -bf, no additional Received: header
4781 is added to the test message. After some thought, I decided that this
4784 This change does not affect the value of $received_for. It is still set
4785 after address rewriting, but before local_scan() is called.
4787 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4789 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4790 gave an unhelpful panic error message, and a defer error. I have managed to
4791 change this behaviour so that it now rejects any supplied certificate,
4792 which seems right, as the list of acceptable certificates is empty.
4794 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4795 gave an unhelpful defer error. I have not managed to make this reject any
4796 supplied certificates, but the error message it gives is "no certificate
4797 supplied", which is not helpful.
4799 70. exigrep's output now also includes lines that are not associated with any
4800 message, but which match the given pattern. Implemented by a patch from
4801 Martin Sluka, which also tidied up the Perl a bit.
4803 71. Recipient callout verification, like sender verification, was using <> in
4804 the MAIL FROM command. This isn't really the right thing, since the actual
4805 sender may affect whether the remote host accepts the recipient or not. I
4806 have changed it to use the actual sender in the callout; this means that
4807 the cache record is now keyed on a recipient/sender pair, not just the
4808 recipient address. There doesn't seem to be a real danger of callout loops,
4809 since a callout by the remote host to check the sender would use <>.
4810 [SEE ABOVE: changed after hitting problems.]
4812 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4813 temporary errors. However, in the case of such a code being given after
4814 the end of a data transmission (i.e. after ".") Exim was failing to write
4815 a retry record for the message. (Yes, there was some broken host that was
4816 actually sending 8xx at this point.)
4818 73. An unknown lookup type in a host list could cause Exim to panic-die when
4819 the list was checked. (An example that provoked this was putting <; in the
4820 middle of a list instead of at the start.) If this happened during a DATA
4821 ACL check, a -D file could be left lying around. This kind of configuration
4822 error no longer causes Exim to die; instead it causes a defer errror. The
4823 incident is still logged to the main and panic logs.
4825 74. Buglet left over from Exim 3 conversion. The message "too many messages
4826 in one connection" was written to the rejectlog but not the mainlog, except
4827 when address rewriting (yes!) was being logged.
4829 75. Added write_rejectlog option.
4831 76. When a system filter was run not as root (that is, when system_filter_user
4832 was set), the values of the $n variables were not being returned to the
4833 main process; thus, they were not subsequently available in the $sn
4836 77. Added +return_path_on_delivery log selector.
4838 78. A connection timeout was being treated differently from recipients deferred
4839 when testing hosts_max_try with a message that was older than the host's
4840 retry timeout. (The host should not be counted, thus allowing all hosts to
4841 be tried at least once before bouncing.) This may have been the cause of an
4842 occasionally reported bug whereby a message would remain on the queue
4843 longer than the retry timeout, but would be bounced if a delivery was
4844 forced. I say "may" because I never totally pinned down the problem;
4845 setting up timeout/retry tests is difficult. See also the next item.
4847 79. The ultimate address timeout was not being applied to errors that involved
4848 a combination of host plus message (for example, a timeout on a MAIL
4849 command). When an address resolved to a number of possible hosts, and they
4850 were not all tried for each delivery (e.g. because of hosts_max_try), a
4851 message could remain on the queue longer than the retry timeout.
4853 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4856 81. Fixed an obscure SMTP outgoing bug which required at least the following
4857 conditions: (a) there was another message waiting for the same server;
4858 (b) the server returned 5xx to all RCPT commands in the first message so
4859 that the message was not completed; (c) the server dropped the connection
4860 or gave a negative response to the RSET that Exim sends to abort the
4861 transaction. The observed case was a dropped connection after DATA that had
4862 been sent in pipelining mode. That is, the server had advertised PIPELINING
4863 but was not implementing it correctly. The effect of the bug was incorrect
4864 behaviour, such as trying another host, and this could lead to a crash.
4870 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4871 and daemon.c were passed as pointers to ints; they should have been
4872 pointers to socklen_t variables (which are typically unsigned ints).
4874 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4877 3. Fixed a really odd bug that affected only the testing scheme; patching a
4878 certain fixed string in the binary changed the value of another string that
4879 happened to be identical to the end of the original first string.
4881 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4882 name", it returns that address as the IP address. On some operating
4883 systems (e.g. Solaris), it also passes back the IP address string as the
4884 "host name". However, on others (e.g. Linux), it passes back an empty
4885 string. Exim wasn't checking for this, and was changing the host name to an
4886 empty string, assuming it had been canonicized.
4888 5. Although rare, it is permitted to have more than one PTR record for a given
4889 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4890 all the names associated with an address, because they do in Solaris.
4891 However, it seems that they do not in Linux for data that comes from the
4892 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4893 I found this out when I moved to a new Linux workstation and tried to run
4894 the Exim test suite.
4896 To get round this problem I have changed the code so that it now does its
4897 own call to the DNS to look up PTR records when searching for a host name.
4898 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4899 addresses that are only in /etc/hosts are still found.
4901 This behaviour is, however, controlled by an option called host_lookup_
4902 order, which defaults to "bydns:byaddr". If people want to use the other
4903 order, or indeed, just use one or the other means of lookup, they can
4904 specify it in this variable.
4906 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4907 some operating systems, this comes back from gethostbyaddr() as an empty
4908 string, and this is what Exim used to test for. However, it seems that in
4909 other systems, "." is yielded. Exim now tests for this case too.
4911 7. The values of check_spool_space and check_log_space are now held internally
4912 as a number of kilobytes instead of an absolute number of bytes. If a
4913 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4914 kilobyte. This means that much larger values can be stored.
4916 8. Exim monitor: an attempt to get the action menu when not actually pointing
4917 at a message produces an empty menu entitled "No message selected". This
4918 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4919 no entries in it ("Shell widget menu has zero width and/or height"). So I
4920 have added a single, blank menu entry in this case.
4922 9. Added ${quote_local_part.
4924 10. MIME decoding is now applied to the contents of Subject: header lines when
4927 11. Now that a reference to $sender_host_address automatically causes a reverse
4928 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4929 host lookup before query-style lookups in lists that might use this
4930 variable. This has therefore been abolished, and the "net-" prefix is no
4931 longer necessary for query-style lookups.
4933 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4934 have been a typo for LFLAGS, so it has been changed.
4936 13. The install script calls Exim with "-C /dev/null" in order to find the
4937 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4938 to be output. Howeve, since Exim outputs its version number before the
4939 error, it didn't break the script. It just looked ugly. I fixed this by
4940 always allowing "-C /dev/null" if the caller is root.
4942 14. Ignore overlarge ACL variable number when reading spool file - insurance
4943 against a later release with more variables having written the file.
4945 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4946 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4947 this, as well as the form without the "IPv6" on the front.
4949 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4950 OS/Makefile-Darwin file.
4952 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4953 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4955 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4956 variable. However, some OS use ints for inodes. I've added cast to long int
4957 to get rid of the compiler warning.
4959 19. I had forgotten to lock out "/../" in configuration file names when
4960 ALT_CONFIG_PREFIX was set.
4962 20. Routers used for verification do not need to specify transports. However,
4963 if such a router generated a host list, and callout was configured, Exim
4964 crashed, because it could not find a port number from the (non-existent)
4965 transport. It now assumes port 25 in this circumstance.
4967 21. Added the -t option to exigrep.
4969 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4970 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4971 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4972 them is set, LOOKUP_LSEARCH is forced.
4974 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4976 24. Added sender and host information to the "rejected by local_scan()" log
4977 line; previously there was no indication of these.
4979 25. Added .include_if_exists.
4981 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4982 Linux. It turns out that not all file systems support this. Apparently some
4983 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4984 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4985 not supported on the file descriptor, is now ignored when Exim is trying to
4986 sync a directory. This applies only to Linux.
4988 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4990 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4991 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4994 29. Added support for maildirsize files from supplied patch (modified a bit).
4996 30. The use of :fail: followed by an empty string could lead Exim to respond to
4997 sender verification failures with (e.g.):
4999 550 Verification failed for <xxx>
5000 550 Sender verify failed
5002 where the first response line was missing the '-' that indicates it is not
5003 the final line of the response.
5005 31. The loop for finding the name of the user that called Exim had a hardwired
5006 limit of 10; it now uses the value of finduser_retries, which is used for
5007 all other user lookups.
5009 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5012 33. Exim was neglecting to zero errno before one call of strtol() when
5013 expanding a string and expecting an integer value. On some systems this
5014 resulted in spurious "integer overflow" errors. Also, it was casting the
5015 result into an int without checking.
5017 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5018 did not work. The code looks as if it has *never* worked, though it appears
5019 to have been documented since at least releast 1.62. I have made it work.
5021 35. The "timeout_DNS" error in retry rules, also documented since at least
5022 1.62, also never worked. As it isn't clear exactly what this means, and
5023 clearly it isn't a major issue, I have abolished the feature by treating it
5024 as "timeout", and writing a warning to the main and panic logs.
5026 36. The display of retry rules for -brt wasn't always showing the error code
5029 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5030 timeout_connect_A, timeout_connect_MX.
5032 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5033 to the empty sender.
5035 39. The daemon was not analysing the content of -oX till after it had closed
5036 stderr and disconnected from the controlling terminal. This meant that any
5037 syntax errors were only noted on the panic log, and the return code from
5038 the command was 0. By re-arranging the code a little, I've made the
5039 decoding happen first, so such errors now appear on stderr, and the return
5040 code is 1. However, the actual setting up of the sockets still happens in
5041 the disconnected process, so errors there are still only recorded on the
5044 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5045 connections (as happens on some IP stacks) was logged at start up time as
5046 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5047 it from "IPv6 and IPv4", which means that two separate sockets are being
5050 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5051 says whether AF_INET or AF_INET6 was passed as an argument.
5053 42. Exiwhat output was messed up when time zones were included in log
5056 43. Exiwhat now gives more information about the daemon's listening ports,
5057 and whether -tls-on-connect was used.
5059 44. The "port" option of the smtp transport is now expanded.
5061 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5062 silently ignored. Now an error message is written to the main and panic
5065 46. There's a new ACL modifier called "logwrite" which writes to a log file
5066 as soon as it is encountered.
5068 47. Added $local_user_uid and $local_user_gid at routing time.
5070 48. Exim crashed when trying to verify a sender address that was being
5073 49. Exim was recognizing only a space character after ".include". It now also
5074 recognizes a tab character.
5076 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5077 extracting the relevant information from the specification. The man page no
5078 longer contains scrambled data for the -d option, and I've added a section
5079 at the front about calling Exim under different names.
5081 51. Added "extra_headers" argument to the "mail" command in filter files.
5083 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5086 53. Installed eximstats 1.29.
5088 54. Added transport_filter_timeout as a generic transport option.
5090 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5091 Cc: header lines. This was required by RFC 822, but it not required by RFC
5094 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5095 incoming messages that did not have them. Now it does so only if the
5096 message originates locally, that is, if there is no associated remote host
5097 address. When Resent- header lines are present, this applies to the Resent-
5098 lines rather than the non-Resent- lines.
5100 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5101 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5103 58. Messages for configuration errors now include the name of the main
5104 configuration file - useful now that there may be more than one file in a
5105 list (.included file names were always shown).
5107 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5108 for those rare installations that do not start the daemon as root or run it
5109 setuid root. I've cut out the call to initgroups() if the daemon is not
5112 60. The Exim user and group can now be bound into the binary as text strings
5113 that are looked up at the start of Exim's processing.
5115 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5117 62. Added $mailstore_basename variable.
5119 63. Installed patch to sieve.c from Michael Haardt.
5121 64. When Exim failed to open the panic log after failing to open the main log,
5122 the original message it was trying to log was written to stderr and debug
5123 output, but if they were not available (the usual case in production), it
5124 was lost. Now it is written to syslog before the two lines that record the
5125 failures to open the logs.
5127 65. Users' Exim filters run in subprocesses under the user's uid. It is
5128 possible for a "deliver" command or an alias in a "personal" command to
5129 provoke an address rewrite. If logging of address rewriting is configured,
5130 this fails because the process is not running as root or exim. There may be
5131 a better way of dealing with this, but for the moment (because 4.30 needs
5132 to be released), I have disabled address rewrite logging when running a
5133 filter in a non-root, non-exim process.
5139 1. The buildconfig auxiliary program wasn't quoting the value set for
5140 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5141 not defined. This bug was masked in 4.22 by the effect that was fixed in
5144 2. Some messages that were rejected after a message id was allocated were
5145 shown as "incomplete" by exigrep. It no longer does this for messages that
5146 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5148 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5149 have allow_domain_literals set, the ID did not get logged in the <= line.
5150 Domain literals are now always recognized in Message-ID: header lines.
5152 4. The first argument for a ${extract expansion item is the key name or field
5153 number. Leading and trailing spaces in this item were not being ignored,
5154 causing some misleading effects.
5156 5. When deliver_drop_privilege was set, single queue runner processes started
5157 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5158 same command in the process it spins off) were not dropping privilege.
5160 6. When the daemon running as "exim" started a queue runner, it always
5161 re-executed Exim in the spun-off process. This is a waste of effort when
5162 deliver_drop_privilege is set. The new process now just calls the
5163 queue-runner function directly.
5169 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5172 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5173 ignored. Though the use of -oP was forcing the writing of a pid file, it
5174 was always written to the default place.
5176 3. If the message "no IP address found for host xxxx" is generated during
5177 incoming verification, it is now followed by identification of the incoming
5178 connection (so you can more easily find what provoked it).
5180 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5182 5. Added some features to "harden" Exim a bit more against certain attacks:
5184 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5185 be put in Local/Makefile. This is like the never_users runtime option,
5186 but it cannot be overridden. The default setting is "root".
5188 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5189 prefix string with which any file named in a -C command line option
5192 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5193 is retained for -C and -D only if the caller of Exim is root. Without
5194 it, the exim user may also use -C and -D and retain privilege.
5196 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5197 command line option is disabled.
5199 6. Macro names set by the -D option must start with an upper case letter, just
5200 like macro names defined in the configuration file.
5202 7. Added "dereference=" facility to LDAP.
5204 8. Two instances of the typo "uknown" in the source files are fixed.
5206 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5207 the Configure-Makefile script screwed up while processing it.
5209 10. Incorporated PCRE 4.4.
5211 11. The SMTP synchronization check was not operating right at the start of an
5212 SMTP session. For example, it could not catch a HELO sent before the client
5213 waited for the greeting. There is now a check for outstanding input at the
5214 point when the greeting is written. Because of the duplex, asynchronous
5215 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5216 way, but not yet received, when the check is performed.
5218 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5219 on TCP/IP sockets, because this apparently causes some broken clients to
5222 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5223 unchanged) from the Cygwin maintainer.
5225 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5226 was supported instead of testing for mbx. Effectively a typo.
5228 15. The spa authenticator server code was not checking that the input it
5229 received was valid base64.
5231 16. The debug output line for the "set" modifier in ACLs was not showing the
5232 name of the variable that was being set.
5234 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5236 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5237 lookup on its own. Something else was needed to trigger the lookup. For
5238 example, a match in host_lookup or the need for a host name in a host list.
5239 Now, if $sender_host_name is referenced and the host name has not yet been
5240 looked up, a lookup is performed. If the lookup fails, the variable remains
5241 empty, and $host_lookup_failed is set to "1".
5243 19. Added "eqi" as a case-independent comparison operator.
5245 20. The saslauthd authentication condition could segfault if neither service
5246 nor realm was specified.
5248 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5249 error message that was logged was misleading, and incoming SMTP
5250 connections were dropped. The message is now more accurate, and temporary
5251 errors are given to SMTP connections.
5253 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5254 (except RSET) with a 421 error, until QUIT is received. However, it was
5255 failing to send a response to QUIT.
5257 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5258 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5259 ACL is now run after the helo_try_verify_hosts code.
5261 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5262 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5263 case-independent, so other case variants are also recognized). Apparently
5264 some systems use these upper case variants.
5266 25. If more than two messages were waiting for the same host, and a transport
5267 filter was specified for the transport, Exim sent two messages over the
5268 same TCP/IP connection, and then failed with "socket operation on non-
5269 socket" when it tried to send the third.
5271 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5273 27. The extern definition of crypt16() in expand.c was not being excluded when
5274 the OS had its own crypt16() function.
5276 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5277 as a preferred synonym for return_size_limit, both as an option and as an
5280 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5282 30. Changed the default configuration ACL to relax the local part checking rule
5283 for addresses that are not in any local domains. For these addresses,
5284 slashes and pipe symbols are allowed within local parts, but the sequence
5285 /../ is explicitly forbidden.
5287 31. SPA server authentication was not clearing the challenge buffer before
5290 32. log_message in a "warn" ACL statement was writing to the reject log as
5291 well as to the main log, which contradicts the documentation and doesn't
5292 seem right (because no rejection is happening). So I have stopped it.
5294 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5295 However, I am unable to do any testing of this.
5297 34. Fixed an infelicity in the appendfile transport. When checking directories
5298 for a mailbox, to see if any needed to be created, it was accidentally
5299 using path names with one or more superfluous leading slashes; tracing
5300 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5302 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5303 contents are added to the log line that is written for every discarded
5304 recipient. (Previously a log_message setting was ignored.)
5306 36. The ${quote: operator now quotes the string if it is empty.
5308 37. The install script runs exim in order to find its version number. If for
5309 some reason other than non-existence or emptiness, which it checks, it
5310 could not run './exim', it was installing it with an empty version number,
5311 i.e. as "exim-". This error state is now caught, and the installation is
5314 38. An argument was missing from the function that creates an error message
5315 when Exim fails to connect to the socket for saslauthd authentication.
5316 This could cause Exim to crash, or give a corrupted message.
5318 39. Added isip, isip4, and isip6 to ${if conditions.
5320 40. The ACL variables $acl_xx are now saved with the message, and can be
5321 accessed later in routers, transports, and filters.
5323 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5324 strings in the file are not string-expanded.
5326 42. If a MAIL command specified a SIZE value that was too large to fit into an
5327 int variable, the check against message_size_limit failed. Such values are
5328 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5329 day this will have to be increased, but I don't think I want to be around
5330 when emails are that large.
5337 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5338 iconv() is not standard in FreeBSD.
5340 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5341 IPv6 enabled. The observed symptom was a segmentation fault on return from
5342 the function os_common_find_running_interfaces() in src/os.c.
5344 3. In the check_special_case() function in daemon.c I had used "errno" as an
5345 argument name, which causes warnings on some systems. This was basically a
5346 typo, since it was named "eno" in the comments!
5348 4. The code that waits for the clock to tick (at a resolution of some fraction
5349 of a second) so as to ensure message-id uniqueness was always waiting for
5350 at least one whole tick, when it could have waited for less. [This is
5351 almost certainly not relevant at current processor speeds, where it is
5352 unlikely to ever wait at all. But we try to future-proof.]
5354 5. The function that sleeps for a time interval that includes fractions of a
5355 second contained a race. It did not block SIGALRM between setting the
5356 timer, and suspending (a couple of lines later). If the interval was short
5357 and the sigsuspend() was delayed until after it had expired, the suspension
5358 never ended. On busy systems this could lead to processes getting stuck for
5361 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5362 process, before it forks any delivery processes. The open lookup caching
5363 mechanism meant that the open file or database connection was passed into
5364 the delivery process. The problem was that delivery processes always tidy
5365 up cached lookup data. This could cause a problem for the next delivery
5366 process started by the queue runner, because the external queue runner
5367 process does not know about the closure. So the next delivery process
5368 still has data in the lookup cache. In the case of a file lookup, there was
5369 no problem because closing a file descriptor in a subprocess doesn't affect
5370 the parent. However, if the lookup was caching a connection to a database,
5371 the connection was closed, and the second delivery process was likely to
5372 see errors such as "PGSQL: query failed: server closed the connection
5373 unexpectedly". The problem has been fixed by closing all cached lookups
5374 in a queue runner before running a delivery process.
5376 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5377 seem to have the "const" qualifier which it has on other OS. I've
5380 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5381 *of the same type* (client or server) with the same public name that an
5382 error should be diagnosed.
5384 9. When Exim looked up a host name for an IP address, but failed to find the
5385 original IP address when looking up the host name (a safety check), it
5386 output the message "<ip address> does not match any IP for NULL", which was
5387 confusing, to say the least. The bug was that the host name should have
5388 appeared instead of "NULL".
5390 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5391 user that is built into the binary, and the -C or -D options is used, root
5392 privilege is dropped before the configuration file is read. In addition,
5393 logging is switched to stderr instead of the normal log files. If the
5394 configuration then re-defines the Exim user, the unprivileged environment
5395 is probably not what is expected, so Exim logs a panic warning message (but
5398 However, if deliver_drop_privilege is set, the unprivileged state may well
5399 be exactly what is intended, so the warning has been cut out in that case,
5400 and Exim is allowed to try to write to its normal log files.
5406 1. smtp_return_error_details was not giving details for temporary sender
5407 or receiver verification errors.
5409 2. Diagnose a configuration error if two authenticators have the same public
5412 3. Exim used not to create the message log file for a message until the first
5413 delivery attempt. This could be confusing when incoming messages were held
5414 for policy or load reasons. The message log file is now created at the time
5415 the message is received, and an initial "Received" line is written to it.
5417 4. The automatically generated man page for command line options had a minor
5418 bug that caused no ill effects; however, a more serious problem was that
5419 the procedure for building the man page automatically didn't always
5420 operate. Consequently, release 4.20 contains an out-of-date version. This
5421 shouldn't happen again.
5423 5. When building Exim with embedded Perl support, the script that builds the
5424 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5425 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5427 6. The freeze_tell option was not being used for messages that were frozen on
5428 arrival, either by an ACL or by local_scan().
5430 7. Added the smtp_incomplete_transaction log selector.
5432 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5433 was accepting AUTH without a new EHLO.
5435 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5436 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5439 10. Logging of TCP/IP connections (when configured) now happens in the main
5440 daemon process instead of the child process, so that the TCP/IP connection
5441 count is more accurate (but it can never be perfect).
5443 11. The use of "drop" in a nested ACL was not being handled correctly in the
5444 outer ACL. Now, if condition failure induced by the nested "drop" causes
5445 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5446 or "require"), the connection is dropped.
5448 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5449 that yield "discard" can now be used with an "accept" or a "discard" verb,
5450 but an error is generated for any others (because I can't see a useful way
5451 to define what should happen).
5453 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5454 are now processed in the same way as lines in the Exim configuration file.
5455 In particular, continuation lines are supported.
5457 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5459 15. Added -ti meaning -t -i.
5461 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5462 domains, and warn by logging if others are found.
5464 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5465 returned from GIFCONF when Exim is trying to find the list of interfaces on
5466 a host. The code in os.c has been modified to copy each ifreq to an aligned
5467 structure in all cases.
5469 Also, in some cases, the returned ifreq's were being copied to a 'struct
5470 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5471 means the last couple of bytes of an IPv6 address could be chopped if the
5472 ifreq contained only a normal sockaddr (14 bytes storage).
5474 18. Named domain lists were not supported in the hosts_treat_as_local option.
5475 An entry such as +xxxx was not recognized, and was treated as a literal
5478 19. Ensure that header lines added by a DATA ACL are included in the reject log
5479 if the ACL subsequently rejects the message.
5481 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5482 MD5 (which is deprecated).
5484 21. When testing a filter file using -bf, Exim was writing a message when it
5485 took the sender from a "From " line in the message, but it was not doing so
5486 when it took $return_path from a Return-Path: header line. It now does.
5488 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5489 with a valid header line field name (a series of printing characters
5490 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5492 23. Changed "disc" in the source to "disk" to conform to the documentation and
5493 the book and for uniformity.
5495 24. Ignore Sendmail's -Ooption=value command line item.
5497 25. When execve() failed while trying to run a command in a pipe transport,
5498 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5499 could be confused with a return value of 69 from the command itself. This
5500 has been changed to 127, the value the shell returns if it is asked to run
5501 a non-existent command. The wording for the related log line suggests a
5502 non-existent command as the problem.
5504 26. If received_header_text expands to an empty string, do not add a Received:
5505 header line to the message. (Well, it adds a token one on the spool, but
5506 marks it "old" so that it doesn't get used or transmitted.)
5508 27. Installed eximstats 1.28 (addition of -nt option).
5510 28. There was no check for failure on the call to getsockname() in the daemon
5511 code. This can fail if there is a shortage of resources on the system, with
5512 ENOMEM, for example. A temporary error is now given on failure.
5514 29. Contrary to the C standard, it seems that in some environments, the
5515 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5516 program. Exim now does this explicitly; it affects the formatting of
5517 timestamps using strftime().
5519 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5520 complaints. I've now initialized all the variables, to avoid this.
5522 32. Header lines added by a system filter were not being "seen" during
5523 transport-time rewrites.
5525 33. The info_callback() function passed to OpenSSL is set up with type void
5526 (*)(SSL *, int, int), as described somewhere. However, when calling the
5527 function (actually a macro) that sets it up, the type void(*)() is
5528 expected. I've put in a cast to prevent warnings from picky compilers.
5530 34. If a DNS black list lookup found a CNAME record, but there were no A
5531 records associated with the domain it pointed at, Exim crashed.
5533 35. If a DNS black list lookup returned more than one A record, Exim ignored
5534 all but the first. It now scans all returned addresses if a particular IP
5535 value is being sought. In this situation, the contents of the
5536 $dnslist_value variable are a list of all the addresses, separated by a
5539 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5540 to accept a host name and all its aliases if the forward lookup for any of
5541 them yielded the IP address of the incoming connection. Now it accepts only
5542 those names whose forward lookup yields the correct IP address. Any other
5543 names are discarded. This closes a loophole whereby a rogue DNS
5544 administrator could create reverse DNS records to break through a
5545 wildcarded host restriction in an ACL.
5547 37. If a user filter or a system filter that ran in a subprocess used any of
5548 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5549 the wrong values were passed to the pipe command ($thisaddress had the
5550 value of $0, $0 had the value of $1, etc). This bug was introduced by
5551 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5553 38. Improved the line breaking for long SMTP error messages from ACLs.
5554 Previously, if there was no break point between 40 and 75 characters, Exim
5555 left the rest of the message alone. Two changes have been made: (a) I've
5556 reduced the minimum length to 35 characters; (b) if it can't find a break
5557 point between 35 and 75 characters, it looks ahead and uses the first one
5558 that it finds. This may give the occasional overlong line, but at least the
5559 remaining text gets split now.
5561 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5562 file descriptors might be low, and that setting 1000 would always raise it.
5563 It turns out that in some environments, the limit is already over 1000 and
5564 that lowering it causes trouble. So now Exim takes care not to decrease it.
5566 40. When delivering a message, the value of $return_path is set to $sender_
5567 address at the start of routing (routers may change the value). By an
5568 oversight, this default was not being set up when an address was tested by
5569 -bt or -bv, which affected the outcome if any router or filter referred to
5572 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5573 log only when "message" or "log_message" are set. However, if one of the
5574 conditions was an address verification, or a call to a nested ACL, the
5575 messages generated by the underlying test were being passed through. This
5576 no longer happens. The underlying message is available in $acl_verify_
5577 message for both "message" and "log_message" expansions, so it can be
5578 passed through if needed.
5580 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5581 new expansion $bh_ to give the encoded byte string without charset
5582 translation. Translation happens only if iconv() is available; HAVE_ICONV
5583 indicates this at build time. HEADERS_CHARSET gives the charset to
5584 translate to; headers_charset can change it in the configuration, and
5585 "headers charset" can change it in an individual filter file.
5587 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5588 that creates RFC 2047 encoded "words" labels them as that charset instead
5589 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5590 expansion operator; (ii) when Exim creates a From: line for a local
5591 message; (iii) when a header line is rewritten to include a "phrase" part.
5593 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5594 buggy, causing it to skip the first lines of messages whose message ID
5595 ended in 'D'. This would not have bitten before Exim release 4.14, because
5596 message IDs were unlikely to end in 'D' before then. The effect was to have
5597 incorrect size information for certain domains.
5599 45. #include "config.h" was missing at the start of the crypt16.c module. This
5600 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5603 46. If there was a timeout during a "random" callout check, Exim treated it as
5604 a failure of the random address, and carried on sending RSET and the real
5605 address. If the delay was just some slowness somewhere, the response to the
5606 original RCPT would be taken as a response to RSET and so on, causing
5607 mayhem of various kinds.
5609 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5610 when I implemented it. It didn't allow for the fact that some option values
5611 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5612 the right test for positive values.
5614 48. Domain names in DNS records are case-independent. Exim always looks them up
5615 in lower case. Some resolvers return domain names in exactly the case they
5616 appear in the zone file, that is, they may contain uppercase letters. Not
5617 all resolvers do this - some return always lower case. Exim was treating a
5618 change of case by a resolver as a change of domain, similar to a widening
5619 of a domain abbreviation. This triggered its re-routing code and so it was
5620 trying to route what was effectively the same domain again. This normally
5621 caused routing to fail (because the router wouldn't handle the domain
5622 twice). Now Exim checks for this case specially, and just changes the
5623 casing of the domain that it ultimately uses when it transmits the message
5626 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5629 50. If a filter generated a file delivery with a non-absolute name (possible if
5630 no home directory exists for the router), the forbid_file option was not
5633 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5634 the existing equality matching.
5636 52. Exim was using ints instead of ino_t variables in some places where it was
5637 dealing with inode numbers.
5639 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5640 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5641 called TMPDIR, and if it finds it is different, it changes its value.
5643 54. The smtp_printf() function is now made available to local_scan() so
5644 additional output lines can be written before returning. There is also an
5645 smtp_fflush() function to enable the detection of a dropped connection.
5646 The variables smtp_input and smtp_batched_input are exported to
5649 55. Changed the default runtime configuration: the message "Unknown user"
5650 has been removed from the ACL, and instead placed on the localuser router,
5651 using the cannot_route_message feature. This means that any verification
5652 failures that generate their own messages won't get overridden. Similarly,
5653 the "Unrouteable address" message that was in the ACL for unverifiable
5654 relay addresses has also been removed.
5656 56. Added hosts_avoid_esmtp to the smtp transport.
5658 57. The exicyclog script was not checking for the esoteric option
5659 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5660 will work only if exicyclog is run under the appropriate euid.
5662 58. Following a discussion on the list, the rules by which Exim recognises line
5663 endings on incoming messages have been changed. The -dropcr and drop_cr
5664 options are now no-ops, retained only for backwards compatibility. The
5665 following line terminators are recognized: LF CRLF CR. However, special
5666 processing applies to CR:
5668 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5669 nor a local message in the state where . is a terminator.
5671 (ii) If a bare CR is encountered in a header line, an extra space is added
5672 after the line terminator so as not to end the header. The reasoning
5673 behind this is that bare CRs in header lines are most likely either
5674 to be mistakes, or people trying to play silly games.
5676 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5677 was being incorrectly given as 18 bytes larger than it should have been.
5678 This is a VOB (very old bug).
5680 60. This may never have affected anything current, but just in case it has:
5681 When the local host is found other than at the start of a list of hosts,
5682 the local host, those with the same MX, and any that follow, are discarded.
5683 When the list in question was part of a longer list of hosts, the following
5684 hosts (not currently being processed) were also being discarded. This no
5685 longer happens. I'm not sure if this situation could ever has previously
5688 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5691 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5692 header. This is something that is recommended in a new Internet Draft, and
5693 is something that is documented as being done by Sendmail. There are two
5694 possible values. For messages generated by the autoreply transport, Exim
5697 Auto-Submitted: auto-replied
5699 whereas for all other generated messages (e.g. bounces) it adds
5701 Auto-Submitted: auto-generated
5703 63. The "personal" condition in filters now includes a test for the
5704 Auto-Submitted: header. If it contains the string "auto-" the message it
5705 not considered personal.
5707 64. Added rcpt_include_affixes as a generic transport option.
5709 65. Added queue_only_override (default true).
5711 66. Added the syslog_duplication option.
5713 67. If what should have been the first header line of a message consisted of
5714 a space followed by a colon, Exim was mis-interpreting it as a header line.
5715 It isn't of course - it is syntactically invalid and should therefore be
5716 treated as the start of the message body. The misbehaviour could have
5717 caused a number of strange effects, including loss of data in subsequent
5718 header lines, and spool format errors.
5720 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5721 client host had authenticated. This control can now be exercised by an ACL
5722 for more flexibility.
5724 69. By default, callouts do not happen when testing with -bh. There is now a
5725 variant, -bhc, which does actually run the callout code, including
5726 consulting and updating the callout cache.
5728 70. Added support for saslauthd authentication, courtesy of Alexander
5731 71. If statvfs() failed on the spool or log directories while checking their
5732 size for availability, Exim confusingly gave the error "space shortage".
5733 Furthermore, in debugging mode it crashed with a floating point exception.
5734 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5735 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5736 serious problem, Exim now writes to the main and panic logs when this
5737 happens, with details of the failure. It then refuses to accept the
5738 incoming message, giving the message "spool directory problem" or "log
5739 directory problem" with a 421 code for SMTP messages.
5741 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5742 0, 1, and 2 exist, because some OS complain for execs without them (see
5743 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5744 descriptors. However, the code omitted to check that the open succeeded,
5745 causing mysterious errors if for some reason the permissions on /dev/null
5746 got screwed. Now Exim writes a message to the main and panic logs, and
5747 bombs out if it can't open /dev/null.
5749 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5750 interact so that it is all more flexible. It is supposed to remain
5751 backwards compatible. Also added extra_local_interfaces.
5753 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5754 to bomb out with an assertion failure - to the client this appears as a
5755 connection drop. This problem occurs in the part of the code that was taken
5756 from the Samba project. Fortunately, the assertion is in a very simple
5757 function, so I have fixed this by reproducing the function inline in the
5758 one place where it is called, and arranging for authentication to fail
5759 instead of killing the process with assert().
5761 75. The SPA client code was not working when the server requested OEM rather
5762 than Unicode encoding.
5764 76. Added code to make require_files with a specific uid setting more usable in
5765 the case where statting the file as root fails - usually a non-root-mounted
5766 NFS file system. When this happens and the failure is EACCES, Exim now
5767 forks a subprocess and does the per-uid checking as the relevant uid.
5769 77. Added process_log_path.
5771 78. If log_file_path was not explicitly set, a setting of check_log_space or
5772 check_log_inodes was ignored.
5774 79. If a space check for the spool or log partitions fails, the incident is now
5775 logged. Of course, in the latter case the data may get lost...
5777 80. Added the %p formatting code to string_format() so that it can be used to
5778 print addresses in debug_print(). Adjusted all the address printing in the
5779 debugging in store.c to use %p rather than %d.
5781 81. There was a concern that a line of code in smtp_in.c could overflow a
5782 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5783 initially expressed, the concern was not well-founded, because trailing
5784 spaces are removed early. However, if the trailing spaces were followed by
5785 a NULL, they did not get removed, so the overflow was possible. Two fixes
5788 (a) I re-wrote the offending code in a cleaner fashion.
5789 (b) If an incoming SMTP command contains a NULL character, it is rejected
5792 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5793 runs initgroups(), so that if the Exim user is in any additional groups,
5794 they will be used during message reception.
5800 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob