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 mainlog rather than paniclog.
142 Debug lines prefixed by "DCC: " to remove any ambiguity.
148 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
150 This, or similar/improved, will also be change PP/11 of 4.82.
156 PP/01 Handle short writes when writing local log-files.
157 In practice, only affects FreeBSD (8 onwards).
158 Bugzilla 1053, with thanks to Dmitry Isaikin.
160 NM/01 Bugzilla 949 - Documentation tweak
162 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
165 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
167 PP/02 Implemented gsasl authenticator.
169 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
171 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
172 `pkg-config foo` for cflags/libs.
174 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
175 with rest of GSASL and with heimdal_gssapi.
177 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
178 `pkg-config foo` for cflags/libs for the TLS implementation.
180 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
181 properties get this fed in as external SSF. A number of robustness
182 and debugging improvements to the cyrus_sasl authenticator.
184 PP/08 cyrus_sasl server now expands the server_realm option.
186 PP/09 Bugzilla 1214 - Log authentication information in reject log.
187 Patch by Jeremy Harris.
189 PP/10 Added dbmjz lookup type.
191 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
193 PP/12 MAIL args handles TAB as well as SP, for better interop with
194 non-compliant senders.
195 Analysis and variant patch by Todd Lyons.
197 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
198 Bug report from Lars Müller <lars@samba.org> (via SUSE),
199 Patch from Dirk Mueller <dmueller@suse.com>
201 PP/13 tls_peerdn now print-escaped for spool files.
202 Observed some $tls_peerdn in wild which contained \n, which resulted
203 in spool file corruption.
205 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
206 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
207 or write after TLS renegotiation, which otherwise led to messages
210 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
211 as a tracking header (ie: a signed header comes before the signature).
212 Patch from Wolfgang Breyha.
214 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
215 comma-sep list; embedded commas doubled.
217 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
219 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
221 Report and patch from Dmitry Banschikov.
223 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
224 Removed SSL_clear() after SSL_new() which led to protocol negotiation
225 failures. We appear to now support TLS1.1+ with Exim.
227 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
228 lets Exim select keys and certificates based upon TLS SNI from client.
229 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
230 before an outbound SMTP session. New log_selector, +tls_sni.
232 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
233 NULL dereference. Report and patch from Alun Jones.
235 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
236 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
237 Not seeing resolver debug output on NetBSD, but suspect this is a
238 resolver implementation change.
240 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
241 Left warnings. Added "eximon gdb" invocation mode.
243 PP/21 Defaulting "accept_8bitmime" to true, not false.
245 PP/22 Added -bw for inetd wait mode support.
247 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
248 locate the relevant includes and libraries. Made this the default.
250 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
251 Bugzilla 1246, report and most of solution from Tomasz Kusy.
253 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
254 This may cause build issues on older platforms.
256 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
257 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
258 gnutls_require_mac & gnutls_require_protocols (no longer supported).
259 Added SNI support via GnuTLS too.
260 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
262 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
264 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
265 Applied second patch from Janne, implementing suggestion to default
266 multiple-strings-in-record handling to match SPF spec.
268 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
270 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
271 read-only, out of scope).
272 Patch from Wolfgang Breyha, report from Stuart Northfield.
274 PP/29 Fix three issues highlighted by clang analyser static analysis.
275 Only crash-plausible issue would require the Cambridge-specific
276 iplookup router and a misconfiguration.
277 Report from Marcin Mirosław.
279 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
281 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
282 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
283 As part of this, removing so much warning spew let me fix some minor
284 real issues in debug logging.
286 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
287 assignment on my part. Fixed.
289 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
290 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
291 Janne Snabb (who went above and beyond: thank you).
293 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
294 string otherwise requires a connection and a bunch more work and it's
295 relatively easy to get wrong. Should also expose TLS library linkage
298 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
299 64-bit ${eval} (JH/03).
301 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
302 GNU libc to support some of the 64-bit stuff, should not lead to
303 conflicts. Defined before os.h is pulled in, so if a given platform
304 needs to override this, it can.
306 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
307 protection layer was required, which is not implemented.
308 Bugzilla 1254, patch from Wolfgang Breyha.
310 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
311 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
312 tls_dhparam take prime identifiers. Also unbreak combination of
313 OpenSSL+DH_params+TLSSNI.
315 PP/39 Disable SSLv2 by default in OpenSSL support.
321 PP/01 Solaris build fix for Oracle's LDAP libraries.
322 Bugzilla 1109, patch from Stephen Usher.
324 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
326 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
329 TF/02 Fix a couple more cases where we did not log the error message
330 when unlink() failed. See also change 4.74-TF/03.
332 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
333 lock up or crash if it happened to be inside a call to libc when it
334 got a SIGUSR1 from exiwhat.
336 The SIGUSR1 handler appends the current process status to the process
337 log which is later printed by exiwhat. It used to use the general
338 purpose logging code to do this, but several functions it calls are
339 not safe for signals.
341 The new output code in the SIGUSR1 handler is specific to the process
342 log, and simple enough that it's easy to inspect for signal safety.
343 Removing some special cases also simplifies the general logging code.
344 Removing the spurious timestamps from the process log simplifies
347 TF/04 Improved ratelimit ACL condition.
349 The /noupdate option has been deprecated in favour of /readonly which
350 has clearer semantics. The /leaky, /strict, and /readonly update modes
351 are mutually exclusive. The update mode is no longer included in the
352 database key; it just determines when the database is updated. (This
353 means that when you upgrde Exim will forget old rate measurements.)
355 Exim now checks that the per_* options are used with an update mode that
356 makes sense for the current ACL. For example, when Exim is processing a
357 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
358 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
359 must specify per_mail/readonly. If you omit the update mode it defaults to
360 /leaky where that makes sense (as before) or /readonly where required.
362 The /noupdate option is now undocumented but still supported for
363 backwards compatibility. It is equivalent to /readonly except that in
364 ACLs where /readonly is required you may specify /leaky/noupdate or
365 /strict/noupdate which are treated the same as /readonly.
367 A useful new feature is the /count= option. This is a generalization
368 of the per_byte option, so that you can measure the throughput of other
369 aggregate values. For example, the per_byte option is now equivalent
370 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
372 The per_rcpt option has been generalized using the /count= mechanism
373 (though it's more complicated than the per_byte equivalence). When it is
374 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
375 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
376 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
377 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
378 Note that using per_rcpt with a non-readonly update mode in more than
379 one ACL will cause the recipients to be double-counted. (The per_mail
380 and per_byte options don't have this problem.)
382 The handling of very low rates has changed slightly. If the computed rate
383 is less than the event's count (usually one) then this event is the first
384 after a long gap. In this case the rate is set to the same as this event's
385 count, so that the first message of a spam run is counted properly.
387 The major new feature is a mechanism for counting the rate of unique
388 events. The new per_addr option counts the number of different
389 recipients that someone has sent messages to in the last time period. It
390 behaves like per_rcpt if all the recipient addresses are different, but
391 duplicate recipient addresses do not increase the measured rate. Like
392 the /count= option this is a general mechanism, so the per_addr option
393 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
394 example, measure the rate that a client uses different sender addresses
395 with the options per_mail/unique=$sender_address. There are further
396 details in the main documentation.
398 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
400 TF/06 Removed a few PCRE remnants.
402 TF/07 Automatically extract Exim's version number from tags in the git
403 repository when doing development or release builds.
405 PP/02 Raise smtp_cmd_buffer_size to 16kB.
406 Bugzilla 879. Patch from Paul Fisher.
408 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
409 Heavily based on revision 40f9a89a from Simon Arlott's tree.
412 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
414 PP/05 Variable $av_failed, true if the AV scanner deferred.
415 Bugzilla 1078. Patch from John Horne.
417 PP/06 Stop make process more reliably on build failure.
418 Bugzilla 1087. Patch from Heiko Schlittermann.
420 PP/07 Make maildir_use_size_file an _expandable_ boolean.
421 Bugzilla 1089. Patch from Heiko Schlittermann.
423 PP/08 Handle ${run} returning more data than OS pipe buffer size.
424 Bugzilla 1131. Patch from Holger Weiß.
426 PP/09 Handle IPv6 addresses with SPF.
427 Bugzilla 860. Patch from Wolfgang Breyha.
429 PP/10 GnuTLS: support TLS 1.2 & 1.1.
431 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
434 PP/11 match_* no longer expand right-hand-side by default.
435 New compile-time build option, EXPAND_LISTMATCH_RHS.
436 New expansion conditions, "inlist", "inlisti".
438 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
440 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
442 PP/14 fix log_write() format string regression from TF/03.
443 Bugzilla 1152. Patch from Dmitry Isaikin.
449 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
451 PP/02 Harmonised TLS library version reporting; only show if debugging.
452 Layout now matches that introduced for other libraries in 4.74 PP/03.
454 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
456 PP/04 New "dns_use_edns0" global option.
458 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
461 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
462 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
464 TK/01 Updated PolarSSL code to 0.14.2.
465 Bugzilla 1097. Patch from Andreas Metzler.
467 PP/07 Catch divide-by-zero in ${eval:...}.
470 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
473 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
474 format-string attack -- SECURITY: remote arbitrary code execution.
476 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
477 time unintentionally subject to list matching rules, letting the header
478 cause arbitrary Exim lookups (of items which can occur in lists, *not*
479 arbitrary string expansion). This allowed for information disclosure.
481 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
482 INT_MIN/-1 -- value coerced to INT_MAX.
488 NM/01 Workround for PCRE version dependency in version reporting
491 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
492 This fixes portability to compilers other than gcc, notably
493 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
495 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
496 makefiles for portability to HP-UX and POSIX correctness.
498 PP/01 Permit LOOKUP_foo enabling on the make command-line.
499 Also via indented variable definition in the Makefile.
500 (Debugging by Oliver Heesakkers).
502 PP/02 Restore caching of spamd results with expanded spamd_address.
503 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
505 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
506 Improves build reliability. Fix from: Frank Elsner
508 NM/02 Fix wide character breakage in the rfc2047 coding
509 Fixes bug 1064. Patch from Andrey N. Oktyabrski
511 NM/03 Allow underscore in dnslist lookups
512 Fixes bug 1026. Patch from Graeme Fowler
514 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
515 Code patches from Adam Ciarcinski of NetBSD.
517 NM/04 Fixed exiqgrep to cope with mailq missing size issue
520 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
521 is logged, to avoid truncation. Patch from John Horne.
523 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
524 Patch from Jakob Hirsch.
526 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
527 SQL string expansion failure details.
528 Patch from Andrey Oktyabrski.
530 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
531 Patch from Simon Arlott.
533 PP/09 New lookups functionality failed to compile on old gcc which rejects
534 extern declarations in function scope.
535 Patch from Oliver Fleischmann
537 PP/10 Use sig_atomic_t for flags set from signal handlers.
538 Check getgroups() return and improve debugging.
539 Fixed developed for diagnosis in bug 927 (which turned out to be
542 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
543 Patch from Mark Zealey.
545 PP/12 Bugzilla 1056: Improved spamd server selection.
546 Patch from Mark Zealey.
548 PP/13 Bugzilla 1086: Deal with maildir quota file races.
549 Based on patch from Heiko Schlittermann.
551 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
552 Patch from Uwe Doering, sign-off by Michael Haardt.
554 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
555 variable declaration deep within a block. Bug and patch from
558 PP/15 lookups-Makefile IRIX compatibilty coercion.
560 PP/16 Make DISABLE_DKIM build knob functional.
562 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
563 Patch by Simon Arlott
565 TF/03 Fix valgrind.h portability to C89 compilers that do not support
566 variable argument macros. Our copy now differs from upstream.
572 TF/01 Failure to get a lock on a hints database can have serious
573 consequences so log it to the panic log.
575 TF/02 Log LMTP confirmation messages in the same way as SMTP,
576 controlled using the smtp_confirmation log selector.
578 TF/03 Include the error message when we fail to unlink a spool file.
580 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
581 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
582 for maintaining out-of-tree patches for some time.
584 PP/01 Bugzilla 139: Documentation and portability issues.
585 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
586 Handle per-OS dynamic-module compilation flags.
588 PP/02 Let /dev/null have normal permissions.
589 The 4.73 fixes were a little too stringent and complained about the
590 permissions on /dev/null. Exempt it from some checks.
591 Reported by Andreas M. Kirchwitz.
593 PP/03 Report version information for many libraries, including
594 Exim version information for dynamically loaded libraries. Created
595 version.h, now support a version extension string for distributors
596 who patch heavily. Dynamic module ABI change.
598 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
599 privilege escalation vulnerability whereby the Exim run-time user
600 can cause root to append content of the attacker's choosing to
603 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
606 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
607 If dropping privileges for untrusted macros, we disabled normal logging
608 on the basis that it would fail; for the Exim run-time user, this is not
609 the case, and it resulted in successful deliveries going unlogged.
610 Fixed. Reported by Andreas Metzler.
616 PP/01 Date: & Message-Id: revert to normally being appended to a message,
617 only prepend for the Resent-* case. Fixes regression introduced in
618 Exim 4.70 by NM/22 for Bugzilla 607.
620 PP/02 Include check_rfc2047_length in configure.default because we're seeing
621 increasing numbers of administrators be bitten by this.
623 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
625 PP/03 Bugzilla 994: added openssl_options main configuration option.
627 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
629 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
631 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
633 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
634 without a peer certificate, leading to a segfault because of an
635 assumption that peers always have certificates. Be a little more
636 paranoid. Problem reported by Martin Tscholak.
638 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
639 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
640 NB: ClamAV planning to remove STREAM in "middle of 2010".
641 CL also introduces -bmalware, various -d+acl logging additions and
642 more caution in buffer sizes.
644 PP/09 Implemented reverse_ip expansion operator.
646 PP/10 Bugzilla 937: provide a "debug" ACL control.
648 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
650 PP/12 Bugzilla 973: Implement --version.
652 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
654 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
656 PP/15 Bugzilla 816: support multiple condition rules on Routers.
658 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
659 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
660 ignore trailing whitespace.
662 JJ/02 prevent non-panic DKIM error from being sent to paniclog
664 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
667 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
668 Notification from Dr Andrew Aitchison.
670 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
671 ExtendedDetectionInfo response format.
672 Notification from John Horne.
674 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
677 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
678 XSL and documented dependency on system catalogs, with examples of how
681 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
684 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
685 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
686 configuration file which is writeable by the Exim user or group.
688 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
689 of configuration files to cover files specified with the -C option if
690 they are going to be used with root privileges, not just the default
693 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
694 option (effectively making it always true).
696 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
697 files to be used while preserving root privileges.
699 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
700 that rogue child processes cannot use them.
702 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
703 run-time user, instead of root.
705 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
706 Exim run-time user without dropping privileges.
708 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
709 result string, instead of calling string_vformat() twice with the same
712 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
713 for other users. Others should always drop root privileges if they use
714 -C on the command line, even for a whitelisted configure file.
716 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
718 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
724 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
725 $data_path, and $header_path variables; fixed documentation bugs and
728 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
729 exipick to access non-standard spools, including the "frozen" queue
732 NM/01 Bugzilla 965: Support mysql stored procedures.
733 Patch from Alain Williams
735 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
737 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
738 Patch from Andreas Metzler
740 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
741 Patch from Kirill Miazine
743 NM/05 Bugzilla 671: Added umask to procmail example.
745 JJ/03 installed exipick 20100323.0, fixing doc bug
747 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
748 directory. Notification and patch from Dan Rosenberg.
750 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
752 TK/02 Improve log output when DKIM signing operation fails.
754 MH/01 Treat the transport option dkim_domain as a colon separated
755 list, not as a single string, and sign the message with each element,
756 omitting multiple occurences of the same signer.
758 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
759 Bugzilla 985, 986. Patch by Simon Arlott
761 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
762 Patch by Simon Arlott
764 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
765 MBX locking. Notification from Dan Rosenberg.
771 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
773 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
775 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
777 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
779 NM/04 Bugzilla 915: Flag broken perl installation during build.
785 TK/01 Added patch by Johannes Berg that expands the main option
786 "spamd_address" if it starts with a dollar sign.
788 TK/02 Write list of recipients to X-Envelope-Sender header when building
789 the mbox-format spool file for content scanning (suggested by Jakob
792 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
793 (http://www.dcc-servers.net/) support via dccifd. Activated by
794 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
796 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
797 by Mark Daniel Reidel <mr@df.eu>.
799 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
800 When building exim an external PCRE library is now needed -
801 PCRE is a system library on the majority of modern systems.
802 See entry on PCRE_LIBS in EDITME file.
804 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
805 conversation. Added nologin parameter to request.
806 Patch contributed by Kirill Miazine.
808 TF/01 Do not log submission mode rewrites if they do not change the address.
810 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
812 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
813 log files in place. Contributed by Roberto Lima.
815 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
817 TF/03 Bugzilla 615: When checking the local_parts router precondition
818 after a local_part_suffix or local_part_prefix option, Exim now
819 does not use the address's named list lookup cache, since this
820 contains cached lookups for the whole local part.
822 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
823 Robert Millan. Documentation is in experimental-spec.txt.
825 TF/04 Bugzilla 668: Fix parallel build (make -j).
827 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
829 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
830 Patch provided by Jan Srzednicki.
832 TF/05 Leading white space used to be stripped from $spam_report which
833 wrecked the formatting. Now it is preserved.
835 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
836 that they are available at delivery time.
838 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
840 TF/08 TLS error reporting now respects the incoming_interface and
841 incoming_port log selectors.
843 TF/09 Produce a more useful error message if an SMTP transport's hosts
844 setting expands to an empty string.
846 NM/06 Bugzilla 744: EXPN did not work under TLS.
847 Patch provided by Phil Pennock.
849 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
850 Patch provided by Richard Godbee.
852 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
853 acl_smtp_notquit, added index entry.
855 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
856 Patch provided by Eugene Bujak.
858 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
859 accept(). Patch provided by Maxim Dounin.
861 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
862 Patch provided by Phil Pennock.
864 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
866 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
867 Patch provided by Brad "anomie" Jorsch.
869 NM/14 Bugzilla 622: Added timeout setting to transport filter.
870 Patch provided by Dean Brooks.
872 TK/05 Add native DKIM support (does not depend on external libraries).
874 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
875 Patch provided by Graeme Fowler.
877 NM/16 Bugzilla 851: Documentation example syntax fix.
879 NM/17 Changed NOTICE file to remove references to embedded PCRE.
881 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
884 NM/19 Bugzilla 745: TLS version reporting.
885 Patch provided by Phil Pennock.
887 NM/20 Bugzilla 167: bool: condition support.
888 Patch provided by Phil Pennock.
890 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
891 clients. Patch provided by Phil Pennock.
893 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
894 Patch provided by Brad "anomie" Jorsch.
896 NM/23 Bugzilla 687: Fix misparses in eximstats.
897 Patch provided by Heiko Schlittermann.
899 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
900 Patch provided by Heiko Schlittermann.
902 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
903 plus update to original patch.
905 NM/26 Bugzilla 799: Documentation correction for ratelimit.
907 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
908 Patch provided by David Brownlee.
910 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
912 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
914 NM/30 Bugzilla 888: TLS documentation bugfixes.
916 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
918 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
919 Unlike the original bugzilla I have changed all shell scripts in src tree.
921 NM/33 Bugzilla 898: Transport filter timeout fix.
922 Patch by Todd Rinaldo.
924 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
925 Patch by Serge Demonchaux.
927 NM/35 Bugzilla 39: Base64 decode bug fixes.
928 Patch by Jakob Hirsch.
930 NM/36 Bugzilla 909: Correct connect() call in dcc code.
932 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
934 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
936 NM/39 Bugzilla 911: Fixed MakeLinks build script.
942 TK/01 Add preliminary DKIM support. Currently requires a forked version of
943 ALT-N's libdkim that I have put here:
944 http://duncanthrax.net/exim-experimental/
946 Note to Michael Haardt: I had to rename some vars in sieve.c. They
947 were called 'true' and it seems that C99 defines that as a reserved
948 keyword to be used with 'bool' variable types. That means you could
949 not include C99-style headers which use bools without triggering
950 build errors in sieve.c.
952 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
953 as mailq or other aliases. Changed the --help handling significantly
954 to do whats expected. exim_usage() emits usage/help information.
956 SC/01 Added the -bylocaldomain option to eximstats.
958 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
960 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
962 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
968 PH/01 Another patch from the Sieve maintainer.
970 PH/02 When an IPv6 address is converted to a string for single-key lookup
971 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
972 dots are used instead of colons so that keys in lsearch files need not
973 contain colons. This was done some time before quoting was made available
974 in lsearch files. However, iplsearch files do require colons in IPv6 keys
975 (notated using the quote facility) so as to distinguish them from IPv4
976 keys. This meant that lookups for IP addresses in host lists did not work
977 for iplsearch lookups.
979 This has been fixed by arranging for IPv6 addresses to be expressed with
980 colons if the lookup type is iplsearch. This is not incompatible, because
981 previously such lookups could never work.
983 The situation is now rather anomolous, since one *can* have colons in
984 ordinary lsearch keys. However, making the change in all cases is
985 incompatible and would probably break a number of configurations.
987 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
990 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
991 conversion specification without a maximum field width, thereby enabling
992 a rogue spamd server to cause a buffer overflow. While nobody in their
993 right mind would setup Exim to query an untrusted spamd server, an
994 attacker that gains access to a server running spamd could potentially
995 exploit this vulnerability to run arbitrary code as the Exim user.
997 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
998 $primary_hostname instead of what libspf2 thinks the hosts name is.
1000 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1001 a directory entry by the name of the lookup key. Previously, if a
1002 symlink pointed to a non-existing file or a file in a directory that
1003 Exim lacked permissions to read, a lookup for a key matching that
1004 symlink would fail. Now it is enough that a matching directory entry
1005 exists, symlink or not. (Bugzilla 503.)
1007 PH/03 The body_linecount and body_zerocount variables are now exported in the
1010 PH/04 Added the $dnslist_matched variable.
1012 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1013 This means they are set thereafter only if the connection becomes
1016 PH/06 Added the client_condition to authenticators so that some can be skipped
1017 by clients under certain conditions.
1019 PH/07 The error message for a badly-placed control=no_multiline_responses left
1020 "_responses" off the end of the name.
1022 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1024 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1025 (without spaces) instead of just copying the configuration text.
1027 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1029 PH/11 Added $max_received_linelength.
1031 PH/12 Added +ignore_defer and +include_defer to host lists.
1033 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1034 way in which PCRE > 7.0 is built.
1036 PH/14 Implemented queue_only_load_latch.
1038 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1039 MAIL command. The effect was to mangle the value on 64-bit systems.
1041 PH/16 Another patch from the Sieve maintainer.
1043 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1045 PH/18 If a system quota error occurred while trying to create the file for
1046 a maildir delivery, the message "Mailbox is full" was not appended to the
1047 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1048 only to a quota excession during the actual writing of the file.
1050 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1051 characters?) which causes problems in log lines. The DN values are now
1052 passed through string_printing() before being added to log lines.
1054 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1055 and InterBase are left for another time.)
1057 PH/21 Added message_body_newlines option.
1059 PH/22 Guard against possible overflow in moan_check_errorcopy().
1061 PH/23 POSIX allows open() to be a macro; guard against that.
1063 PH/24 If the recipient of an error message contained an @ in the local part
1064 (suitably quoted, of course), incorrect values were put in $domain and
1065 $local_part during the evaluation of errors_copy.
1071 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1072 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1075 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1076 issue a MAIL command.
1078 PH/02 In an ACL statement such as
1080 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1082 if a client was not listed at all, or was listed with a value other than
1083 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1084 the condition was not true (as it should be), so access was not denied.
1085 The bug was that the ! inversion was incorrectly passed on to the second
1086 item. This has been fixed.
1088 PH/03 Added additional dnslists conditions == and =& which are different from
1089 = and & when the dns lookup returns more than one IP address.
1091 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1092 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1094 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1095 FSYNC, which compiles an option called disable_fsync that allows for
1096 bypassing fsync(). The documentation is heavily laced with warnings.
1098 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1100 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1101 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1102 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1103 including adding "make clean"; (3) Added -fPIC when compiling the test
1104 dynamically loaded module, to get rid of a warning.
1106 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1107 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1108 = 0s. The bug is otherwise harmless.
1110 PH/07 There was a bug in the dovecot authenticator such that the value of
1111 $auth1 could be overwritten, and so not correctly preserved, after a
1112 successful authentication. This usually meant that the value preserved by
1113 the server_setid option was incorrect.
1115 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1117 PH/09 Installed PCRE release 7.0.
1119 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1120 run for batched SMTP input. It is now run at the start of every message
1121 in the batch. While fixing this I discovered that the process information
1122 (output by running exiwhat) was not always getting set for -bs and -bS
1123 input. This is fixed, and it now also says "batched" for BSMTP.
1125 PH/11 Added control=no_pipelining.
1127 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1128 patch, slightly modified), and move the expansion of helo_data till after
1129 the connection is made in the smtp transport (so it can use these
1132 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1134 PH/14 Added log_selector = +pid.
1136 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1138 PH/16 Add ${if forany and ${if forall.
1140 PH/17 Added dsn_from option to vary the From: line in DSNs.
1142 PH/18 Flush SMTP output before performing a callout, unless control =
1143 no_callout_flush is set.
1145 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1146 was true (the default) a successful delivery failed to delete the retry
1147 item, thus causing premature timeout of the address. The bug is now
1150 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1152 PH/21 Long custom messages for fakedefer and fakereject are now split up
1153 into multiline reponses in the same way that messages for "deny" and
1154 other ACL rejections are.
1156 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1157 with slight modification.
1159 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1160 draft, changing the syntax and factoring some duplicate code".
1162 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1163 for deliveries of the second and subsequent messages over the same SMTP
1166 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1167 ${reduce, with only minor "tidies".
1169 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1171 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1172 expansion side effects.
1174 PH/27 When a message times out after an over-quota error from an Exim-imposed
1175 quota, the bounce message says "mailbox is full". This message was not
1176 being given when it was a system quota that was exceeded. It now should
1179 MH/03 Made $recipients available in local_scan(). local_scan() already has
1180 better access to the recipient list through recipients_list[], but
1181 $recipients can be useful in postmaster-provided expansion strings.
1183 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1184 in the case of a MAIL command with additional options following the
1185 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1186 were accidentally chopped off.
1188 PH/29 SMTP synchronization checks are implemented when a command is read -
1189 there is a check that no more input is waiting when there shouldn't be
1190 any. However, for some commands, a delay in an ACL can mean that it is
1191 some time before the response is written. In this time, more input might
1192 arrive, invalidly. So now there are extra checks after an ACL has run for
1193 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1194 pipelining has not been advertised.
1196 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1198 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1199 returned in the SMTP response when smtp_return_error_details was false.
1200 This has been fixed.
1202 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1203 instead of the C I/O that was originally supplied, because problems were
1204 reported on Solaris.
1206 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1207 Exim which did not show up earlier: it was assuming that a call to
1208 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1209 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1210 was a macro that became an assignment, so it seemed to work. This has
1211 changed to a proper function call with a void return, hence the compile
1212 error. Exim's code has been fixed.
1214 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1217 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1219 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1221 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1222 because it turns out to be too big or there is a timeout) there is no
1223 "Completed" line in the log. When some messages of this type were
1224 selected by exigrep, they were listed as "not completed". Others were
1225 picked up by some special patterns. I have improved the selection
1226 criteria to be more general.
1228 PH/38 The host_find_failed option in the manualroute router can now be set
1229 to "ignore", to completely ignore a host whose IP address cannot be
1230 found. If all hosts are ignored, the behaviour is controlled by the new
1231 host_all_ignored option.
1233 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1234 homing or because of multiple MX records with /mx) generated more than
1235 one IP address, and the following item turned out to be the local host,
1236 all the secondary addresses of the first item were incorrectly removed
1237 from the list, along with the local host and any following hosts (which
1238 is what is supposed to happen).
1240 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1241 whoever called Exim into the -H file. In the case of the daemon it was
1242 behaving confusingly. When first started, it used values for whoever
1243 started the daemon, but after a SIGHUP it used the Exim user (because it
1244 calls itself on a restart). I have changed the code so that it now always
1247 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1248 message are rejected with the same error (e.g. no authentication or bad
1249 sender address), and a DATA command is nevertheless sent (as can happen
1250 with PIPELINING or a stupid MUA), the error message that was given to the
1251 RCPT commands is included in the rejection of the DATA command. This is
1252 intended to be helpful for MUAs that show only the final error to their
1255 PH/42 Another patch from the Sieve maintainer.
1257 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1258 Eximstats - Fixed some broken HTML links and added missing column headers
1260 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1261 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1263 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1265 PH/43 Yet another patch from the Sieve maintainer.
1267 PH/44 I found a way to check for a TCP/IP connection going away before sending
1268 the response to the final '.' that terminates a message, but only in the
1269 case where the client has not sent further data following the '.'
1270 (unfortunately, this is allowed). However, in many cases there won't be
1271 any further data because there won't be any more messages to send. A call
1272 to select() can be used: if it shows that the input is "ready", there is
1273 either input waiting, or the socket has been closed. An attempt to read
1274 the next input character can distinguish the two cases. Previously, Exim
1275 would have sent an OK response which the client would never have see.
1276 This could lead to message repetition. This fix should cure that, at
1277 least in a lot of common cases.
1279 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1280 advertised in response to EHLO.
1286 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1287 fixed by 4.65/MH/01 (is this a record?) are fixed:
1289 (i) An empty string was always treated as zero by the numeric comparison
1290 operators. This behaviour has been restored.
1292 (ii) It is documented that the numeric comparison operators always treat
1293 their arguments as decimal numbers. This was broken in that numbers
1294 starting with 0 were being interpreted as octal.
1296 While fixing these problems I realized that there was another issue that
1297 hadn't been noticed. Values of message_size_limit (both the global option
1298 and the transport option) were treated as octal if they started with 0.
1299 The documentation was vague. These values are now always treated as
1300 decimal, and I will make that clear in the documentation.
1306 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1307 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1310 MH/01 Don't check that the operands of numeric comparison operators are
1311 integers when their expansion is in "skipping" mode (fixes bug
1312 introduced by 4.64-PH/07).
1314 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1315 child addresses, Exim now panics and dies. Previously, because the count
1316 is held in a short int, deliveries were likely to be lost. As such a
1317 large number of recipients for a single message is ridiculous
1318 (performance will be very, very poor), I have chosen to impose a limit
1319 rather than extend the field.
1325 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1326 leftover -K file (the existence of which was triggered by #402).
1327 While we were at it, introduced process PID as part of the -K
1328 filename. This should rule out race conditions when creating
1331 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1332 processing considerably. Previous code took too long for large mails,
1333 triggering a timeout which in turn triggers #401.
1335 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1336 in the DK code in transports.c. sendfile() is not really portable,
1337 hence the _LINUX specificness.
1339 TF/01 In the add_headers option to the mail command in an Exim filter,
1340 there was a bug that Exim would claim a syntax error in any
1341 header after the first one which had an odd number of characters
1344 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1345 callout verification, Exim cached a "reject" for the entire domain. This
1346 is correct for most verifications, but it is not correct for a recipient
1347 verification with use_sender or use_postmaster set, because in that case
1348 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1349 case of MAIL FROM:<> rejection from other early rejections (e.g.
1350 rejection of HELO). When verifying a recipient using a non-null MAIL
1351 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1352 Whatever the result of the callout, the value of the domain cache is
1353 left unchanged (for any other kind of callout, getting as far as trying
1354 RCPT means that the domain itself is ok).
1356 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1359 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1360 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1361 session, the daemon ignores ECONNECT errors and logs others; it now
1362 ignores EPIPE as well.
1364 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1365 (quoted-printable decoding).
1367 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1368 later the small subsequent patch to fix an introduced bug.
1370 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1372 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1374 PH/08 An error is now given if message_size_limit is specified negative.
1376 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1377 to be given (somewhat) arbitrary names.
1379 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1382 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1383 miscellaneous code fixes
1385 PH/10 Added the log_reject_target ACL modifier to specify where to log
1388 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1389 hostname. This is wrong, because it relates to the incoming message (and
1390 probably the interface on which it is arriving) and not to the outgoing
1391 callout (which could be using a different interface). This has been
1392 changed to use the value of the helo_data option from the smtp transport
1393 instead - this is what is used when a message is actually being sent. If
1394 there is no remote transport (possible with a router that sets up host
1395 addresses), $smtp_active_hostname is used.
1397 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1398 tweaks were necessary in order to get it to work (see also 21 below):
1399 (a) The code assumed that strncpy() returns a negative number on buffer
1400 overflow, which isn't the case. Replaced with Exim's string_format()
1402 (b) There were several signed/unsigned issues. I just did the minimum
1403 hacking in of casts. There is scope for a larger refactoring.
1404 (c) The code used strcasecmp() which is not a standard C function.
1405 Replaced with Exim's strcmpic() function.
1406 (d) The code set only $1; it now sets $auth1 as well.
1407 (e) A simple test gave the error "authentication client didn't specify
1408 service in request". It would seem that Dovecot has changed its
1409 interface. Fortunately there's a specification; I followed it and
1410 changed what the client sends and it appears to be working now.
1412 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1415 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1416 address A is aliased to B and C, where B exists and C does not. Without
1417 -v the output is "A verified" because verification stops after a
1418 successful redirection if more than one address is generated. However,
1419 with -v the child addresses are also verified. Exim was outputting "A
1420 failed to verify" and then showing the successful verification for C,
1421 with its parentage. It now outputs "B failed to verify", showing B's
1422 parentage before showing the successful verification of C.
1424 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1425 look up a TXT record in a specific list after matching in a combined
1428 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1429 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1430 they consult the DNS. I had assumed they would set it the way they
1431 wanted; and indeed my experiments on Linux seem to show that in some
1432 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1433 To be on the safe side, however, I have now made the interface to
1434 host_find_byname() similar to host_find_bydns(), with an argument
1435 containing the DNS resolver options. The host_find_byname() function now
1436 sets these options at its start, just as host_find_bydns() does. The smtp
1437 transport options dns_qualify_single and dns_search_parents are passed to
1438 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1439 of host_find_byname() use the default settings of RES_DEFNAMES
1440 (qualify_single) but not RES_DNSRCH (search_parents).
1442 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1443 spool_read_header() do less string testing, by means of a preliminary
1444 switch on the second character of optional "-foo" lines. (This is
1445 overdue, caused by the large number of possibilities that now exist.
1446 Originally there were few.) While I was there, I also converted the
1447 str(n)cmp tests so they don't re-test the leading "-" and the first
1448 character, in the hope this might squeeze out yet more improvement.
1450 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1451 flag allowing group syntax was set by the header_syntax check but not
1452 turned off, possible causing trouble later; (2) The flag was not being
1453 set at all for the header_verify test, causing "group"-style headers to
1454 be rejected. I have now set it in this case, and also caused header_
1455 verify to ignore an empty address taken from a group. While doing this, I
1456 came across some other cases where the code for allowing group syntax
1457 while scanning a header line wasn't quite right (mostly, not resetting
1458 the flag correctly in the right place). These bugs could have caused
1459 trouble for malformed header lines. I hope it is now all correct.
1461 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1462 with the "reply" argument non-NULL. The code, however (which originally
1463 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1464 but it didn't always do it. This confused somebody who was copying the
1465 code for some other use. I have removed all the tests.
1467 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1468 feature that was used to support insecure browsers during the U.S. crypto
1469 embargo. It requires special client support, and Exim is probably the
1470 only MTA that supported it -- and would never use it because real RSA is
1471 always available. This code has been removed, because it had the bad
1472 effect of slowing Exim down by computing (never used) parameters for the
1473 RSA_EXPORT functionality.
1475 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1476 authenticator to fail if there's a tab character in the incoming data
1477 (there should never be unless someone is messing about, as it's supposed
1478 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1479 if the connection is using TLS or if the remote IP is the same as the
1480 local IP, and the "valid-client-cert option" if a client certificate has
1483 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1484 authenticators. This can be used for authorization after authentication
1485 succeeds. (In the case of plaintext, it servers for both authentication
1488 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1489 if any retry times were supplied.
1491 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1492 connection, where there is no client IP address to check. In this
1493 situation, the verify now always succeeds.
1495 PH/25 Applied John Jetmore's -Mset patch.
1497 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1499 PH/27 In a string expansion for a processed (not raw) header when multiple
1500 headers of the same name were present, leading whitespace was being
1501 removed from all of them, but trailing whitespace was being removed only
1502 from the last one. Now trailing whitespace is removed from each header
1503 before concatenation. Completely empty headers in a concatenation (as
1504 before) are ignored.
1506 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1507 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1509 PH/29 [Removed. This was a change that I later backed out, and forgot to
1510 correct the ChangeLog entry (that I had efficiently created) before
1511 committing the later change.]
1513 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1514 address errors (4xx response to RCPT) over the same connection as other
1515 messages routed to the same hosts. Such deliveries are always "forced",
1516 so retry times are not inspected. This resulted in far too many retries
1517 for the affected addresses. The effect occurred only when there were more
1518 hosts than the hosts_max_try setting in the smtp transport when it had
1519 the 4xx errors. Those hosts that it had tried were not added to the list
1520 of hosts for which the message was waiting, so if all were tried, there
1521 was no problem. Two fixes have been applied:
1523 (i) If there are any address or message errors in an SMTP delivery, none
1524 of the hosts (tried or untried) are now added to the list of hosts
1525 for which the message is waiting, so the message should not be a
1526 candidate for sending over the same connection that was used for a
1527 successful delivery of some other message. This seems entirely
1528 reasonable: after all the message is NOT "waiting for some host".
1529 This is so "obvious" that I'm not sure why it wasn't done
1530 previously. Hope I haven't missed anything, but it can't do any
1531 harm, as the worst effect is to miss an optimization.
1533 (ii) If, despite (i), such a delivery is accidentally attempted, the
1534 routing retry time is respected, so at least it doesn't keep
1535 hammering the server.
1537 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1538 in ${readsocket because some servers need this prod.
1540 PH/32 Added some extra debug output when updating a wait-xxx database.
1542 PH/33 The hint "could be header name not terminated by colon", which has been
1543 given for certain expansion errors for a long time, was not being given
1544 for the ${if def:h_colon_omitted{... case.
1546 PH/34 The spec says: "With one important exception, whenever a domain list is
1547 being scanned, $domain contains the subject domain." There was at least
1548 one case where this was not true.
1550 PH/35 The error "getsockname() failed: connection reset by peer" was being
1551 written to the panic log as well as the main log, but it isn't really
1552 panic-worthy as it just means the connection died rather early on. I have
1553 removed the panic log writing for the ECONNRESET error when getsockname()
1556 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1557 runs only) independently of the message's sender address. This meant
1558 that, if the 4xx error was in fact related to the sender, a different
1559 message to the same recipient with a different sender could confuse
1560 things. In particualar, this can happen when sending to a greylisting
1561 server, but other circumstances could also provoke similar problems.
1562 I have changed the default so that the retry time for these errors is now
1563 based a combination of the sender and recipient addresses. This change
1564 can be overridden by setting address_retry_include_sender=false in the
1567 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1568 remote server are returned as part of bounce messages. This was not
1569 happening for LMTP over a pipe (the lmtp transport), but now it is the
1570 same for both kinds of LMTP.
1572 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1573 in header lines that were in fact CNAMEs. This is no longer the case.
1575 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1576 and queue runs started by the daemon processed all messages. This has
1577 been fixed so that -R and -S can now usefully be given with -q<time>.
1579 PH/40 Import PCRE release 6.7 (fixes some bugs).
1581 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1583 PH/42 Give an error if -q is specified more than once.
1585 PH/43 Renamed the variables $interface_address and $interface_port as
1586 $received_ip_address and $received_port, to make it clear that these
1587 values apply to message reception, and not to the outgoing interface when
1588 a message is delivered. (The old names remain recognized, of course.)
1590 PH/44 There was no timeout on the connect() call when using a Unix domain
1591 socket in the ${readsocket expansion. There now is.
1593 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1594 be meaningful with "accept".
1596 SC/01 Eximstats V1.43
1597 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1599 SC/02 Eximstats V1.44
1600 Use a glob alias rather than an array ref in the generated
1601 parser. This improves both readability and performance.
1603 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1604 Collect SpamAssassin and rejection statistics.
1605 Don't display local sender or destination tables unless
1606 there is data to show.
1607 Added average volumes into the top table text output.
1609 SC/04 Eximstats V1.46
1610 Collect data on the number of addresses (recipients)
1611 as well as the number of messages.
1613 SC/05 Eximstats V1.47
1614 Added 'Message too big' to the list of mail rejection
1615 reasons (thanks to Marco Gaiarin).
1617 SC/06 Eximstats V1.48
1618 Mainlog lines which have GMT offsets and are too short to
1619 have a flag are now skipped.
1621 SC/07 Eximstats V1.49 (Alain Williams)
1622 Added the -emptyok flag.
1624 SC/08 Eximstats V1.50
1625 Fixes for obtaining the IP address from reject messages.
1627 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1628 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1629 whitesspace changes from 4.64-PH/27
1631 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1634 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1635 are found, allow negative numbers in numeric criteria)
1637 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1639 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1640 to match changes made in 4.64-PH/43
1642 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1644 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1645 "domain" is actually an IP address, and force a failure. This locks out
1646 those revolvers/nameservers that support "A-for-A" lookups, in
1647 contravention of the specifications.
1649 PH/48 When a host name was looked up from an IP address, and the subsequent
1650 forward lookup of the name timed out, the host name was left in
1651 $sender_host_name, contrary to the specification.
1653 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1654 restricted to single-key lookups, Exim was not diagnosing an error if
1655 * or *@ was used with a query-style lookup.
1657 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1659 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1660 long ago, but noone interested enough thought of it. Let's just say that
1661 the "1.1" means that there are some new functions that weren't there at
1662 some point in the past.
1664 PH/51 Error processing for expansion failure of helo_data from an smtp
1665 transport during callout processing was broken.
1667 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1668 tested/used via the -bh/-bhc/-bs options.
1670 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1671 bug, fixed in subsequent PCRE releases).
1673 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1674 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1676 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1682 SC/01 Use a glob alias rather than an array ref in eximstats generated
1683 parser. This improves both readability and performance.
1685 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1686 Don't display local sender or destination tables in eximstats unless
1687 there is data to show.
1688 Added average volumes into the eximstats top table text output.
1690 SC/03 Collect data on the number of addresses (recipients) as well
1691 as the number of messages in eximstats.
1693 TF/01 Correct an error in the documentation for the redirect router. Exim
1694 does (usually) call initgroups() when daemonizing.
1696 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1697 with consistent privilege compared to when running as a daemon.
1699 TF/03 Note in the spec that $authenticated_id is not set for local
1700 submissions from trusted users.
1702 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1703 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1705 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1706 by adding some example configuration directives to the default
1707 configuration file. A little bit of work is required to uncomment the
1708 directives and define how usernames and passwords are checked, but
1709 there is now a framework to start from.
1711 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1712 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1713 without this. I don't know how relevant this is to other LDAP libraries.
1715 PH/02 Add the verb name to the "unknown ACL verb" error.
1717 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1719 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1721 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1722 directory not expanded when it should be if an expanded home directory
1723 was set for the address (which is overridden by the transport).
1725 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1728 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1729 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1730 because it is too late at that time, and has no effect.
1732 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1733 security issue with \' (bugzilla #107). I could not use the
1734 PQescapeStringConn() function, because it needs a PGconn value as one of
1737 PH/08 When testing addresses using -bt, indicate those final addresses that
1738 are duplicates that would not cause an additional delivery. At least one
1739 person was confused, thinking that -bt output corresponded to deliveries.
1740 (Suppressing duplicates isn't a good idea as you lose the information
1741 about possibly different redirections that led to the duplicates.)
1743 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1744 systems where poll() doesn't work, in particular OS X.
1746 PH/10 Added more information to debugging output for retry time not reached.
1748 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1749 operations in malware.c.
1751 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1754 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1755 syslog_duplication set false, log lines that would normally be written
1756 both the the main log and to the reject log were not written to syslog at
1759 PH/14 In the default configuration, change the use of "message" in ACL warn
1760 statements to "add_header".
1762 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1763 not followed by a command (e.g. "seen endif").
1765 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1766 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1769 PH/17 Added extra conditions to the default value of delay_warning_condition
1773 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1774 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1775 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1778 The Auto-Submitted: and various List- headers are standardised, whereas I
1779 don't think Precedence: ever was.
1781 PH/18 Refactored debugging code in route_finduser() to show more information,
1782 in particular, the error code if getpwnam() issues one.
1784 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1785 This is apparently needed in addition to the PH/07 change above to avoid
1786 any possible encoding problems.
1788 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1789 but not after initializing Perl.
1791 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1792 output them only if debugging. By default they are written stderr,
1793 apparently, which is not desirable.
1795 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1798 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1801 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1803 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1804 authenticated or an ident call has been made. Suppress the default
1805 values for $authenticated_id and $authenticated_sender (but permit -oMai
1806 and -oMas) when testing with -bh.
1808 PH/24 Re-jigged the order of the tests in the default configuration so that the
1809 tests for valid domains and recipients precede the DNS black list and CSA
1810 tests, on the grounds that those ones are more expensive.
1812 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1813 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1814 This bug exists in every version of Exim that I still have, right back to
1817 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1818 However, an attempt to turn on case-sensitivity in a regex key by
1819 including (?-i) didn't work because the subject string was already
1820 lowercased, and the effects were non-intuitive. It turns out that a
1821 one-line patch can be used to allow (?-i) to work as expected.
1827 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1828 other effects) broke the use of negated acl sub-conditions.
1830 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1833 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1834 "Deny" causes Exim to reject the incoming connection with a 554 error.
1835 Unfortunately, if there is a major crisis, such as a disk failure,
1836 tcp-wrappers gives "deny", whereas what one would like would be some
1837 kind of temporary error. A kludge has been added to help with this.
1838 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1839 554 error is used if errno is still zero or contains ENOENT (which occurs
1840 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1843 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1845 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1846 errors. Otherwise a message that provokes a temporary error (when other
1847 messages do not) can cause a whole host to time out.
1849 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1850 addresses were routed directly to files or pipes from a redirect router.
1851 File deliveries just didn't batch; pipe deliveries might have suffered
1854 PH/06 A failure to get a lock for a hints database would erroneously always say
1855 "Failed to get write lock", even when it was really a read lock.
1857 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1858 of 0600. This has been changed to use the value of the lockfile_mode
1859 option (which defaults to 0600).
1861 PH/08 Applied small patch from the Sieve maintainer.
1863 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1864 folder from quota calculations, a direct delivery into this folder messed
1865 up the contents of the maildirsize file. This was because the regex was
1866 used only to exclude .Trash (or whatever) when the size of the mailbox
1867 was calculated. There was no check that a delivery was happening into an
1868 excluded directory. This bug has been fixed by ignoring all quota
1869 processing for deliveries into excluded directories.
1871 PH/10 Added the maildirfolder_create_regex option to appendfile.
1877 PH/01 The code for finding all the local interface addresses on a FreeBSD
1878 system running IPv6 was broken. This may well have applied to all BSD
1879 systems, as well as to others that have similar system calls. The broken
1880 code found IPv4 interfaces correctly, but gave incorrect values for the
1881 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1882 that it would not match correctly against @[] and not recognize the IPv6
1885 PH/02 The ipliteral router was not recognizing addresses of the form user@
1886 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1888 PH/03 Added disable_ipv6.
1890 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1891 decoded form, because it is most often used to construct To: headers
1892 lines in autoreplies, and the decoded form may well be syntactically
1893 invalid. However, $reply_address has leading white space removed, and all
1894 newlines turned into spaces so that the autoreply transport does not
1897 PH/05 If group was specified without a user on a router, and no group or user
1898 was specified on a transport, the group from the router was ignored.
1900 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1901 for visible compile-time settings that can be used to change these
1902 numbers, for those that want even more. Backwards compatibility with old
1903 spool files has been maintained. However, going back to a previous Exim
1904 release will lost any variables that are in spool files.
1906 PH/07 Two small changes when running in the test harness: increase delay when
1907 passing a TCP/IP connection to a new process, in case the original
1908 process has to generate a bounce, and remove special handling of
1909 127.0.0.2 (sic), which is no longer necessary.
1911 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1912 be the same on different OS.
1914 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1917 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1918 whether --show-vars was specified or not
1920 JJ/02 exipick: Added support for new ACL variable spool format introduced
1923 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1924 syntactically invalid From: or Reply-to: line, and a filter used this to
1925 generate an autoreply, and therefore failed to obtain an address for the
1926 autoreply, Exim could try to deliver to a non-existent relative file
1927 name, causing unrelated and misleading errors. What now happens is that
1928 it logs this as a hard delivery error, but does not attempt to create a
1931 PH/11 The exinext utility has a -C option for testing purposes, but although
1932 the given file was scanned by exinext itself; it wasn't being passed on
1933 when Exim was called.
1935 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1936 an end-of-file indication when reading a command response.
1938 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1939 compiled. In many other places in Exim, IPv6 addresses are always
1940 recognized, so I have changed this. It also means that IPv4 domain
1941 literals of the form [IPV4:n.n.n.n] are now always recognized.
1943 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1944 used if the router is not running as root, for example, when verifying at
1945 ACL time, or when using -bh. The debugging output from this situation was
1946 non-existent - all you got was a failure to exec. I have made two
1949 (a) Failures to set uid/gid, the current directory, or a process leader
1950 in a subprocess such as that created by queryprogram now generate
1951 suitable debugging ouput when -d is set.
1953 (b) The queryprogram router detects when it is not running as root,
1954 outputs suitable debugging information if -d is set, and then runs
1955 the subprocess without attempting to change uid/gid.
1957 PH/15 Minor change to Makefile for building test_host (undocumented testing
1960 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1961 additional section of a DNS packet that returns MX or SRV records.
1962 Instead, it always explicitly searches for A/AAAA records. This avoids
1963 major problems that occur when a DNS server includes only records of one
1964 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1965 fixed another bug: if SRV records were looked up and the corresponding
1966 address records were *not* found in the additional section, the port
1967 values from the SRV records were lost.
1969 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1970 using the correct key (the original address) when searching the retry
1971 rules in order to find which one to use for generating the retry hint.
1973 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1974 adding the default one. Similarly, if it contains a Reply-To: header, the
1975 errors_reply_to option, if set, is not used.
1977 PH/19 When calculating a retry time, Exim used to measure the "time since
1978 failure" by looking at the "first failed" field in the retry record. Now
1979 it does not use this if it is later than than the arrival time of the
1980 message. Instead it uses the arrival time. This makes for better
1981 behaviour in cases where some deliveries succeed, thus re-setting the
1982 "first failed" field. An example is a quota failure for a huge message
1983 when small messages continue to be delivered. Without this change, the
1984 "time since failure" will always be short, possible causing more frequent
1985 delivery attempts for the huge message than are intended.
1986 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1988 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1989 $1, $2, $3) because the numerical variables can be reset during some
1990 expansion items (e.g. "match"), thereby losing the authentication data.
1992 PH/21 Make -bV show the size of off_t variables so that the test suite can
1993 decide whether to run tests for quotas > 2G.
1995 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1996 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1997 filecount value is greater than 2G or if a quota value is greater than 2G
1998 on a system where the size of off_t is not greater than 4, a panic error
2001 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2002 never match. The debug and -bh output now contains an explicit error
2003 message indicating a malformed IPv4 address or mask.
2005 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2006 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2007 PH/23 above applies.
2009 PH/25 Do not write to syslog when running in the test harness. The only
2010 occasion when this arises is a failure to open the main or panic logs
2011 (for which there is an explicit test).
2013 PH/26 Added the /no_tell option to "control=freeze".
2015 PH/27 If a host name lookup failed very early in a connection, for example, if
2016 the IP address matched host_lookup and the reverse lookup yielded a name
2017 that did not have a forward lookup, an error message of the form "no IP
2018 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2019 could be logged. Now it outputs the IP address instead of "NULL".
2021 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2022 allows the sender and the authenticated sender to be set when
2023 submitting a message from within Exim. Since child_open_exim() is
2024 documented for local_scan(), the new function should be too.
2026 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2027 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2028 results in an empty string is now treated as unset.
2030 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2032 PH/31 Added sender_verify_fail logging option.
2034 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2035 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2036 tidied the source and removed it altogether.
2038 PH/33 When a queue run was abandoned because the load average was too high, a
2039 log line was always written; now it is written only if the queue_run log
2040 selector is set. In addition, the log line for abandonment now contains
2041 information about the queue run such as the pid. This is always present
2042 in "start" and "stop" lines but was omitted from the "abandon" line.
2044 PH/34 Omit spaces between a header name and the colon in the error message that
2045 is given when verify = headers_syntax fails (if there are lots of them,
2046 the message gets confusing).
2048 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2049 names, as there are now some PTR records that contain slashes. This check
2050 is only to protect against broken name servers that fall over on strange
2051 characters, so the fact that it applies to all lookups doesn't matter.
2053 PH/36 Now that the new test suite is complete, we can remove some of the
2054 special code in Exim that was needed for the old test suite. For example,
2055 sorting DNS records because real resolvers return them in an arbitrary
2056 order. The new test suite's fake resolver always returns records in the
2059 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2060 bounces) except when queue_only is set, to avoid logging races between
2061 the different processes.
2063 PH/38 Panic-die if .include specifies a non-absolute path.
2065 PH/39 A tweak to the "H" retry rule from its user.
2067 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2068 a label. They prevented compilation on older perls.
2070 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2071 a warning to be raised on newish perls.
2073 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2074 on queue. Changes to match documented behaviour of showing count of
2075 messages matching specified criteria.
2077 PH/40 Changed the default ident timeout from 30s to 5s.
2079 PH/41 Added support for the use of login_cap features, on those BSD systems
2080 that have them, for controlling the resources used by pipe deliveries.
2082 PH/42 The content-scanning code uses fopen() to create files in which to put
2083 message data. Previously it was not paying any attention to the mode of
2084 the files. Exim runs with umask(0) because the rest of the code creates
2085 files with open(), and sets the required mode explicitly. Thus, these
2086 files were ending up world-writeable. This was not a big issue, because,
2087 being within the spool directory, they were not world-accessible. I have
2088 created a function called modefopen, which takes an additional mode
2089 argument. It sets umask(777), creates the file, chmods it to the required
2090 mode, then resets the umask. All the relevant calls to fopen() in the
2091 content scanning code have been changed to use this function.
2093 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2094 to 24 hours. This avoids potential overflow problems when processing G
2095 and H retry rules. I suspect nobody ever tinkers with this value.
2097 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2099 PH/45 When the plaintext authenticator is running as a client, the server's
2100 challenges are checked to ensure they are valid base64 strings. By
2101 default, the authentication attempt is cancelled if an invalid string is
2102 received. Setting client_ignore_invalid_base64 true ignores these errors.
2103 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2104 they are received. Thus, the responses can be made to depend on the
2105 challenges. If an invalid string is ignored, an empty string is placed in
2108 PH/46 Messages that are created by the autoreply transport now contains a
2109 References: header, in accordance with RFCs 2822 and 3834.
2111 PH/47 Added authenticated_sender_force to the smtp transport.
2113 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2115 PH/49 Installed latest patch from the Sieve maintainer.
2117 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2118 was also set, the appendfile transport was unnecessarily scanning a
2119 directory of message files (e.g. for maildir delivery) to find the count
2120 of files (along with the size), even though it did not need this
2121 information. It now does the scan only if it needs to find either the
2122 size of the count of files.
2124 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2126 PH/52 Two bugs concerned with error handling when the smtp transport is
2129 (i) Exim was not creating retry information for temporary errors given
2130 for individual recipients after the DATA command when the smtp transport
2131 was used in LMTP mode. This meant that they could be retried too
2132 frequently, and not timed out correctly.
2134 (ii) Exim was setting the flag that allows error details to be returned
2135 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2136 recipients that were returned after the DATA command.
2138 PH/53 This is related to PH/52, but is more general: for any failing address,
2139 when detailed error information was permitted to be returned to the
2140 sender, but the error was temporary, then after the final timeout, only
2141 "retry timeout exceeded" was returned. Now it returns the full error as
2142 well as "retry timeout exceeded".
2144 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2145 do this, and (what is worse) MTAs that accept it.
2147 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2148 will now be deprecated.
2150 PH/56 New os.c-cygwin from the Cygwin maintainer.
2152 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2153 formats (previously only available in exim formats via -bpr, -bpru,
2154 and -bpra. Now also available in native and exiqgrep formats)
2156 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2157 with very large, slow to parse queues
2159 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2161 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2163 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2164 responses to authentication challenges, though it was showing the
2165 challenges; (ii) I've removed the CR characters from the debug output for
2168 PH/58 Allow for the insertion of a newline as well as a space when a string
2169 is turned into more than one encoded-word during RFC 2047 encoding. The
2170 Sieve code now uses this.
2172 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2173 data_4xx, lost_connection, tls_required.
2175 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2176 message was being sent as an SMTP response.
2178 PH/61 Add -l and -k options to exicyclog.
2180 PH/62 When verifying, if an address was redirected to one new address, so that
2181 verification continued, and the new address failed or deferred after
2182 having set something in $address_data, the value of $address_data was not
2183 passed back to the ACL. This was different to the case when no
2184 redirection occurred. The value is now passed back in both cases.
2186 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2187 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2188 use login_cap.h, so on its own it isn't the distinguishing feature. The
2189 new name refers directly to the setclassresources() function.
2191 PH/65 Added configuration files for NetBSD3.
2193 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2195 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2196 is preferred over IPv4.
2198 PH/68 The bounce_return_message and bounce_return_body options were not being
2199 honoured for bounces generated during the reception of non-SMTP messages.
2200 In particular, this applied to messages rejected by the ACL. This bug has
2201 been fixed. However, if bounce_return_message is true and bounce_return_
2202 body is false, the headers that are returned for a non-SMTP message
2203 include only those that have been read before the error was detected.
2204 (In the case of an ACL rejection, they have all been read.)
2206 PH/69 The HTML version of the specification is now built in a directory called
2207 spec_html instead of spec.html, because the latter looks like a path with
2208 a MIME-type, and this confuses some software.
2210 PH/70 Catch two compiler warnings in sieve.c.
2212 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2213 function verify_get_ident() calls ip_connect() to connect a socket, but
2214 if the "connect()" function timed out, ip_connect() used to close the
2215 socket. However, verify_get_ident() also closes the socket later, and in
2216 between Exim writes to the log, which may get opened at this point. When
2217 the socket was closed in ip_connect(), the log could get the same file
2218 descriptor number as the socket. This naturally causes chaos. The fix is
2219 not to close the socket in ip_connect(); the socket should be closed by
2220 the function that creates it. There was only one place in the code where
2221 this was missing, in the iplookup router, which I don't think anybody now
2222 uses, but I've fixed it anyway.
2224 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2225 well as to direct DNS lookups. Otherwise the handling of names in host
2226 lists is inconsistent and therefore confusing.
2232 PH/01 Two changes to the default runtime configuration:
2234 (1) Move the checks for relay_from_hosts and authenticated clients from
2235 after to before the (commented out) DNS black list checks.
2237 (2) Add control=submission to the relay_from_hosts and authenticated
2238 clients checks, on the grounds that messages accepted by these
2239 statements are most likely to be submissions.
2241 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2243 (1) Generate an error if the third argument for the ${prvs expansion is
2246 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2249 (3) Reset the variables that are obtained from the first argument of
2250 ${prvscheck and used in the second argument before leaving the code,
2251 because their memory is reclaimed, so using them afterwards may do
2254 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2255 one (it's much easier than Tom thought :-).
2257 (5) Because of (4), we can now allow for the use of $prvscheck_result
2258 inside the third argument.
2260 PH/03 For some reason, the default setting of PATH when running a command from
2261 a pipe transport was just "/usr/bin". I have changed it to
2264 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2265 anything to be listed in the output from -bV.
2267 PH/05 When a filter generated an autoreply, the entire To: header line was
2268 quoted in the delivery log line, like this:
2270 => >A.N.Other <ano@some.domain> <original@ddress> ...
2272 This has been changed so that it extracts the operative address. There
2273 may be more than one such address. If so, they are comma-separated, like
2276 => >ano@some.domain,ona@other.domain <original@ddress> ...
2278 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2279 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2280 not being looked up in the rDNS to get a host name, Exim was showing the
2281 IP address twice in Received: lines, even though the IP addresses were
2282 identical. For example:
2284 Received: from [1.2.3.4] (helo=[1.2.3.4])
2286 However, if the real host name was known, it was omitting the HELO data
2287 if it matched the actual IP address. This has been tidied up so that it
2288 doesn't show the same IP address twice.
2290 PH/07 When both +timestamp and +memory debugging was on, the value given by
2291 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2292 called by the debug printing, thereby overwriting the timestamp buffer.
2293 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2295 PH/08 When the original message was included in an autoreply transport, it
2296 always said "this is a copy of the message, including all the headers",
2297 even if body_only or headers_only was set. It now gives an appropriate
2300 PH/09 Applied a patch from the Sieve maintainer which:
2302 o fixes some comments
2303 o adds the (disabled) notify extension core
2304 o adds some debug output for the result of if/elsif tests
2305 o points to the current vacation draft in the documentation
2306 and documents the missing references header update
2310 o fixes a bug in processing the envelope test (when testing
2311 multiple envelope elements, the last element determinted the
2314 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2315 Electronic Mail") by including:
2317 Auto-submitted: auto-generated
2319 in the messages that it generates (bounce messages and others, such as
2320 warnings). In the case of bounce messages for non-SMTP mesages, there was
2321 also a typo: it was using "Auto_submitted" (underscore instead of
2322 hyphen). Since every message generated by Exim is necessarily in response
2323 to another message, thes have all been changed to:
2325 Auto-Submitted: auto-replied
2327 in accordance with these statements in the RFC:
2329 The auto-replied keyword:
2331 - SHOULD be used on messages sent in direct response to another
2332 message by an automatic process,
2334 - MUST NOT be used on manually-generated messages,
2336 - MAY be used on Delivery Status Notifications (DSNs) and Message
2337 Disposition Notifications (MDNs),
2339 - MUST NOT be used on messages generated by automatic or periodic
2340 processes, except for messages which are automatic responses to
2343 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2344 to the default Received: header definition.
2346 PH/12 Added log selector acl_warn_skipped (default on).
2348 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2349 variables because (a) they are in the wrong storage pool and (b) even if
2350 they were copied, it wouldn't work properly because of the caching.
2352 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2353 checking when decoding. Apparently there are clients that generate
2354 overlong encoded strings. Why am I not surprised?
2356 PH/15 If the first argument of "${if match_address" was not empty, but did not
2357 contain an "@" character, Exim crashed. Now it writes a panic log message
2358 and treats the condition as false.
2360 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2362 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2363 extension "envelope-auth". The code is finished and in agreement with
2364 other implementations, but there is no documentation so far and in fact,
2365 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2366 not changing the active code.
2368 Print executed "if" and "elsif" statements when debugging is used. This
2369 helps a great deal to understand what a filter does.
2371 Document more things not specified clearly in RFC3028. I had all this
2372 sorted out, when out of a sudden new issues came to my mind. Oops."
2374 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2377 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2378 canonical form (as documented). However, after a host name lookup from
2379 the IP address, check_host() was doing a simple string comparison with
2380 addresses acquired from the DNS when checking that the found name did
2381 have the original IP as one of its addresses. Since any found IPv6
2382 addresses are likely to be in abbreviated form, the comparison could
2383 fail. Luckily, there already exists a function for doing the comparison
2384 by converting both addresses to binary, so now that is used instead of
2385 the text comparison.
2387 PH/20 There was another similar case to PH/19, when a complete host name was
2388 given in a host list; looking up its IP address could give an abbreviated
2389 form, whereas the current host's name might or might not be abbreviated.
2390 The same fix has been applied.
2396 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2397 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2400 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2401 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2403 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2405 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2406 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2407 Also, for "Auto-Submitted", treat anything other than "no" as
2408 non-personal, in accordance with RFC 3834. (Previously it treated
2409 anything starting "auto-" as non-personal.)
2411 TF/01 The control=submission/name=... option had a problem with syntax
2412 errors if the name included a slash character. The /name= option
2413 now slurps the rest of the string, so it can include any characters
2414 but it must come last in the list of options (after /sender_retain
2417 PH/05 Some modifications to the interface to the fake nameserver for the new
2425 TK/01 Added the "success_on_redirect" address verification option. See
2426 NewStuff for rationale and an example.
2428 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2430 PH/02 Patch to exigrep to allow it to work on syslog lines.
2432 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2433 fread() to skip over the body file's header line, because in Cygwin the
2434 header line is locked and is inaccessible.
2436 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2437 co-exist for some time) to make it clear that it is the Exim ID that is
2438 referenced, not the Message-ID: header line.
2440 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2441 string_format() function, because snprintf() does not exist on all
2444 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2445 ${stat: expansion item.
2447 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2448 protocol synchronization error", to keep the pedants happy.
2450 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2451 well as for IRIX systems, when gcc is being used. See the host.c source
2454 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2456 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2459 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2460 transport and to the smtp transport in LMTP mode.
2462 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2464 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2465 run a filter in a subprocess. This could lead to confusion in subsequent
2466 lookups in the parent process. There should also be a search_tidyup() at
2467 the end of the subprocess.
2469 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2470 only if the host matched helo_try_verify_hosts, which caused the
2471 verification to occur when the EHLO/HELO command was issued. The ACL just
2472 tested the remembered result. Now, if a previous verification attempt has
2473 not happened, "verify = helo" does it there and then.
2475 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2477 TK/03 Fix log output including CR from clamd.
2479 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2480 exist provoked a memory error which could cause a segfault.
2482 PH/15 Installed PCRE 6.2
2484 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2486 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2487 of the problem. Specifically, suggested +O2 rather than +O1 for the
2490 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2492 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2493 record, the port was not correctly logged when the outgoing_port log
2494 selector was set (it logged the transort's default port).
2496 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2497 fallback_hosts, and "hosts" in the smtp transport.
2499 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2500 host errors such as "Connection refused".
2502 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2503 authentication with radiusclient 0.4.9:
2505 - Error returned from rc_read_config was caught wrongly
2506 - Username/password not passed on to radius server due to wrong length.
2508 The presumption is that some radiusclient API changes for 4.51/PH/17
2509 were not taken care of correctly. The code is still untested by me (my
2510 Linux distribution still has 0.3.2 of radiusclient), but it was
2511 contributed by a Radius user.
2513 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2514 expanding the "port" option of the smtp transport.
2516 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2517 while reading a MIME header. Thanks to Tom Hughes for a patch.
2519 PH/24 Include config.h inside local_scan.h so that configuration settings are
2522 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2523 that in an ACL for RCPT (for example), you can examine exactly what was
2526 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2527 commands, but it was not correctly comparing the address with the actual
2528 client host address. Thus, it would show the EHLO address in Received:
2529 header lines when this was not necessary.
2531 PH/27 Added the % operator to ${eval:}.
2533 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2534 it should be ignoring failures (because with -C, for example, it has lost
2535 privilege). It wasn't ignoring creation failures other than "already
2538 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2541 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2542 because an input line was too long, either on its own, or by virtue of
2543 too many continuations, the temporary file was not being removed, and the
2544 return code was incorrect.
2546 PH/31 Missing "BOOL" in function definition in filtertest.c.
2548 PH/32 Applied Sieve patches from the maintainer.
2550 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2552 PH/33 Added "verify = not_blind".
2554 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2555 Local/Makefile (with some defaults set). These are used in built scripts
2556 such as exicyclog, but they have never been used in the exim_install
2557 script (though there are many overriding facilities there). I have
2558 arranged that the exim_install script now takes note of these two
2561 PH/35 Installed configuration files for Dragonfly.
2563 PH/36 When a locally submitted message by a trusted user did not contain a
2564 From: header, and the sender address was obtained from -f or from an SMTP
2565 MAIL command, and the trusted user did not use -F to supply a sender
2566 name, $originator_name was incorrectly used when constructing a From:
2567 header. Furthermore, $originator_name was used for submission mode
2568 messages from external hosts without From: headers in a similar way,
2569 which is clearly wrong.
2571 PH/37 Added control=suppress_local_fixups.
2573 PH/38 When log_selector = +received_sender was set, and the addition of the
2574 sender made the log line's construction buffer exactly full, or one byte
2575 less than full, an overflow happened when the terminating "\n" was
2578 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2579 when the result of a list match is failure because a DNS lookup failed.
2581 PH/40 RM_COMMAND is now used in the building process.
2583 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2584 the "build-* directories that it finds.
2586 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2587 address in a domain literal was a prefix of an interface address.
2589 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2590 when verifying a sender address, unless rewrite_headers is false.
2592 PH/44 Wrote a long comment about why errors_to addresses are verified as
2593 recipients, not senders.
2595 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2596 the ratelimit ACL was added.
2598 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2600 PH/46 Added extra information about PostgreSQL errors to the error string.
2602 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2603 avoiding the need to install special zones in a real server. This is
2604 backwards compatible; if it can't find the fake resolver, it drops back.
2605 Thus, both old and new test suites can be run.
2607 TF/02 Added util/ratelimit.pl
2609 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2610 clock is set back in time.
2612 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2613 Candler <B.Candler@pobox.com>.
2615 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2616 OK for addresses that are the result of redirections.
2618 PH/48 A number of further additions for the benefit of the new test suite,
2619 including a fake gethostbyname() that interfaces to the fake DNS resolver
2622 TF/06 The fix for widen_domains has also been applied to qualify_single and
2623 search_parents which are the other dnslookup options that can cause
2626 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2629 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2631 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2632 service" if it encountered an MX record with an empty target hostname.
2633 The message is now "an MX or SRV record indicated no SMTP service".
2635 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2636 if the DNS of the sending site is misconfigured. This is quite a
2637 common situation. This change restores the behaviour of treating a
2638 helo verification defer as a failure.
2640 PH/51 If self=fail was set on a router, the bounce message did not include the
2641 actual error message.
2647 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2649 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2650 command itself ended in error, the underlying message about the transport
2651 filter timeout was being overwritten with the pipe command error. Now the
2652 underlying error message should be appended to the second error message.
2654 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2656 PH/02 Reduce the amount of output that "make" produces by default. Full output
2657 can still be requested.
2659 PH/03 The warning log line about a condition test deferring for a "warn" verb
2660 was being output only once per connection, rather than after each
2661 occurrence (because it was using the same function as for successful
2662 "warn" verbs). This seems wrong, so I have changed it.
2664 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2665 it should not have, which might have caused a crash in the right
2666 circumstances, but probably never did.
2668 PH/04 Installed a modified version of Tony Finch's patch to make submission
2669 mode fix the return path as well as the Sender: header line, and to
2670 add a /name= option so that you can make the user's friendly name appear
2673 TF/03 Added the control = fakedefer ACL modifier.
2675 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2676 Mark Lowes for thorough testing.
2678 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2680 TK/03 Merged latest SRS patch from Miles Wilton.
2682 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2683 with the definition in sysexits.h (which is #included earlier).
2684 Fortunately, Exim does not actually use EX_OK. The code used to try to
2685 preserve the sysexits.h value, by assumimg that macro definitions were
2686 scanned for macro replacements. I have been disabused of this notion,
2687 so now the code just undefines EX_OK before #including unistd.h.
2689 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2690 in the smtp transport. When a block could not be written in a single
2691 write() function, the timeout was being re-applied to each part-write.
2692 This seems wrong - if the receiver was accepting one byte at a time it
2693 would take for ever. The timeout is now adjusted when this happens. It
2694 doesn't have to be particularly precise.
2696 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2697 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2699 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2700 without a domain if the check to <postmaster@domain> fails.
2702 SC/01 Eximstats: added -xls and the ability to specify output files
2703 (patch written by Frank Heydlauf).
2705 SC/02 Eximstats: use FileHandles for outputing results.
2707 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2709 SC/04 Eximstats: fixed display of large numbers with -nvr option
2711 SC/05 Eximstats: fixed merging of reports with empty tables.
2713 SC/06 Eximstats: added the -include_original_destination flag
2715 SC/07 Eximstats: removed tabs and trailing whitespace.
2717 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2719 TK/06 MBOX spool code: Add real "From " MBOX separator line
2720 so the .eml file is really in mbox format (even though
2721 most programs do not really care). Patch from Alex Miller.
2723 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2724 The latter is generated from $received_to and is only set if the
2725 message has one envelope recipient. SA can use these headers,
2726 obviously out-of-the-box. Patch from Alex Miller.
2728 PH/08 The ${def test on a variable was returning false if the variable's
2729 value was "0", contrary to what the specification has always said!
2730 The result should be true unless the variable is empty.
2732 PH/09 The syntax error of a character other than { following "${if
2733 def:variable_name" (after optional whitespace) was not being diagnosed.
2734 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2735 accidental colon was present, for example, could give incorrect results.
2737 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2738 result is used (not including appendfile, where other changes are about
2741 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2742 This involved changing a lot of size variables from int to off_t. It
2743 should work with maildirs and everything.
2745 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2746 spamd dying while we are connected to it.
2748 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2751 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2752 function for BDB 4.3.
2754 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2756 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2757 the sources for PCRE have been re-arranged and more files are now
2760 PH/15 The code I had for printing potentially long long variables in PH/11
2761 above was not the best (it lost precision). The length of off_t variables
2762 is now inspected at build time, and an appropriate printing format (%ld
2763 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2764 to be "long long int" or "long int". This is needed for the internal
2765 formatting function string_vformat().
2767 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2768 the configuration file to be ":syslog", then the script "guesses" where
2769 the logs files are, rather than using the compiled in default. In our
2770 case the guess is not the same as the compiled default, so the script
2771 suddenly stopped working when I started to use syslog. The patch checks
2772 to see if log_file_path is "". If so, it attempts to read it from exim
2773 with no configuration file to get the compiled in version, before it
2774 falls back to the previous guessing code."
2776 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2777 implementing BATV in an Exim configuration. See NewStuff for the gory
2780 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2781 Makefile that are specific to HP-UX.
2783 PH/18 If the "use_postmaster" option was set for a recipient callout together
2784 with the "random" option, the postmaster address was used as the MAIL
2785 FROM address for the random test, but not for the subsequent recipient
2786 test. It is now used for both.
2788 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2789 patch removes a few documentation additions to RFC 3028, because the
2790 latest draft now contains them. It adds the new en;ascii-case comparator
2791 and a new error check for 8bit text in MIME parts. Comparator and
2792 require names are now matched exactly. I enabled the subaddress
2793 extension, but it is not well tested yet (read: it works for me)."
2795 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2796 rework some of the code of TK/09 above to avoid the hardwired use of
2797 "%lld" and "long long". Replaced the call to snprintf() with a call to
2800 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2801 records point to non-existent hosts", "retry timeout exceeded", and
2802 "retry time not reached for any host after a long failure period".
2804 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2805 experimental DomainKeys support:
2807 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2808 (2) On an error such as an illegally used "control", the wrong name for
2809 the control was given.
2811 These problems did NOT occur unless DomainKeys support was compiled.
2813 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2815 PH/24 Added ${if match_ip condition.
2817 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2818 databases so that it will be absolutely obvious if a crash occurs in the
2819 DB library. This is a regular occurrence (often caused by mis-matched
2822 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2823 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2824 picked up on a user's system that detects such things. There doesn't seem
2825 to be a gcc warning option for this - only an attribute that has to be
2826 put on the function's prototype. It seems that in Fedora Core 4 they have
2827 set this on a number of new functions. No doubt there will be more in due
2830 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2831 specify a transport. However, if an address that was verified by such a
2832 router was the subject of a callout, Exim crashed because it tried to
2833 read the rcpt_include_affixes from the non-existent transport. Now it
2834 just assumes that the setting of that option is false. This bug was
2835 introduced by 4.51/PH/31.
2837 PH/28 Changed -d+all to exclude +memory, because that information is very
2838 rarely of interest, but it makes the output a lot bigger. People tend to
2839 do -d+all out of habit.
2841 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2842 code in os-type was giving problems when libc.so lives in lib64, like on
2845 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2846 aren't the modern standard, and it seems that some systems' include files
2847 don't always have them. Exim was already checking for some of the newer
2848 ones like T_AAAA, and defining it itself. I've added checks for all the
2849 record types that Exim uses.
2851 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2852 not automatically generating a new one, as it is supposed to. This
2853 prevented TLS from working. If the file did exist, but contained invalid
2854 data, a new version was generated, as expected. It was only the case of a
2855 non-existent file that was broken.
2857 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2858 with a change in libdomainkeys > 0.64.
2860 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2861 from DNS. If the selector record carries the flag, it now has
2862 precedence over the domain-wide flag.
2864 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2866 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2867 the use of a port name that isn't defined in /etc/services) occurred, the
2868 message was deferred as in a normal delivery, and thus remained on the
2869 spool, instead of being failed because of the mua_wrapper setting. This
2870 is now fixed, and I tidied up some of the mua_wrapper messages at the
2873 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2874 the messages in a hash of arrays rather than using individual hashes.
2875 This is a bit cleaner and results in dramatic memory savings, albeit
2876 at a slight CPU cost.
2878 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2879 as requested by Marc Sherman.
2881 SC/10 Eximstats: added histograms for user specified patterns as requested
2884 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2886 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2887 fopen() in the content-scanning modules that did not already have it.
2893 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2894 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2896 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2898 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2900 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2901 to test Sieve filters that use "vacation".
2903 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2904 that changes the way the GnuTLS parameters are stored in the cache file.
2905 The new format can be generated externally. For backward compatibility,
2906 if the data in the cache doesn't make sense, Exim assumes it has read an
2907 old-format file, and it generates new data and writes a new file. This
2908 means that you can't go back to an older release without removing the
2911 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2912 work if there are any delivery delays because "one_time" forces the
2913 parent to be marked "delivered", so its unseen clone is never tried
2914 again. For this reason, Exim now forbids the simultaneous setting of
2917 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2918 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2919 entry if you want to know the details. The fix, however, neglected to
2920 consider the case when local delivery batching is involved. The test for
2921 "previously delivered" was not happening when checking to see if an
2922 address could be batched with a previous (undelivered) one; under
2923 certain circumstances this could lead to multiple deliveries to the same
2926 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2927 in its include files, and this causes problems building Exim.
2929 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2930 header_syntax) but Exim was just ignoring anything given after a slash.
2931 In particular, this caused confusion with an attempt to use "verify =
2932 reverse_host_lookup/defer_ok". An error is now given when options are
2933 supplied for verify items that do not have them. (Maybe reverse_host_
2934 lookup should have a defer_ok option, but that's a different point.)
2936 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2937 defined by RFC 821) to 2048, because there were problems with some AUTH
2938 commands, and RFC 1869 says the size should be increased for extended
2939 SMTP commands that take arguments.
2941 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2944 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2945 "unknown" error; now it says that the functionality isn't in the binary.
2947 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2948 an address' error message when a string expansion fails (syntax or
2949 whatever). Otherwise the password may appear in the log. Following change
2950 PH/42 below, there is no longer a chance of it appearing in a bounce
2953 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2955 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2956 of its addresses was ever tried. (Bugzilla bug #2.)
2958 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2959 the result incorrectly in the debug output. (It correctly added a newline
2960 to what was transported.)
2962 TF/01 Added $received_time.
2964 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2965 commented out examples of how to interface to a virus scanner and to
2966 SpamAssassin. Also added commented examples of av_scanner and
2967 spamd_address settings.
2969 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2970 and controls are allowed in which ACLs. There were a couple of minor
2971 errors. Some of the entries in the conditions table (which is a table of
2972 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2973 negation of where the condition IS allowed.
2975 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2977 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2978 header file does not have a version number, so I've had to invent a new
2979 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2980 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2981 radiusclient), but it was contributed by a Radius user.
2983 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2984 files or directories, for OpenSSL.
2986 PH/19 When an Exim process that is running as root has to create an Exim log
2987 file, it does so in a subprocess that runs as exim:exim so as to get the
2988 ownership right at creation (otherwise, other Exim processes might see
2989 the file with the wrong ownership). There was no test for failure of this
2990 fork() call, which would lead to the process getting stuck as it waited
2991 for a non-existent subprocess. Forks do occasionally fail when resources
2992 run out. I reviewed all the other calls to fork(); they all seem to check
2995 PH/20 When checking for unexpected SMTP input at connect time (before writing
2996 the banner), Exim was not dealing correctly with a non-positive return
2997 from the read() function. If the client had disconnected by this time,
2998 the result was a log entry for a synchronization error with an empty
2999 string after "input=" when read() returned zero. If read() returned -1
3000 (an event I could not check), uninitialized data bytes were printed.
3001 There were reports of junk text (parts of files, etc) appearing after
3004 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3006 PH/22 Added support for macro redefinition, and (re)definition in between
3007 driver and ACL definitions.
3009 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3010 forgetting to use the resulting value; it was using the unexpanded value.
3012 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3013 hadn't been configured. The fix is from Juergen Kreileder, who
3014 understands it better than I do:
3016 "Here's what I see happening with three configured cyrus_sasl
3017 authenticators configured (plain, login, cram-md5):
3019 On startup auth_cyrus_sasl_init() gets called for each of these.
3020 This means three calls to sasl_listmech() without a specified mech_list.
3021 => SASL tests which mechs of all available mechs actually work
3022 => three warnings about OTP not working
3023 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3025 With the patch, sasl_listmech() also gets called three times. But now
3026 SASL's mech_list option is set to the server_mech specified in the the
3027 authenticator. Or in other words, the answer from sasl_listmech()
3028 gets limited to just the mech you're testing for (which is different
3030 => the return list contains just 'plain' or 'login', 'cram-md5' or
3031 nothing depending on the value of ob->server_mech.
3033 I've just tested the patch: Authentication still works fine,
3034 unavailable mechs specified in the exim configuration are still
3035 caught, and the auth.log warnings about OTP are gone."
3037 PH/25 When debugging is enabled, the contents of the command line are added
3038 to the debugging output, even when log_selector=+arguments is not
3041 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3042 answer is "GNU", and only if the return is "GNU/something" is the answer
3045 PH/27 $acl_verify_message is now set immediately after the failure of a
3046 verification in an ACL, and so is available in subsequent modifiers. In
3047 particular, the message can be preserved by coding like this:
3049 warn !verify = sender
3050 set acl_m0 = $acl_verify_message
3052 Previously, $acl_verify_message was set only while expanding "message"
3053 and "log_message" when a very denied access.
3055 PH/28 Modified OS/os.c-Linux with
3057 -#ifndef OS_LOAD_AVERAGE
3058 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3060 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3061 nomenclature these days.)
3063 PH/29 Installed patch from the Sieve maintainer that adds the options
3064 sieve_useraddress and sieve_subaddress to the redirect router.
3066 PH/30 In these circumstances:
3067 . Two addresses routed to the same list of hosts;
3068 . First host does not offer TLS;
3069 . First host accepts first address;
3070 . First host gives temporary error to second address;
3071 . Second host offers TLS and a TLS session is established;
3072 . Second host accepts second address.
3073 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3074 and peerdn, if requested) that were in fact used only for the second
3077 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3078 attention to any local part prefix or suffix that was matched by the
3079 router that accepted the address. It now behaves in the same way as it
3080 does for delivery: the affixes are removed from the local part unless
3081 rcpt_include_affixes is set on the transport.
3083 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3084 timeout during the DATA phase of an incoming message.
3086 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3087 applied a patch sanctioned by the Sieve maintainer.
3089 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3090 the uid or gid is negative. A case of a negative gid caused this to be
3091 noticed. The fix allows for either to be negative.
3093 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3094 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3097 PH/36 The change PH/12 above was broken. Fixed it.
3099 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3100 the grounds that routing the same address twice would always produce the
3101 same answer. This might have been true once, but it is certainly no
3102 longer true now. Routing a child address may depend on the previous
3103 routing that produced that child. Some complicated redirection strategies
3104 went wrong when messages had multiple recipients, and made Exim's
3105 behaviour dependent on the order in which the addresses were given.
3107 I have moved the duplicate checking until after the routing is complete.
3108 Exim scans the addresses that are assigned to local and remote
3109 transports, and removes any duplicates. This means that more work will be
3110 done, as duplicates will always all be routed, but duplicates are
3111 presumably rare, so I don't expect this is of any significance.
3113 For deliveries to pipes, files, and autoreplies, the duplicate checking
3114 still happens during the routing process, since they are not going to be
3117 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3118 It corrects a timeout issue with spamd. This is Ian's comment: "The
3119 background is that sometimes spamd either never reads data from a
3120 connection it has accepted, or it never writes response data. The exiscan
3121 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3122 blindly assumes that writes won't block so it may never time out."
3124 PH/39 Allow G after quota size as well as K and M.
3126 PH/40 The value set for $authenticated_id in an authenticator may not contain
3127 binary zeroes or newlines because the value is written to log lines and
3128 to spool files. There was no check on this. Now the value is run through
3129 the string_printing() function so that such characters are converted to
3130 printable escape sequences.
3132 PH/41 $message_linecount is a new variable that contains the total number of
3133 lines in the message. Compare $body_linecount, which is the count for the
3136 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3137 bounce and delay warning messages, except in certain special cases, which
3140 (a) An SMTP error message from a remote host;
3141 (b) A message specified in a :fail: redirection;
3142 (c) A message specified in a "fail" command in a system filter;
3143 (d) A message specified in a FAIL return from the queryprogram router;
3144 (e) A message specified by the cannot_route_message router option.
3146 In these cases only, Exim does include the error details in bounce and
3147 warning messages. There are also a few cases where bland messages such
3148 as "unrouteable address" or "local delivery error" are given.
3150 PH/43 $value is now also set for the "else" part of a ${run expansion.
3152 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3153 being worked on, but at least Exim now implements the latest version to
3156 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3157 process to complete was treated as a delivery failure, a timeout while
3158 writing the message to the pipe was logged, but erroneously treated as a
3159 successful delivery. Such timeouts include transport filter timeouts. For
3160 consistency with the overall process timeout, these timeouts are now
3161 treated as errors, giving rise to delivery failures by default. However,
3162 there is now a new Boolean option for the pipe transport called
3163 timeout_defer, which, if set TRUE, converts the failures into defers for
3164 both kinds of timeout. A transport filter timeout is now identified in
3167 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3168 systems where "make" and "gmake" are different, calling "gmake" at top
3169 level broke things. I've arranged for the value of $(MAKE) to be passed
3170 from the Makefile to this script so that it can call the same version of
3174 A note about Exim versions 4.44 and 4.50
3175 ----------------------------------------
3177 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3178 changes of various kinds. As a consequence, a big documentation update was
3179 needed. This delayed the release for rather longer than seemed good, especially
3180 in the light of a couple of (minor) security issues. Therefore, the changes
3181 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3182 release. So 4.44 and 4.50 are in effect two different branches that both start
3185 I have left the 4.50 change log unchanged; it contains all the changes since
3186 4.43. The change log for 4.44 is below; many of its items are identical to
3187 those for 4.50. This seems to be the most sensible way to preserve the
3188 historical information.
3194 1. Minor wording change to the doc/README.SIEVE file.
3196 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3197 computation of the current number of files was incorrect.
3199 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3200 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3203 4. Give more explanation in the error message when the command for a transport
3204 filter fails to execute.
3206 5. There are several places where Exim runs a non-Exim command in a
3207 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3208 was being done only for the command run by the queryprogram router. It is
3209 now done for all such subprocesses. The other cases are: ${run, transport
3210 filters, and the commands run by the lmtp and pipe transports.
3212 6. Added CONFIGURE_GROUP build-time option.
3214 7. Some older OS have a limit of 256 on the maximum number of file
3215 descriptors. Exim was using setrlimit() to set 1000 as a large value
3216 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3217 systems. I've change it so that if it can't get 1000, it tries for 256.
3219 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3220 was an oversight, and furthermore, ever since the addition of extra
3221 controls (e.g. 4.43/32), the checks on when to allow different forms of
3222 "control" were broken. There should now be diagnostics for all cases when a
3223 control that does not make sense is encountered.
3225 9. Added the /retain_sender option to "control=submission".
3227 10. $recipients is now available in the predata ACL (oversight).
3229 11. Tidy the search cache before the fork to do a delivery from a message
3230 received from the command line. Otherwise the child will trigger a lookup
3231 failure and thereby defer the delivery if it tries to use (for example) a
3232 cached ldap connection that the parent has called unbind on.
3234 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3235 of $address_data from the recipient verification was clobbered by the
3236 sender verification.
3238 13. The value of address_data from a sender verification is now available in
3239 $sender_address_data in subsequent conditions in the ACL statement.
3241 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3243 15. Added a new option "connect=<time>" to callout options, to set a different
3246 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3247 was its contents. (It was OK if the option was not defined at all.)
3249 17. A "Completed" log line is now written for messages that are removed from
3250 the spool by the -Mrm option.
3252 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3253 information about exactly what failed.
3255 19. Added -dd to debug only the daemon process.
3257 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3258 handles timeouts, both on the server side and network timeouts. Renamed the
3259 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3261 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3262 It is now set to "smtps".
3264 22. $host_address is now set to the target address during the checking of
3265 ignore_target_hosts.
3267 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3268 being passed; this would have caused $sender_host_name to have been used if
3269 matching the list had actually called for a host name (not very likely,
3270 since this list is usually IP addresses). A host name is now passed as
3273 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3274 code that specifies a non-restarting handler (typically sigaction() in
3275 modern systems) in an attempt to fix a rare and obscure crash bug.
3277 25. Narrowed the window for a race in the daemon that could cause it to ignore
3278 SIGCHLD signals. This is not a major problem, because they are used only to
3279 wake it up if nothing else does.
3281 26. A malformed maildirsize file could cause Exim to calculate negative values
3282 for the mailbox size or file count. Odd effects could occur as a result.
3283 The maildirsize information is now recalculated if the size or filecount
3286 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3287 support for a long time. Removed HAVE_SYS_VFS_H.
3289 28. Installed the latest version of exipick from John Jetmore.
3291 29. In an address list, if the pattern was not a regular expression, an empty
3292 subject address (from a bounce message) matched only if the pattern was an
3293 empty string. Non-empty patterns were not even tested. This was the wrong
3294 because it is perfectly reasonable to use an empty address as part of a
3295 database query. An empty address is now tested by patterns that are
3296 lookups. However, all the other forms of pattern expect the subject to
3297 contain a local part and a domain, and therefore, for them, an empty
3298 address still always fails if the pattern is not itself empty.
3300 30. Exim went into a mad DNS loop when attempting to do a callout where the
3301 host was specified on an smtp transport, and looking it up yielded more
3302 than one IP address.
3304 31. Re-factored the code for checking spool and log partition space into a
3305 function that finds that data and another that does the check. The former
3306 is then used to implement four new variables: $spool_space, $log_space,
3307 $spool_inodes, and $log_inodes.
3309 32. The RFC2047 encoding function was originally intended for short strings
3310 such as real names; it was not keeping to the 75-character limit for
3311 encoded words that the RFC imposes. It now respects the limit, and
3312 generates multiple encoded words if necessary. To be on the safe side, I
3313 have increased the buffer size for the ${rfc2047: expansion operator from
3316 33. It is now permitted to omit both strings after an "if" condition; if the
3317 condition is true, the result is "true". As before, when the second string
3318 is omitted, a false condition yields an empty string. This makes it less
3319 cumbersome to write custom ACL and router conditions.
3321 34. Failure to deliver a bounce message always caused it to be frozen, even if
3322 there was an errors_to setting on the router. The errors_to setting is now
3325 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3326 canonical form (fully expanded) before being placed in
3327 $sender_host_address.
3329 36. The table in the code that translates DNS record types into text (T_A to
3330 "A" for instance) was missing entries for NS and CNAME. It is just possible
3331 that this could have caused confusion if both these types were looked up
3332 for the same domain, because the text type is used as part of Exim's
3333 per-process caching. But the chance of anyone hitting this buglet seems
3336 37. The dnsdb lookup has been extended in a number of ways.
3338 (1) There is a new type, "zns", which walks up the domain tree until it
3339 finds some nameserver records. It should be used with care.
3341 (2) There is a new type, "mxh", which is like "mx" except that it returns
3342 just the host names, not the priorities.
3344 (3) It is now possible to give a list of domains (or IP addresses) to be
3345 looked up. The behaviour when one of the lookups defers can be
3346 controlled by a keyword.
3348 (4) It is now possible to specify the separator character for use when
3349 multiple records are returned.
3351 38. The dnslists ACL condition has been extended: it is now possible to supply
3352 a list of IP addresses and/or domains to be looked up in a particular DNS
3355 39. Added log_selector=+queue_time_overall.
3357 40. When running the queue in the test harness, wait just a tad after forking a
3358 delivery process, to get repeatability of debugging output.
3360 41. Include certificate and key file names in error message when GnuTLS fails
3361 to set them up, because the GnuTLS error message doesn't include the name
3362 of the failing file when there is a problem reading it.
3364 42. Allow both -bf and -bF in the same test run.
3366 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3368 44. The "Exiscan patch" is now merged into the mainline Exim source.
3370 45. Sometimes the final signoff response after QUIT could fail to get
3371 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3372 < 0 before doing a fflush(). This bug looks as though it goes back to the
3373 introduction of TLS in release 3.20, but "sometimes" must have been rare
3374 because the tests only now provoked it.
3376 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3377 (this can affect the format of dates).
3379 47. exim_tidydb, when checking for the continued existence of a message for
3380 which it has found a message-specific retry record, was not finding
3381 messages that were in split spool directories. Consequently, it was
3382 deleting retry records that should have stayed in existence.
3384 48. Steve fixed some bugs in eximstats.
3386 49. The SPA authentication driver was not abandoning authentication and moving
3387 on to the next authenticator when an expansion was forced to fail,
3388 contradicting the general specification for all authenticators. Instead it
3389 was generating a temporary error. It now behaves as specified.
3391 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3392 (the order specifies the preference for clients). The order is now AES256,
3393 AES128, 3DES, ARCFOUR128.
3395 51. Small patch to Sieve code - explicitly set From: when generating an
3398 52. Exim crashed if a remote delivery caused a very long error message to be
3399 recorded - for instance if somebody sent an entire SpamAssassin report back
3400 as a large number of 550 error lines. This bug was coincidentally fixed by
3401 increasing the size of one of Exim's internal buffers (big_buffer) that
3402 happened as part of the Exiscan merge. However, to be on the safe side, I
3403 have made the code more robust (and fixed the comments that describe what
3406 53. Now that there can be additional text after "Completed" in log lines (if
3407 the queue_time_overall log selector is set), a one-byte patch to exigrep
3408 was needed to allow it to recognize "Completed" as not the last thing in
3411 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3412 patch that reportedly fixes this has been added. I am not expert enough to
3413 create a test for it. This is what the patch creator wrote:
3415 "I found a little strange behaviour of ldap code when working with
3416 Windows 2003 AD Domain, where users was placed in more than one
3417 Organization Units. When I tried to give exim partial DN, the exit code
3418 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3419 But simultaneously result of request was absolutely normal ldap result,
3420 so I produce this patch..."
3422 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3423 REFERENCE, so I have modified the code to exclude the patch when that macro
3426 55. Some experimental protocols are using DNS PTR records for new purposes. The
3427 keys for these records are domain names, not reversed IP addresses. The
3428 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3429 leaves it alone. Component reversal etc. now happens only for IP addresses.
3432 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3434 57. Double the size of the debug message buffer (to 2048) so that more of very
3435 long debug lines gets shown.
3437 58. The exicyclog utility now does better if the number of log files to keep
3438 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3440 59. Two changes related to the smtp_active_hostname option:
3442 (1) $smtp_active_hostname is now available as a variable.
3443 (2) The default for smtp_banner uses $smtp_active_hostname instead
3444 of $primary_hostname.
3446 60. The host_aton() function is supposed to be passed a string that is known
3447 to be a valid IP address. However, in the case of IPv6 addresses, it was
3448 not checking this. This is a hostage to fortune. Exim now panics and dies
3449 if the condition is not met. A case was found where this could be provoked
3450 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3451 components; fortuitously, this particular loophole had already been fixed
3452 by change 4.50/55 above.
3454 If there are any other similar loopholes, the new check in host_aton()
3455 itself should stop them being exploited. The report I received stated that
3456 data on the command line could provoke the exploit when Exim was running as
3457 exim, but did not say which command line option was involved. All I could
3458 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3459 running as the user.
3462 61. There was a buffer overflow vulnerability in the SPA authentication code
3463 (which came originally from the Samba project). I have added a test to the
3464 spa_base64_to_bits() function which I hope fixes it.
3467 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3468 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3470 63. The daemon start-up calls getloadavg() while still root for those OS that
3471 need the first call to be done as root, but it missed one case: when
3472 deliver_queue_load_max is set with deliver_drop_privilege. This is
3473 necessary for the benefit of the queue runner, because there is no re-exec
3474 when deliver_drop_privilege is set.
3476 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3477 This has been fixed.
3479 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3480 was in use, was not putting the data itself into the right store pool;
3481 consequently, it could be overwritten for a subsequent message in the same
3482 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3485 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3487 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3488 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3489 the function were treating the return as a boolean value, which happened to
3490 work because 0=false and not-0=true, but is not correct code.
3492 68. The host_aton() function was not handling scoped IPv6 addresses (those
3493 with, for example, "%eth0" on the end) correctly.
3495 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3496 negated items (that is, ~something) in unsigned ints. Some compilers
3497 apparently mutter when there is no cast.
3499 70. If an address verification called from an ACL failed, and did not produce a
3500 user-specific message (i.e. there was only a "system" message), nothing was
3501 put in $acl_verify_message. In this situation, it now puts the system
3504 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3505 session; change 4.31/43 added the unwanted input to the log line - except
3506 that it did not do this in the start of session case. It now does.
3508 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3509 This is wrong and can cause the other end to generate a synchronization
3510 error if it is another Exim or anything else that does the synchronization
3511 check. A QUIT command is no longer sent after a timeout.
3513 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3514 during host lookups.
3516 74. The defer_ok option of callout verification was not working if it was used
3517 when verifying addresses in header lines, that is, for this case:
3519 verify = header_sender/callout=defer_ok
3521 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3522 those file descriptors could be used for SMTP connections. If anything
3523 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3524 could be sent to the SMTP client, causing chaos. The daemon now opens
3525 stdin, stdout, and stderr to /dev/null when it puts itself into the
3528 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3529 log by default. The user can override this with suitable Perl magic.
3531 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3532 the log message when discard triggers, was not working for the DATA ACL or
3533 for the non-SMTP ACL.
3535 78. Error message wording change in sieve.c.
3537 79. If smtp_accept_max_per_host was set, the number of connections could be
3538 restricted to fewer than expected, because the daemon was trying to set up
3539 a new connection before checking whether the processes handling previous
3540 connections had finished. The check for completed processes is now done
3541 earlier. On busy systems, this bug wouldn't be noticed because something
3542 else would have woken the daemon, and it would have reaped the completed
3545 80. If a message was submitted locally by a user whose login name contained one
3546 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3547 It caused a spool format error. I have fixed the spool reading code. A
3548 related problem was that the "from" clause in the Received: line became
3549 illegal because of the space(s). It is now covered by ${quote_local_part.
3551 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3554 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3556 83. Patch from Sieve maintainer to fix unterminated string problem in
3557 "vacation" handling.
3559 84. Some minor changes to the Linux configuration files to help with other
3560 OS variants using glibc.
3562 85. One more patch for Sieve to update vacation handling to latest spec.
3565 ----------------------------------------------------
3566 See the note above about the 4.44 and 4.50 releases.
3567 ----------------------------------------------------
3573 1. Change 4.43/35 introduced a bug that caused file counts to be
3574 incorrectly computed when quota_filecount was set in an appendfile
3577 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3578 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3581 3. Give more explanation in the error message when the command for a transport
3582 filter fails to execute.
3584 4. There are several places where Exim runs a non-Exim command in a
3585 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3586 was being done only for the command run by the queryprogram router. It is
3587 now done for all such subprocesses. The other cases are: ${run, transport
3588 filters, and the commands run by the lmtp and pipe transports.
3590 5. Some older OS have a limit of 256 on the maximum number of file
3591 descriptors. Exim was using setrlimit() to set 1000 as a large value
3592 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3593 systems. I've change it so that if it can't get 1000, it tries for 256.
3595 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3596 was an oversight, and furthermore, ever since the addition of extra
3597 controls (e.g. 4.43/32), the checks on when to allow different forms of
3598 "control" were broken. There should now be diagnostics for all cases when a
3599 control that does not make sense is encountered.
3601 7. $recipients is now available in the predata ACL (oversight).
3603 8. Tidy the search cache before the fork to do a delivery from a message
3604 received from the command line. Otherwise the child will trigger a lookup
3605 failure and thereby defer the delivery if it tries to use (for example) a
3606 cached ldap connection that the parent has called unbind on.
3608 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3609 of $address_data from the recipient verification was clobbered by the
3610 sender verification.
3612 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3613 was its contents. (It was OK if the option was not defined at all.)
3615 11. A "Completed" log line is now written for messages that are removed from
3616 the spool by the -Mrm option.
3618 12. $host_address is now set to the target address during the checking of
3619 ignore_target_hosts.
3621 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3622 being passed; this would have caused $sender_host_name to have been used if
3623 matching the list had actually called for a host name (not very likely,
3624 since this list is usually IP addresses). A host name is now passed as
3627 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3628 code that specifies a non-restarting handler (typically sigaction() in
3629 modern systems) in an attempt to fix a rare and obscure crash bug.
3631 15. Narrowed the window for a race in the daemon that could cause it to ignore
3632 SIGCHLD signals. This is not a major problem, because they are used only to
3633 wake it up if nothing else does.
3635 16. A malformed maildirsize file could cause Exim to calculate negative values
3636 for the mailbox size or file count. Odd effects could occur as a result.
3637 The maildirsize information is now recalculated if the size or filecount
3640 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3641 support for a long time. Removed HAVE_SYS_VFS_H.
3643 18. Updated exipick to current release from John Jetmore.
3645 19. Allow an empty sender to be matched against a lookup in an address list.
3646 Previously the only cases considered were a regular expression, or an
3649 20. Exim went into a mad DNS lookup loop when doing a callout where the
3650 host was specified on the transport, if the DNS lookup yielded more than
3653 21. The RFC2047 encoding function was originally intended for short strings
3654 such as real names; it was not keeping to the 75-character limit for
3655 encoded words that the RFC imposes. It now respects the limit, and
3656 generates multiple encoded words if necessary. To be on the safe side, I
3657 have increased the buffer size for the ${rfc2047: expansion operator from
3660 22. Failure to deliver a bounce message always caused it to be frozen, even if
3661 there was an errors_to setting on the router. The errors_to setting is now
3664 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3665 canonical form (fully expanded) before being placed in
3666 $sender_host_address.
3668 24. Updated eximstats to version 1.33
3670 25. Include certificate and key file names in error message when GnuTLS fails
3671 to set them up, because the GnuTLS error message doesn't include the name
3672 of the failing file when there is a problem reading it.
3674 26. Expand error message when OpenSSL has problems setting up cert/key files.
3677 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3678 (this can affect the format of dates).
3680 28. exim_tidydb, when checking for the continued existence of a message for
3681 which it has found a message-specific retry record, was not finding
3682 messages that were in split spool directories. Consequently, it was
3683 deleting retry records that should have stayed in existence.
3685 29. eximstats updated to version 1.35
3686 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3687 1.35 - bugfix such that pie charts by volume are generated correctly
3689 30. The SPA authentication driver was not abandoning authentication and moving
3690 on to the next authenticator when an expansion was forced to fail,
3691 contradicting the general specification for all authenticators. Instead it
3692 was generating a temporary error. It now behaves as specified.
3694 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3695 (the order specifies the preference for clients). The order is now AES256,
3696 AES128, 3DES, ARCFOUR128.
3698 31. Small patch to Sieve code - explicitly set From: when generating an
3701 32. Exim crashed if a remote delivery caused a very long error message to be
3702 recorded - for instance if somebody sent an entire SpamAssassin report back
3703 as a large number of 550 error lines. This bug was coincidentally fixed by
3704 increasing the size of one of Exim's internal buffers (big_buffer) that
3705 happened as part of the Exiscan merge. However, to be on the safe side, I
3706 have made the code more robust (and fixed the comments that describe what
3709 33. Some experimental protocols are using DNS PTR records for new purposes. The
3710 keys for these records are domain names, not reversed IP addresses. The
3711 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3712 leaves it alone. Component reversal etc. now happens only for IP addresses.
3715 34. The host_aton() function is supposed to be passed a string that is known
3716 to be a valid IP address. However, in the case of IPv6 addresses, it was
3717 not checking this. This is a hostage to fortune. Exim now panics and dies
3718 if the condition is not met. A case was found where this could be provoked
3719 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3720 components; fortuitously, this particular loophole had already been fixed
3721 by change 4.50/55 or 4.44/33 above.
3723 If there are any other similar loopholes, the new check in host_aton()
3724 itself should stop them being exploited. The report I received stated that
3725 data on the command line could provoke the exploit when Exim was running as
3726 exim, but did not say which command line option was involved. All I could
3727 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3728 running as the user.
3731 35. There was a buffer overflow vulnerability in the SPA authentication code
3732 (which came originally from the Samba project). I have added a test to the
3733 spa_base64_to_bits() function which I hope fixes it.
3736 36. The daemon start-up calls getloadavg() while still root for those OS that
3737 need the first call to be done as root, but it missed one case: when
3738 deliver_queue_load_max is set with deliver_drop_privilege. This is
3739 necessary for the benefit of the queue runner, because there is no re-exec
3740 when deliver_drop_privilege is set.
3742 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3743 was in use, was not putting the data itself into the right store pool;
3744 consequently, it could be overwritten for a subsequent message in the same
3745 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3748 38. Sometimes the final signoff response after QUIT could fail to get
3749 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3750 < 0 before doing a fflush(). This bug looks as though it goes back to the
3751 introduction of TLS in release 3.20, but "sometimes" must have been rare
3752 because the tests only now provoked it.
3758 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3759 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3760 Now it returns an integer. A number of calls were still expecting a T/F
3761 return. Fortuitously, in all cases, the tests worked in OK situations,
3762 which is the norm. However, things would have gone wrong on any write
3763 failures on the smtp file descriptor. This function is used when sending
3764 messages over SMTP and also when doing verify callouts.
3766 2. When Exim is called to do synchronous delivery of a locally submitted
3767 message (the -odf or -odi options), it no longer closes stderr before doing
3770 3. Implemented the mua_wrapper option.
3772 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3774 5. Implemented the functions header_remove(), header_testname(),
3775 header_add_at_position(), and receive_remove_recipient(), and exported them
3778 6. If an ACL "warn" statement specified the addition of headers, Exim already
3779 inserted X-ACL-Warn: at the start if there was no header name. However, it
3780 was not making this test for the second and subsequent header lines if
3781 there were newlines in the string. This meant that an invalid header could
3782 be inserted if Exim was badly configured.
3784 7. Allow an ACL "warn" statement to add header lines at the start or after all
3785 the Received: headers, as well as at the end.
3787 8. Added the rcpt_4xx retry error code.
3789 9. Added postmaster_mailfrom=xxx to callout verification option.
3791 10. Added mailfrom=xxxx to the callout verification option, for verify=
3794 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3795 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3797 12. Inserted some casts to stop certain compilers warning when using pointer
3798 differences as field lengths or precisions in printf-type calls (mostly
3799 affecting debugging statements).
3801 13. Added optional readline() support for -be (dynamically loaded).
3803 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3804 same clock tick as a message's arrival, so that its received time was the
3805 same as the "first fail" time on the retry record, and that message
3806 remained on the queue past the ultimate address timeout, every queue runner
3807 would try a delivery (because it was past the ultimate address timeout) but
3808 after another failure, the ultimate address timeout, which should have then
3809 bounced the address, did not kick in. This was a "< instead of <=" error;
3810 in most cases the first failure would have been in the next clock tick
3811 after the received time, and all would be well.
3813 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3814 being recognized when the domain list was tested by the match_domain
3815 condition in an expansion string.
3817 16. Added the ${str2b64: operator.
3819 17. Exim was always calling setrlimit() to set a large limit for the number of
3820 processes, without checking whether the existing limit was already
3821 adequate. (It did check for the limit on file descriptors.) Furthermore,
3822 errors from getrlimit() and setrlimit() were being ignored. Now they are
3823 logged to the main and panic logs, but Exim does carry on, to try to do its
3824 job under whatever limits there are.
3826 18. Imported PCRE 5.0.
3828 19. Trivial typo in log message " temporarily refused connection" (the leading
3831 20. If the log selector return_path_on_delivery was set and an address was
3832 redirected to /dev/null, the delivery process crashed because it assumed
3833 that a return path would always be set for a "successful" delivery. In this
3834 case, the whole delivery is bypassed as an optimization, and therefore no
3837 21. Internal re-arrangement: the function for sending a challenge and reading
3838 a response while authentication was assuming a zero-terminated challenge
3839 string. It's now changed to take a pointer and a length, to allow for
3840 binary data in such strings.
3842 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3844 23. Exim was not respecting finduser_retries when seeking the login of the
3845 uid under which it was called; it was always trying 10 times. (The default
3846 setting of finduser_retries is zero.) Also, it was sleeping after the final
3847 failure, which is pointless.
3849 24. Implemented tls_on_connect_ports.
3851 25. Implemented acl_smtp_predata.
3853 26. If the domain in control=submission is set empty, Exim assumes that the
3854 authenticated id is a complete email address when it generates From: or
3855 Sender: header lines.
3857 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3858 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3859 chown and chgrp in /bin and hostname in /usr/bin.
3861 28. Exim was keeping the "process log" file open after each use, just as it
3862 does for the main log. This opens the possibility of it remaining open for
3863 long periods when the USR1 signal hits a daemon. Occasional processlog
3864 errors were reported, that could have been caused by this. Anyway, it seems
3865 much more sensible not to leave this file open at all, so that is what now
3868 29. The long-running daemon process does not normally write to the log once it
3869 has entered its main loop, and it closes the log before doing so. This is
3870 so that log files can straightforwardly be renamed and moved. However,
3871 there are a couple of unusual error situations where the daemon does write
3872 log entries, and I had neglected to close the log afterwards.
3874 30. The text of an SMTP error response that was received during a remote
3875 delivery was being truncated at 512 bytes. This is too short for some of
3876 the long messages that one sometimes sees. I've increased the limit to
3879 31. It is now possible to make retry rules that apply only when a message has a
3880 specific sender, in particular, an empty sender.
3882 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3883 it possible to be selective about when SMTP synchronization is enforced.
3885 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3887 32. Implemented hosts_connection_nolog.
3889 33. Added an ACL for QUIT.
3891 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3894 35. Added mailbox_size and mailbox_filecount to appendfile.
3896 36. Added control = no_multiline_responses to ACLs.
3898 37. There was a bug in the logic of the code that waits for the clock to tick
3899 in the case where the clock went backwards by a substantial amount such
3900 that the microsecond fraction of "now" was more than the microsecond
3901 fraction of "then" (but the whole seconds number was less).
3903 38. Added support for the libradius Radius client library this is found on
3904 FreeBSD (previously only the radiusclient library was supported).
3910 1. When certain lookups returned multiple values in the form name=value, the
3911 quoting of the values was not always being done properly. Specifically:
3912 (a) If the value started with a double quote, but contained no whitespace,
3914 (b) If the value contained whitespace other than a space character (i.e.
3915 tabs or newlines or carriage returns) it was not quoted.
3916 This fix has been applied to the mysql and pgsql lookups by writing a
3917 separate quoting function and calling it from the lookup code. The fix
3918 should probably also be applied to nisplus, ibase and oracle lookups, but
3919 since I cannot test any of those, I have not disturbed their existing code.
3921 2. A hit in the callout cache for a specific address caused a log line with no
3922 reason for rejecting RCPT. Now it says "Previous (cached) callout
3923 verification failure".
3925 3. There was an off-by-one bug in the queryprogram router. An over-long
3926 return line was truncated at 256 instead of 255 characters, thereby
3927 overflowing its buffer with the terminating zero. As well as fixing this, I
3928 have increased the buffer size to 1024 (and made a note to document this).
3930 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3931 when Exim is waiting for an SMTP response from a remote server, Exim
3932 restarts its select() call on the socket, thereby resetting its timeout.
3933 This is not a problem when such interrupts are rare. Somebody set up a cron
3934 job to run exiwhat every 2 minutes, which is less than the normal select()
3935 timeout (5 or 10 minutes). This meant that the select() timeout never
3936 kicked in because it was always reset. I have fixed this by comparing the
3937 time when an interrupt arrives with the time at the start of the first call
3938 to select(). If more time than the timeout has elapsed, the interrupt is
3939 treated as a timeout.
3941 5. Some internal re-factoring in preparation for the addition of Sieve
3942 extensions (by MH). In particular, the "personal" test is moved to a
3943 separate function, and given an option for scanning Cc: and Bcc: (which is
3944 not set for Exim filters).
3946 6. When Exim created an email address using the login of the caller as the
3947 local part (e.g. when creating a From: or Sender: header line), it was not
3948 quoting the local part when it contained special characters such as @.
3950 7. Installed new OpenBSD configuration files.
3952 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3953 try to make them clearer.
3955 9. Callout options, other than the timeout value, were being ignored when
3956 verifying sender addresses in header lines. For example, when using
3958 verify = header_sender/callout=no_cache
3960 the cache was (incorrectly) being used.
3962 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3963 only the Cygwin environment.
3965 11. When return_path_on_delivery was set as a log selector, if different remote
3966 addresses in the same message used different return paths and parallel
3967 remote delivery occurred, the wrong values would sometimes be logged.
3968 (Whenever a remote delivery process finished, the return path value from
3969 the most recently started remote delivery process was logged.)
3971 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3972 lines when AUTH and/or TLS are in use. This is the "received protocol"
3973 field. Exim used to use "asmtp" for authenticated SMTP, without any
3974 indication (in the protocol name) for TLS use. Now it follows the RFC and
3975 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3976 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3977 names appear in log lines as well as in Received: header lines.
3979 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3980 extensions, and comparison tests, and to fix some bugs.
3982 14. Changes to the "personal" filter test:
3984 (1) The test was buggy in that it was just doing the equivalent of
3985 "contains" tests on header lines. For example, if a user's address was
3986 anne@some.where, the "personal" test would incorrectly be true for
3988 To: susanne@some.where
3990 This test is now done by extracting each address from the header in turn,
3991 and checking the entire address. Other tests that are part of "personal"
3992 are now done using regular expressions (for example, to check local parts
3993 of addresses in From: header lines).
3995 (2) The list of non-personal local parts in From: addresses has been
3996 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3997 taken from the Sieve specification recommendations.
3999 (3) If the message contains any header line starting with "List-" it is
4000 treated as non-personal.
4002 (4) The test for "circular" in the Subject: header line has been removed
4003 because it now seems ill-conceived.
4005 15. Minor typos in src/EDITME comments corrected.
4007 16. Installed latest exipick from John Jetmore.
4009 17. If headers_add on a router specified a text string that was too long for
4010 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4011 of string_sprintf() is now avoided.
4013 18. $message_body_size was not set (it was always zero) when running the DATA
4014 ACL and the local_scan() function.
4016 19. For the "mail" command in an Exim filter, no default was being set for
4017 the once_repeat time, causing a random time value to be used if "once" was
4018 specified. (If the value happened to be <= 0, no repeat happened.) The
4019 default is now 0s, meaning "never repeat". The "vacation" command was OK
4020 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4021 (I found it when inspecting the code).
4023 20. There is now an overall timeout for performing a callout verification. It
4024 defaults to 4 times the callout timeout, which applies to individual SMTP
4025 commands during the callout. The overall timeout applies when there is more
4026 than one host that can be tried. The timeout is checked before trying the
4027 next host. This prevents very long delays if there are a large number of
4028 hosts and all are timing out (e.g. when the network connections are timing
4029 out). The value of the overall timeout can be changed by specifying an
4030 additional sub-option for "callout", called "maxwait". For example:
4032 verify = sender/callout=5s,maxwait=20s
4034 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4035 to the end before writing, but this should make it even safer).
4037 22. Exim was forgetting that it had advertised PIPELINING for the second and
4038 subsequent messages on an SMTP connection. It was also not resetting its
4039 memory on STARTTLS and an internal HELO.
4041 23. When Exim logs an SMTP synchronization error within a session, it now
4042 records whether PIPELINING has been advertised or not.
4044 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4045 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4046 rather than long int.
4048 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4050 26. Added the never_mail option to autoreply.
4056 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4057 crash if the getsockname() call failed; this can happen if a connection is
4058 closed very soon after it is established. The problem was simply in the
4059 order in which certain operations were done, causing Exim to try to write
4060 to the SMTP stream before it had set up the file descriptor. The bug has
4061 been fixed by making things happen in the correct order.
4067 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4068 before the connection was closed, thus losing the rejection response.
4070 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4071 some early Solaris releases, but causes trouble in current releases where
4072 socklen_t is defined.
4074 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4077 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4080 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4081 that releasing the top of it at the end releases what was used for sub-
4082 expansions (unless the block got too big). However, discard this block if
4083 the first thing is a variable or header, so that we can use its block when
4084 it is dynamic (useful for very large $message_headers, for example).
4086 6. Lookups now cache *every* query, not just the most recent. A new, separate
4087 store pool is used for this. It can be recovered when all lookup caches are
4088 flushed. Lookups now release memory at the end of their result strings.
4089 This has involved some general refactoring of the lookup sources.
4091 7. Some code has been added to the store_xxx() functions to reduce the amount
4092 of flapping under certain conditions.
4094 8. log_incoming_interface used to affect only the <= reception log lines. Now
4095 it causes the local interface and port to be added to several more SMTP log
4096 lines, for example "SMTP connection from", and rejection lines.
4098 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4100 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4102 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4103 could be overwritten at the end of the current message (or the start of a
4104 new message if it was set in a HELO ACL). The value is now preserved for
4105 the duration of the SMTP connection.
4107 12. If a transport had a headers_rewrite setting, and a matching header line
4108 contained an unqualified address, that address was qualified, even if it
4109 did not match any rewriting rules. The underlying bug was that the values
4110 of the flags that permit the existence of unqualified sender and recipient
4111 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4112 non-local messages, and by -bnq for local messages) were not being
4113 preserved with the message after it was received.
4115 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4116 "next input=" as part of the text comprising the host identity instead of
4117 the correct text. The code was using the same buffer for two different
4118 strings. However, depending on which order the printing function evaluated
4119 its arguments, the bug did not always show up. Under Linux, for example, my
4120 test suite worked just fine.
4122 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4123 doesn't work with some older versions of Perl. It has been changed to "my",
4124 which in any case is probably the better facility to use.
4126 15. A really picky compiler found some instances of statements for creating
4127 error messages that either had too many or two few arguments for the format
4130 16. The size of the buffer for calls to the DNS resolver has been increased
4131 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4132 for addresses that have a lot of PTR records. This alleviates a problem; it
4133 does not fully solve it.
4135 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4136 buffer now truncates the list and logs the incident, which is the same
4137 action as happens when Exim is looking up a host name and its aliases.
4138 Previously in this situation something unpredictable would happen;
4139 sometimes it was "internal error: store_reset failed".
4141 18. If a server dropped the connection unexpectedly when an Exim client was
4142 using GnuTLS and trying to read a response, the client delivery process
4143 crashed while trying to generate an error log message.
4145 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4146 string, for example:
4148 warn message = H1: something\nH2: something
4150 the text was added as a single header line from Exim's point of view
4151 though it ended up OK in the delivered message. However, searching for the
4152 second and subsequent header lines using $h_h2: did not work. This has been
4153 fixed. Similarly, if a system filter added multiple headers in this way,
4154 the routers could not see them.
4156 20. Expanded the error message when iplsearch is called with an invalid key to
4157 suggest using net-iplsearch in a host list.
4159 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4160 ACLs are no longer actually imposed (and a message to that effect is
4163 22. If a "gecos" field in a passwd entry contained escaped characters, in
4164 particular, if it contained a \" sequence, Exim got it wrong when building
4165 a From: or a Sender: header from that name. A second bug also caused
4166 incorrect handling when an unquoted " was present following a character
4167 that needed quoting.
4169 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4170 was not being matched caselessly.
4172 24. Arranged for all hyphens in the exim.8 source to be escaped with
4175 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4176 properly. Recipient callout cache records were still being keyed to include
4177 the sender, even when use_sender was set false. This led to far more
4178 callouts that were necessary. The sender is no longer included in the key
4179 when use_sender is false.
4181 26. Added "control = submission" modifier to ACLs.
4183 27. Added the ${base62d: operator to decode base 62 numbers.
4185 28. dnsdb lookups can now access SRV records.
4187 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4188 the configuration file.
4190 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4191 (-v) mode. This makes the output for a verbose queue run more intelligible.
4193 31. Added a use_postmaster feature to recipient callouts.
4195 32. Added the $body_zerocount variable, containing the number of binary zero
4196 bytes in the message body.
4198 33. The time of last modification of the "new" subdirectory is now used as the
4199 "mailbox time last read" when there is a quota error for a maildir
4202 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4204 35. Added +ignore_unknown as a special item in host lists.
4206 36. Code for decoding IPv6 addresses in host lists is now included, even if
4207 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4208 address was recognized as an IP address, but was then not correctly decoded
4209 into binary, causing unexpected and incorrect effects when compared with
4216 1. Very minor rewording of debugging text in manualroute to say "list of
4217 hosts" instead of "hostlist".
4219 2. If verify=header_syntax was set, and a header line with an unqualified
4220 address (no domain) and a large number of spaces between the end of the
4221 name and the colon was received, the reception process suffered a buffer
4222 overflow, and (when I tested it) crashed. This was caused by some obsolete
4223 code that should have been removed. The fix is to remove it!
4225 3. When running in the test harness, delay a bit after writing a bounce
4226 message to get a bit more predictability in the log output.
4228 4. Added a call to search_tidyup() just before forking a reception process. In
4229 theory, someone could use a lookup in the expansion of smtp_accept_max_
4230 per_host which, without the tidyup, could leave open a database connection.
4232 5. Added the variables $recipient_data and $sender_data which get set from a
4233 lookup success in an ACL "recipients" or "senders" condition, or a router
4234 "senders" option, similar to $domain_data and $local_part_data.
4236 6. Moved the writing of debug_print from before to after the "senders" test
4239 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4240 problems for message scanning, either using a data ACL, or using
4241 local_scan() because the Received: header was not generated till after they
4242 were called (in order to set the time as the time of reception completion).
4243 I have revised the way this works. The header is now generated after the
4244 body is received, but before the ACL or local_scan() are called. After they
4245 are run, the timestamp in the header is updated.
4251 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4252 before starting a queue runner without re-exec. This happened only when
4253 deliver_drop_privilege was set or when the Exim user was set to root. The
4254 effect of the bug was that timeouts during subsequent deliveries caused
4255 crashes instead of being properly handled. The handler is now left at its
4256 default (and expected) setting.
4258 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4259 message, again when deliver_drop_privilege is set or Exim is run as root.
4260 The bug described in (1) was not present in this case, but the tidying up
4261 of the other signals was missing. I have made the two cases consistent.
4263 3. The ignore_target_hosts setting on a manualroute router was being ignored
4264 for hosts that were looked up using the /MX notation.
4266 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4269 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4270 operated on the sender address. After changing the $sender_address to <>
4271 for the sender address verify, Exim was re-instated it as the original
4272 (before rewriting) address, but remembering that it had rewritten it, so it
4273 wasn't rewriting it again. This bug also had the effect of breaking the
4274 sender address verification caching when the sender address was rewritten.
4276 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4277 This has been changed so that if the ip literal address matches
4278 ignore_target_hosts, the router declines.
4280 7. Added expansion conditions match_domain, match_address, and match_local_
4281 part (NOT match_host).
4283 8. The placeholder for the Received: header didn't have a length field set.
4285 9. Added code to Exim itself and to exim_lock to test for a specific race
4286 condition that could lead to file corruption when using MBX delivery. The
4287 issue is with the lockfile that is created in /tmp. If this file is removed
4288 after a process has opened it but before that process has acquired a lock,
4289 there is the potential for a second process to recreate the file and also
4290 acquire a lock. This could lead to two Exim processes writing to the file
4291 at the same time. The added code performs the same test as UW imapd; it
4292 checks after acquiring the lock that its file descriptor still refers to
4293 the same named file.
4295 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4296 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4299 11. Added the smtp_active_hostname option. If used, this will typically be made
4300 to depend on the incoming interface address. Because $interface_address is
4301 not set up until the daemon has forked a reception process, error responses
4302 that can happen earlier (such as "too many connections") no longer contain
4305 12. If an expansion in a condition on a "warn" statement fails because a lookup
4306 defers, the "warn" statement is abandoned, and the next ACL statement is
4307 processed. Previously this caused the whole ACL to be aborted.
4309 13. Added the iplsearch lookup type.
4311 14. Added ident_timeout as a log selector.
4313 15. Added tls_certificate_verified as a log selector.
4315 16. Added a global option tls_require_ciphers (compare the smtp transport
4316 option of the same name). This controls incoming TLS connections.
4318 17. I finally figured out how to make tls_require_ciphers do a similar thing
4319 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4320 before starting the TLS session.
4322 18. Tabs are now shown as \t in -bP output.
4324 19. If the log selector return_path_on_delivery was set, Exim crashed when
4325 bouncing a message because it had too many Received: header lines.
4327 20. If two routers both had headers_remove settings, and the first one included
4328 a superfluous trailing colon, the final name in the first list and the
4329 first name in the second list were incorrectly joined into one item (with a
4330 colon in the middle).
4336 1. Added -C and -D options to the exinext utility, mainly to make it easier
4337 to include in the automated testing, but these could be helpful when
4338 multiple configurations are in use.
4340 2. The exinext utility was not formatting the output nicely when there was
4341 an alternate port involved in the retry record key, nor when there was a
4342 message id as well (for retries that were specific to a specific message
4343 and a specific host). It was also confused by IPv6 addresses, because of
4344 the additional colons they contain. I have fixed the IPv4 problem, and
4345 patched it up to do a reasonable job for IPv6.
4347 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4348 delivery, the log line now contains "pipelined" if PIPELINING was used.
4350 4. An SMTP transport process used to panic and die if the bind() call to set
4351 an explicit outgoing interface failed. This has been changed; it is now
4352 treated in the same way as a connect() failure.
4354 5. A reference to $sender_host_name in the part of a conditional expansion
4355 that was being skipped was still causing a DNS lookup. This no longer
4358 6. The def: expansion condition was not recognizing references to header lines
4359 that used bh_ and bheader_.
4361 7. Added the _cache feature to named lists.
4363 8. The code for checking quota_filecount in the appendfile transport was
4364 allowing one more file than it should have been.
4366 9. For compatibility with Sendmail, the command line option
4374 and sets the incoming protocol and host name (for trusted callers). The
4375 host name and its colon can be omitted when only the protocol is to be set.
4376 Note the Exim already has two private options, -pd and -ps, that refer to
4377 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4378 "s", but I don't think that's a major issue.
4380 10. A number of refactoring changes to the code, none of which should affect
4383 (a) The number of logging options was getting close to filling up the
4384 32-bit word that was used as a bit map. I have split them into two classes:
4385 those that are passed in the argument to log_write(), and those that are
4386 only ever tested independently outside of that function. These are now in
4387 separate 32-bit words, so there is plenty of room for expansion again.
4388 There is no change in the user interface or the logging behaviour.
4390 (b) When building, for example, log lines, the code previously used a
4391 macro that called string_cat() twice, in order to add two strings. This is
4392 not really sufficiently general. Furthermore, there was one instance where
4393 it was actually wrong because one of the argument was used twice, and in
4394 one call a function was used. (As it happened, calling the function twice
4395 did not affect the overall behaviour.) The macro has been replaced by a
4396 function that can join an arbitrary number of extra strings onto a growing
4399 (c) The code for expansion conditions now uses a table and a binary chop
4400 instead of a serial search (which was left over from when there were very
4401 few conditions). Also, it now recognizes conditions like "pam" even when
4402 the relevant support is not compiled in: a suitably worded error message is
4403 given if an attempt is made to use such a condition.
4405 11. Added ${time_interval:xxxxx}.
4407 12. A bug was causing one of the ddress fields not to be passed back correctly
4408 from remote delivery subprocesses. The field in question was not being
4409 subsequently used, so this caused to problems in practice.
4411 13. Added new log selectors queue_time and deliver_time.
4413 14. Might have fixed a bug in maildirsizefile handling that threw up
4414 "unexpected character" debug warnings, and recalculated the data
4415 unnecessarily. In any case, I expanded the warning message to give more
4418 15. Added the message "Restricted characters in address" to the statements in
4419 the default ACL that block characters like @ and % in local parts.
4421 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4422 Three changes have been made:
4424 (a) There was a serious bug; a negative response to MAIL caused the whole
4425 recipient domain to be cached as invalid, thereby blocking all messages
4426 to all local parts at the same domain, from all senders. This bug has
4427 been fixed. The domain is no longer cached after a negative response to
4428 MAIL if the sender used is not empty.
4430 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4433 (c) A new callout option, "use_sender" has been added for people who want
4434 the modified behaviour.
4440 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4443 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4444 indeed breaks things for older releases.
4446 3. Added additional logging to the case where there is a problem reading data
4447 from a filter that is running in a subprocess using a pipe, in order to
4448 try to track down a specific problem.
4450 4. Testing facility fudge: when running in the test harness and attempting
4451 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4452 getting "No route to host". Convert this to a timeout.
4454 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4457 6. Some OS don't have socklen_t but use size_t instead. This affects the
4458 fifth argument of getsockopt() amongst other things. This is now
4459 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4460 can be set for individual OS. I have set it for SunOS5, OSF1, and
4461 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4462 some earlier ones do not.
4464 7. Change 4.30/15 was not doing the test caselessly.
4466 8. The standard form for an IPv6 address literal was being rejected by address
4467 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4468 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4469 this, as well as the form without the "IPv6" on the front (but only when
4470 address literals are enabled, of course).
4472 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4474 10. Exim crashed if a message with an empty sender address specified by -f
4475 encountered a router with an errors_to setting. This could be provoked only
4476 by a command such as
4480 where an empty string was supplied; "<>" did not hit this bug.
4482 11. Installed PCRE release 4.5.
4484 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4485 remained set. It is now erased.
4487 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4488 times from message ids (which are base 36 rather than the normal 62).
4490 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4491 were being counted as actual protocol errors, and logged if the log
4492 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4493 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4494 and DATA following a set of rejected RCPTs do not count as protocol errors.
4495 In other words, Exim assumes they were pipelined, though this may not
4496 actually be the case. Of course, in all cases the client gets an
4497 appropriate error code.
4499 15. If a lookup fails in an ACL condition, a message about the failure may
4500 be available; it is used if testing the ACL cannot continue, because most
4501 such messages specify what the cause of the deferral is. However, some
4502 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4503 that caused an old message to be retained and used if a later statement
4504 caused a defer, replacing the real cause of the deferral.
4506 16. If an IP address had so many PTR records that the DNS lookup buffer
4507 was not large enough to hold them, Exim could crash while trying to process
4508 the truncated data. It now detects and logs this case.
4510 17. Further to 4.21/58, another change has been made: if (and only if) the
4511 first line of a message (the first header line) ends with CRLF, a bare LF
4512 in a subsequent header line has a space inserted after it, so as not to
4513 terminate the header.
4515 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4516 unnecessarily, used atoi() instead of strtol(), and didn't check the
4517 termination when getting file sizes from file names by regex.
4519 19. Completely re-implemented the support for maildirsize files, in the light
4520 of a number of problems with the previous contributed implementation
4521 (4.30/29). In particular:
4523 . If the quota is zero, the maildirsize file is maintained, but no quota is
4526 . If the maildir directory does not exist, it is created before any attempt
4527 to write a maildirsize file.
4529 . The quota value in the file is just a cache; if the quota is changed in
4530 the transport, the new value overrides.
4532 . A regular expression is available for excluding directories from the
4535 20. The autoreply transport checks the characters in options that define the
4536 message's headers; it allows continued headers, but it was checking with
4537 isspace() after an embedded newline instead of explicitly looking for a
4540 21. If all the "regular" hosts to which an address was routed had passed their
4541 expiry times, and had not reached their retry times, the address was
4542 bounced, even if fallback hosts were defined. Now Exim should go on to try
4545 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4546 equivalent code in the SMTP transport. Some hosts send humungous responses
4547 to HELO/EHLO, more than 1024 it seems.
4549 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4550 gives compiler warnings in some environments. I've now done it "properly",
4553 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4554 (because of problems with the built-in one) was declared to return uschar *
4555 instead of char *, causing compiler failure.
4557 25. Fixed a file descriptor leak when processing alias/forward files.
4559 26. Fixed a minor format string issue in dbfn.c.
4561 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4563 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4564 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4567 29. When a sender address is verified, it is cached, to save repeating the test
4568 when there is more than one recipient in a message. However, when the
4569 verification involves a callout, it is possible for different callout
4570 options to be set for different recipients. It is too complicated to keep
4571 track of this in the cache, so now Exim always runs a verification when a
4572 callout is required, relying on the callout cache for the optimization.
4573 The overhead is duplication of the address routing, but this should not be
4576 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4577 to be verified with callout=postmaster, and the main callout worked but the
4578 postmaster check failed, the verification correctly failed. However, if a
4579 subsequent RCPT command asked for sender verification *without* the
4580 postmaster check, incorrect caching caused this verification also to fail,
4583 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4584 it was not caching the DNS options (qualify_single, search_parents) that
4585 were used when the lookup failed. A subsequent lookup with different
4586 options therefore always gave the same answer, though there were cases
4587 where it should not have. (Example: a "domains = !$mx_any" option on a
4588 dnslookup router: the "domains" option is always processed without any
4589 widening, but the router might have qualify_single set.) Now Exim uses the
4590 cached value only when the same options are set.
4592 32. Added John Jetmore's "exipick" utility to the distribution.
4594 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4595 than a timeout (e.g. a certificate is required, and is not provided), an
4596 Exim server now closes the connection immediately. Previously it waited for
4597 the client to close - but if the client is SSL, it seems that they each
4598 wait for each other, leading to a delay before one of them times out.
4600 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4601 maintained 0.8.x compatibility because I don't think many are using it, and
4602 it is clearly obsolete.
4604 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4607 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4608 client certificate was expired. A simple patch fixes this, though I don't
4609 understand the full logic of why the verify callback is called multiple
4612 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4613 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4614 which causes problems with some clients (such as the Certicom SSL Plus
4615 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4616 disables the coutermeasure allowing Eudora to connect."
4618 38. Exim was not checking that a write() to a log file succeeded. This could
4619 lead to Bad Things if a log got too big, in particular if it hit a file
4620 size limit. Exim now panics and dies if it cannot write to a log file, just
4621 as it does if it cannot open a log file.
4623 39. Modified OS/Makefile-Linux so that it now contains
4625 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4627 The two -D definitions ensure that Exim is compiled with large file
4628 support, which makes it possible to handle log files that are bigger than
4631 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4632 instance) a domain was checked against a named list that involved a lookup,
4633 causing $domain_data to be set, then another domain was checked against the
4634 same list, then the first domain was re-checked, the value of $domain_data
4635 after the final check could be wrong. In particular, if the second check
4636 failed, it could be set empty. This bug probably also applied to
4639 41. The strip_trailing_dot option was not being applied to the address given
4640 with the -f command-line option.
4642 42. The code for reading a message's header from the spool was incrementing
4643 $received_count, but never initializing it. This meant that the value was
4644 incorrect (doubled) while delivering a message in the same process in which
4645 it was received. In the most common configuration of Exim, this never
4646 happens - a fresh exec is done - but it can happen when
4647 deliver_drop_privilege is set.
4649 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4650 it now includes up to 150 characters of the unexpected data in the log
4653 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4654 and building data strings. The size of both of these buffers was 10 000
4655 bytes - far larger than anybody would *ever* want, thought I. Needless to
4656 say, somebody hit the limit. I have increased the maximum line length to
4657 20 000 and the maximum data length of concatenated lines to 100 000. I have
4658 also fixed two bugs, because there was no checking on these buffers. Tsk,
4659 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4660 buffer is too small.
4662 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4663 lsearch lookups. Now it does.
4665 46. When parsing a route_list item in a manualroute router, a fixed-length
4666 buffer was used for the list of hosts. I made this 1024 bytes long,
4667 thinking that nobody would ever have a list of hosts that long. Wrong.
4668 Somebody had a whole pile of complicated expansion conditions, and the
4669 string was silently truncated, leading to an expansion error. It turns out
4670 that it is easier to change to an unlimited length (owing to other changes
4671 that have happened since this code was originally written) than to build
4672 structure for giving a limitation error. The length of the item that
4673 expands into the list of hosts is now unlimited.
4675 47. The lsearch lookup could not handle data where the length of text line was
4676 more than 4095 characters. Such lines were truncated, leading to shortened
4677 data being returned. It should now handle lines of any length.
4679 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4680 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4683 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4684 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4685 no longer happens in comment lines. A list of replacements is now placed
4686 at the head of all of the source files, except those whose only change is
4687 to replace PERL_COMMAND in the very first #! line.
4689 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4690 messages on the queue, with a bottom-up merge sort, using code contributed
4691 by Michael Haardt. This should make operations like -bp somewhat faster on
4692 large queues. It won't affect queue runners, except when queue_run_in_order
4695 51. Installed eximstats 1.31 in the distribution.
4697 52. Added support for SRV lookups to the dnslookup router.
4699 53. If an ACL referred to $message_body or $message_body_end, the value was not
4700 reset for any messages that followed in the same SMTP session.
4702 54. The store-handling optimization for building very long strings was not
4703 differentiating between the different store pools. I don't think this
4704 actually made any difference in practice, but I've tidied it.
4706 55. While running the routers to verify a sender address, $sender_address
4707 was still set to the sender address. This is wrong, because when routing to
4708 send a bounce to the sender, it would be empty. Therefore, I have changed
4709 it so that, while verifying a sender address, $sender_address is set to <>.
4710 (There is no change to what happens when verifying a recipient address.)
4712 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4713 target A or AAAA records (if not already returned) without resetting the
4714 qualify_single or search_parents options of the DNS resolver. These are
4715 inappropriate in this case because the targets of MX and SRV records must
4716 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4717 target that, when qualified, matched something in the local domain. These
4718 two options are now turned off when doing these lookups.
4720 57. It seems that at least some releases of Reiserfs (which does not have the
4721 concept of a fixed number of inodes) returns zero and not -1 for the
4722 number of available inodes. This interacted badly with check_spool_inodes,
4723 which assumed that -1 was the "no such thing" setting. What I have done is
4724 to check that the total number of inodes is greater than zero before doing
4725 the test of how many are available.
4727 58. When a "warn" ACL statement has a log_message modifier, the message is
4728 remembered, and not repeated. This is to avoid a lot of repetition when a
4729 message has many recipients that cause the same warning to be written.
4730 Howewer, Exim was preserving the list of already written lines for an
4731 entire SMTP session, which doesn't seem right. The memory is now reset if a
4732 new message is started.
4734 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4735 debugging output unless log_rewrite was also set.
4737 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4738 of (int)(handle) when we know that handle contains (void *)(-1).
4740 61. The Exim daemon panic-logs an error return when it closes the incoming
4741 connection. However "connection reset by peer" seems to be common, and
4742 isn't really an error worthy of noting specially, so that particular error
4745 62. When Exim is trying to find all the local interfaces, it used to panic and
4746 die if the ioctl to get the interface flags failed. However, it seems that
4747 on at least one OS (Solaris 9) it is possible to have an interface that is
4748 included in the list of interfaces, but for which you get a failure error
4749 for this call. This happens when the interface is not "plumbed" into a
4750 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4751 failure of the "get flags" call assumes that the interface is down.
4753 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4754 makes life easier for people who are doing arithmetic on fields extracted
4755 from dates, where you often get leading zeros that should not be
4756 interpreted as octal.
4758 64. Added qualify_domain to the redirect router, to override the global
4761 65. If a pathologically long header line contained very many addresses (the
4762 report of this problem mentioned 10 000) and each of them was rewritten,
4763 Exim could use up a very large amount of memory. (It kept on making new
4764 copies of the header line as it rewrote, and never released the old ones.)
4765 At the expense of a bit more processing, the header rewriting function has
4766 been changed so that it no longer eats memory in this way.
4768 66. The generation of the Received: header has been moved from the time that a
4769 message starts to be received, to the time that it finishes. The timestamp
4770 in the Received: header should now be very close to that of the <= log
4771 line. There are two side-effects of this change:
4773 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4774 logged header lines no longer include the local Received: line, because
4775 it has not yet been created. The same applies to a copy of the message
4776 that is returned to a non-SMTP sender when a message is rejected.
4778 (b) When a filter file is tested using -bf, no additional Received: header
4779 is added to the test message. After some thought, I decided that this
4782 This change does not affect the value of $received_for. It is still set
4783 after address rewriting, but before local_scan() is called.
4785 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4787 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4788 gave an unhelpful panic error message, and a defer error. I have managed to
4789 change this behaviour so that it now rejects any supplied certificate,
4790 which seems right, as the list of acceptable certificates is empty.
4792 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4793 gave an unhelpful defer error. I have not managed to make this reject any
4794 supplied certificates, but the error message it gives is "no certificate
4795 supplied", which is not helpful.
4797 70. exigrep's output now also includes lines that are not associated with any
4798 message, but which match the given pattern. Implemented by a patch from
4799 Martin Sluka, which also tidied up the Perl a bit.
4801 71. Recipient callout verification, like sender verification, was using <> in
4802 the MAIL FROM command. This isn't really the right thing, since the actual
4803 sender may affect whether the remote host accepts the recipient or not. I
4804 have changed it to use the actual sender in the callout; this means that
4805 the cache record is now keyed on a recipient/sender pair, not just the
4806 recipient address. There doesn't seem to be a real danger of callout loops,
4807 since a callout by the remote host to check the sender would use <>.
4808 [SEE ABOVE: changed after hitting problems.]
4810 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4811 temporary errors. However, in the case of such a code being given after
4812 the end of a data transmission (i.e. after ".") Exim was failing to write
4813 a retry record for the message. (Yes, there was some broken host that was
4814 actually sending 8xx at this point.)
4816 73. An unknown lookup type in a host list could cause Exim to panic-die when
4817 the list was checked. (An example that provoked this was putting <; in the
4818 middle of a list instead of at the start.) If this happened during a DATA
4819 ACL check, a -D file could be left lying around. This kind of configuration
4820 error no longer causes Exim to die; instead it causes a defer errror. The
4821 incident is still logged to the main and panic logs.
4823 74. Buglet left over from Exim 3 conversion. The message "too many messages
4824 in one connection" was written to the rejectlog but not the mainlog, except
4825 when address rewriting (yes!) was being logged.
4827 75. Added write_rejectlog option.
4829 76. When a system filter was run not as root (that is, when system_filter_user
4830 was set), the values of the $n variables were not being returned to the
4831 main process; thus, they were not subsequently available in the $sn
4834 77. Added +return_path_on_delivery log selector.
4836 78. A connection timeout was being treated differently from recipients deferred
4837 when testing hosts_max_try with a message that was older than the host's
4838 retry timeout. (The host should not be counted, thus allowing all hosts to
4839 be tried at least once before bouncing.) This may have been the cause of an
4840 occasionally reported bug whereby a message would remain on the queue
4841 longer than the retry timeout, but would be bounced if a delivery was
4842 forced. I say "may" because I never totally pinned down the problem;
4843 setting up timeout/retry tests is difficult. See also the next item.
4845 79. The ultimate address timeout was not being applied to errors that involved
4846 a combination of host plus message (for example, a timeout on a MAIL
4847 command). When an address resolved to a number of possible hosts, and they
4848 were not all tried for each delivery (e.g. because of hosts_max_try), a
4849 message could remain on the queue longer than the retry timeout.
4851 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4854 81. Fixed an obscure SMTP outgoing bug which required at least the following
4855 conditions: (a) there was another message waiting for the same server;
4856 (b) the server returned 5xx to all RCPT commands in the first message so
4857 that the message was not completed; (c) the server dropped the connection
4858 or gave a negative response to the RSET that Exim sends to abort the
4859 transaction. The observed case was a dropped connection after DATA that had
4860 been sent in pipelining mode. That is, the server had advertised PIPELINING
4861 but was not implementing it correctly. The effect of the bug was incorrect
4862 behaviour, such as trying another host, and this could lead to a crash.
4868 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4869 and daemon.c were passed as pointers to ints; they should have been
4870 pointers to socklen_t variables (which are typically unsigned ints).
4872 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4875 3. Fixed a really odd bug that affected only the testing scheme; patching a
4876 certain fixed string in the binary changed the value of another string that
4877 happened to be identical to the end of the original first string.
4879 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4880 name", it returns that address as the IP address. On some operating
4881 systems (e.g. Solaris), it also passes back the IP address string as the
4882 "host name". However, on others (e.g. Linux), it passes back an empty
4883 string. Exim wasn't checking for this, and was changing the host name to an
4884 empty string, assuming it had been canonicized.
4886 5. Although rare, it is permitted to have more than one PTR record for a given
4887 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4888 all the names associated with an address, because they do in Solaris.
4889 However, it seems that they do not in Linux for data that comes from the
4890 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4891 I found this out when I moved to a new Linux workstation and tried to run
4892 the Exim test suite.
4894 To get round this problem I have changed the code so that it now does its
4895 own call to the DNS to look up PTR records when searching for a host name.
4896 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4897 addresses that are only in /etc/hosts are still found.
4899 This behaviour is, however, controlled by an option called host_lookup_
4900 order, which defaults to "bydns:byaddr". If people want to use the other
4901 order, or indeed, just use one or the other means of lookup, they can
4902 specify it in this variable.
4904 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4905 some operating systems, this comes back from gethostbyaddr() as an empty
4906 string, and this is what Exim used to test for. However, it seems that in
4907 other systems, "." is yielded. Exim now tests for this case too.
4909 7. The values of check_spool_space and check_log_space are now held internally
4910 as a number of kilobytes instead of an absolute number of bytes. If a
4911 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4912 kilobyte. This means that much larger values can be stored.
4914 8. Exim monitor: an attempt to get the action menu when not actually pointing
4915 at a message produces an empty menu entitled "No message selected". This
4916 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4917 no entries in it ("Shell widget menu has zero width and/or height"). So I
4918 have added a single, blank menu entry in this case.
4920 9. Added ${quote_local_part.
4922 10. MIME decoding is now applied to the contents of Subject: header lines when
4925 11. Now that a reference to $sender_host_address automatically causes a reverse
4926 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4927 host lookup before query-style lookups in lists that might use this
4928 variable. This has therefore been abolished, and the "net-" prefix is no
4929 longer necessary for query-style lookups.
4931 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4932 have been a typo for LFLAGS, so it has been changed.
4934 13. The install script calls Exim with "-C /dev/null" in order to find the
4935 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4936 to be output. Howeve, since Exim outputs its version number before the
4937 error, it didn't break the script. It just looked ugly. I fixed this by
4938 always allowing "-C /dev/null" if the caller is root.
4940 14. Ignore overlarge ACL variable number when reading spool file - insurance
4941 against a later release with more variables having written the file.
4943 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4944 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4945 this, as well as the form without the "IPv6" on the front.
4947 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4948 OS/Makefile-Darwin file.
4950 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4951 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4953 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4954 variable. However, some OS use ints for inodes. I've added cast to long int
4955 to get rid of the compiler warning.
4957 19. I had forgotten to lock out "/../" in configuration file names when
4958 ALT_CONFIG_PREFIX was set.
4960 20. Routers used for verification do not need to specify transports. However,
4961 if such a router generated a host list, and callout was configured, Exim
4962 crashed, because it could not find a port number from the (non-existent)
4963 transport. It now assumes port 25 in this circumstance.
4965 21. Added the -t option to exigrep.
4967 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4968 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4969 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4970 them is set, LOOKUP_LSEARCH is forced.
4972 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4974 24. Added sender and host information to the "rejected by local_scan()" log
4975 line; previously there was no indication of these.
4977 25. Added .include_if_exists.
4979 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4980 Linux. It turns out that not all file systems support this. Apparently some
4981 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4982 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4983 not supported on the file descriptor, is now ignored when Exim is trying to
4984 sync a directory. This applies only to Linux.
4986 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4988 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4989 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4992 29. Added support for maildirsize files from supplied patch (modified a bit).
4994 30. The use of :fail: followed by an empty string could lead Exim to respond to
4995 sender verification failures with (e.g.):
4997 550 Verification failed for <xxx>
4998 550 Sender verify failed
5000 where the first response line was missing the '-' that indicates it is not
5001 the final line of the response.
5003 31. The loop for finding the name of the user that called Exim had a hardwired
5004 limit of 10; it now uses the value of finduser_retries, which is used for
5005 all other user lookups.
5007 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5010 33. Exim was neglecting to zero errno before one call of strtol() when
5011 expanding a string and expecting an integer value. On some systems this
5012 resulted in spurious "integer overflow" errors. Also, it was casting the
5013 result into an int without checking.
5015 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5016 did not work. The code looks as if it has *never* worked, though it appears
5017 to have been documented since at least releast 1.62. I have made it work.
5019 35. The "timeout_DNS" error in retry rules, also documented since at least
5020 1.62, also never worked. As it isn't clear exactly what this means, and
5021 clearly it isn't a major issue, I have abolished the feature by treating it
5022 as "timeout", and writing a warning to the main and panic logs.
5024 36. The display of retry rules for -brt wasn't always showing the error code
5027 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5028 timeout_connect_A, timeout_connect_MX.
5030 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5031 to the empty sender.
5033 39. The daemon was not analysing the content of -oX till after it had closed
5034 stderr and disconnected from the controlling terminal. This meant that any
5035 syntax errors were only noted on the panic log, and the return code from
5036 the command was 0. By re-arranging the code a little, I've made the
5037 decoding happen first, so such errors now appear on stderr, and the return
5038 code is 1. However, the actual setting up of the sockets still happens in
5039 the disconnected process, so errors there are still only recorded on the
5042 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5043 connections (as happens on some IP stacks) was logged at start up time as
5044 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5045 it from "IPv6 and IPv4", which means that two separate sockets are being
5048 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5049 says whether AF_INET or AF_INET6 was passed as an argument.
5051 42. Exiwhat output was messed up when time zones were included in log
5054 43. Exiwhat now gives more information about the daemon's listening ports,
5055 and whether -tls-on-connect was used.
5057 44. The "port" option of the smtp transport is now expanded.
5059 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5060 silently ignored. Now an error message is written to the main and panic
5063 46. There's a new ACL modifier called "logwrite" which writes to a log file
5064 as soon as it is encountered.
5066 47. Added $local_user_uid and $local_user_gid at routing time.
5068 48. Exim crashed when trying to verify a sender address that was being
5071 49. Exim was recognizing only a space character after ".include". It now also
5072 recognizes a tab character.
5074 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5075 extracting the relevant information from the specification. The man page no
5076 longer contains scrambled data for the -d option, and I've added a section
5077 at the front about calling Exim under different names.
5079 51. Added "extra_headers" argument to the "mail" command in filter files.
5081 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5084 53. Installed eximstats 1.29.
5086 54. Added transport_filter_timeout as a generic transport option.
5088 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5089 Cc: header lines. This was required by RFC 822, but it not required by RFC
5092 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5093 incoming messages that did not have them. Now it does so only if the
5094 message originates locally, that is, if there is no associated remote host
5095 address. When Resent- header lines are present, this applies to the Resent-
5096 lines rather than the non-Resent- lines.
5098 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5099 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5101 58. Messages for configuration errors now include the name of the main
5102 configuration file - useful now that there may be more than one file in a
5103 list (.included file names were always shown).
5105 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5106 for those rare installations that do not start the daemon as root or run it
5107 setuid root. I've cut out the call to initgroups() if the daemon is not
5110 60. The Exim user and group can now be bound into the binary as text strings
5111 that are looked up at the start of Exim's processing.
5113 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5115 62. Added $mailstore_basename variable.
5117 63. Installed patch to sieve.c from Michael Haardt.
5119 64. When Exim failed to open the panic log after failing to open the main log,
5120 the original message it was trying to log was written to stderr and debug
5121 output, but if they were not available (the usual case in production), it
5122 was lost. Now it is written to syslog before the two lines that record the
5123 failures to open the logs.
5125 65. Users' Exim filters run in subprocesses under the user's uid. It is
5126 possible for a "deliver" command or an alias in a "personal" command to
5127 provoke an address rewrite. If logging of address rewriting is configured,
5128 this fails because the process is not running as root or exim. There may be
5129 a better way of dealing with this, but for the moment (because 4.30 needs
5130 to be released), I have disabled address rewrite logging when running a
5131 filter in a non-root, non-exim process.
5137 1. The buildconfig auxiliary program wasn't quoting the value set for
5138 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5139 not defined. This bug was masked in 4.22 by the effect that was fixed in
5142 2. Some messages that were rejected after a message id was allocated were
5143 shown as "incomplete" by exigrep. It no longer does this for messages that
5144 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5146 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5147 have allow_domain_literals set, the ID did not get logged in the <= line.
5148 Domain literals are now always recognized in Message-ID: header lines.
5150 4. The first argument for a ${extract expansion item is the key name or field
5151 number. Leading and trailing spaces in this item were not being ignored,
5152 causing some misleading effects.
5154 5. When deliver_drop_privilege was set, single queue runner processes started
5155 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5156 same command in the process it spins off) were not dropping privilege.
5158 6. When the daemon running as "exim" started a queue runner, it always
5159 re-executed Exim in the spun-off process. This is a waste of effort when
5160 deliver_drop_privilege is set. The new process now just calls the
5161 queue-runner function directly.
5167 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5170 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5171 ignored. Though the use of -oP was forcing the writing of a pid file, it
5172 was always written to the default place.
5174 3. If the message "no IP address found for host xxxx" is generated during
5175 incoming verification, it is now followed by identification of the incoming
5176 connection (so you can more easily find what provoked it).
5178 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5180 5. Added some features to "harden" Exim a bit more against certain attacks:
5182 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5183 be put in Local/Makefile. This is like the never_users runtime option,
5184 but it cannot be overridden. The default setting is "root".
5186 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5187 prefix string with which any file named in a -C command line option
5190 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5191 is retained for -C and -D only if the caller of Exim is root. Without
5192 it, the exim user may also use -C and -D and retain privilege.
5194 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5195 command line option is disabled.
5197 6. Macro names set by the -D option must start with an upper case letter, just
5198 like macro names defined in the configuration file.
5200 7. Added "dereference=" facility to LDAP.
5202 8. Two instances of the typo "uknown" in the source files are fixed.
5204 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5205 the Configure-Makefile script screwed up while processing it.
5207 10. Incorporated PCRE 4.4.
5209 11. The SMTP synchronization check was not operating right at the start of an
5210 SMTP session. For example, it could not catch a HELO sent before the client
5211 waited for the greeting. There is now a check for outstanding input at the
5212 point when the greeting is written. Because of the duplex, asynchronous
5213 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5214 way, but not yet received, when the check is performed.
5216 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5217 on TCP/IP sockets, because this apparently causes some broken clients to
5220 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5221 unchanged) from the Cygwin maintainer.
5223 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5224 was supported instead of testing for mbx. Effectively a typo.
5226 15. The spa authenticator server code was not checking that the input it
5227 received was valid base64.
5229 16. The debug output line for the "set" modifier in ACLs was not showing the
5230 name of the variable that was being set.
5232 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5234 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5235 lookup on its own. Something else was needed to trigger the lookup. For
5236 example, a match in host_lookup or the need for a host name in a host list.
5237 Now, if $sender_host_name is referenced and the host name has not yet been
5238 looked up, a lookup is performed. If the lookup fails, the variable remains
5239 empty, and $host_lookup_failed is set to "1".
5241 19. Added "eqi" as a case-independent comparison operator.
5243 20. The saslauthd authentication condition could segfault if neither service
5244 nor realm was specified.
5246 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5247 error message that was logged was misleading, and incoming SMTP
5248 connections were dropped. The message is now more accurate, and temporary
5249 errors are given to SMTP connections.
5251 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5252 (except RSET) with a 421 error, until QUIT is received. However, it was
5253 failing to send a response to QUIT.
5255 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5256 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5257 ACL is now run after the helo_try_verify_hosts code.
5259 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5260 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5261 case-independent, so other case variants are also recognized). Apparently
5262 some systems use these upper case variants.
5264 25. If more than two messages were waiting for the same host, and a transport
5265 filter was specified for the transport, Exim sent two messages over the
5266 same TCP/IP connection, and then failed with "socket operation on non-
5267 socket" when it tried to send the third.
5269 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5271 27. The extern definition of crypt16() in expand.c was not being excluded when
5272 the OS had its own crypt16() function.
5274 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5275 as a preferred synonym for return_size_limit, both as an option and as an
5278 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5280 30. Changed the default configuration ACL to relax the local part checking rule
5281 for addresses that are not in any local domains. For these addresses,
5282 slashes and pipe symbols are allowed within local parts, but the sequence
5283 /../ is explicitly forbidden.
5285 31. SPA server authentication was not clearing the challenge buffer before
5288 32. log_message in a "warn" ACL statement was writing to the reject log as
5289 well as to the main log, which contradicts the documentation and doesn't
5290 seem right (because no rejection is happening). So I have stopped it.
5292 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5293 However, I am unable to do any testing of this.
5295 34. Fixed an infelicity in the appendfile transport. When checking directories
5296 for a mailbox, to see if any needed to be created, it was accidentally
5297 using path names with one or more superfluous leading slashes; tracing
5298 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5300 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5301 contents are added to the log line that is written for every discarded
5302 recipient. (Previously a log_message setting was ignored.)
5304 36. The ${quote: operator now quotes the string if it is empty.
5306 37. The install script runs exim in order to find its version number. If for
5307 some reason other than non-existence or emptiness, which it checks, it
5308 could not run './exim', it was installing it with an empty version number,
5309 i.e. as "exim-". This error state is now caught, and the installation is
5312 38. An argument was missing from the function that creates an error message
5313 when Exim fails to connect to the socket for saslauthd authentication.
5314 This could cause Exim to crash, or give a corrupted message.
5316 39. Added isip, isip4, and isip6 to ${if conditions.
5318 40. The ACL variables $acl_xx are now saved with the message, and can be
5319 accessed later in routers, transports, and filters.
5321 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5322 strings in the file are not string-expanded.
5324 42. If a MAIL command specified a SIZE value that was too large to fit into an
5325 int variable, the check against message_size_limit failed. Such values are
5326 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5327 day this will have to be increased, but I don't think I want to be around
5328 when emails are that large.
5335 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5336 iconv() is not standard in FreeBSD.
5338 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5339 IPv6 enabled. The observed symptom was a segmentation fault on return from
5340 the function os_common_find_running_interfaces() in src/os.c.
5342 3. In the check_special_case() function in daemon.c I had used "errno" as an
5343 argument name, which causes warnings on some systems. This was basically a
5344 typo, since it was named "eno" in the comments!
5346 4. The code that waits for the clock to tick (at a resolution of some fraction
5347 of a second) so as to ensure message-id uniqueness was always waiting for
5348 at least one whole tick, when it could have waited for less. [This is
5349 almost certainly not relevant at current processor speeds, where it is
5350 unlikely to ever wait at all. But we try to future-proof.]
5352 5. The function that sleeps for a time interval that includes fractions of a
5353 second contained a race. It did not block SIGALRM between setting the
5354 timer, and suspending (a couple of lines later). If the interval was short
5355 and the sigsuspend() was delayed until after it had expired, the suspension
5356 never ended. On busy systems this could lead to processes getting stuck for
5359 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5360 process, before it forks any delivery processes. The open lookup caching
5361 mechanism meant that the open file or database connection was passed into
5362 the delivery process. The problem was that delivery processes always tidy
5363 up cached lookup data. This could cause a problem for the next delivery
5364 process started by the queue runner, because the external queue runner
5365 process does not know about the closure. So the next delivery process
5366 still has data in the lookup cache. In the case of a file lookup, there was
5367 no problem because closing a file descriptor in a subprocess doesn't affect
5368 the parent. However, if the lookup was caching a connection to a database,
5369 the connection was closed, and the second delivery process was likely to
5370 see errors such as "PGSQL: query failed: server closed the connection
5371 unexpectedly". The problem has been fixed by closing all cached lookups
5372 in a queue runner before running a delivery process.
5374 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5375 seem to have the "const" qualifier which it has on other OS. I've
5378 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5379 *of the same type* (client or server) with the same public name that an
5380 error should be diagnosed.
5382 9. When Exim looked up a host name for an IP address, but failed to find the
5383 original IP address when looking up the host name (a safety check), it
5384 output the message "<ip address> does not match any IP for NULL", which was
5385 confusing, to say the least. The bug was that the host name should have
5386 appeared instead of "NULL".
5388 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5389 user that is built into the binary, and the -C or -D options is used, root
5390 privilege is dropped before the configuration file is read. In addition,
5391 logging is switched to stderr instead of the normal log files. If the
5392 configuration then re-defines the Exim user, the unprivileged environment
5393 is probably not what is expected, so Exim logs a panic warning message (but
5396 However, if deliver_drop_privilege is set, the unprivileged state may well
5397 be exactly what is intended, so the warning has been cut out in that case,
5398 and Exim is allowed to try to write to its normal log files.
5404 1. smtp_return_error_details was not giving details for temporary sender
5405 or receiver verification errors.
5407 2. Diagnose a configuration error if two authenticators have the same public
5410 3. Exim used not to create the message log file for a message until the first
5411 delivery attempt. This could be confusing when incoming messages were held
5412 for policy or load reasons. The message log file is now created at the time
5413 the message is received, and an initial "Received" line is written to it.
5415 4. The automatically generated man page for command line options had a minor
5416 bug that caused no ill effects; however, a more serious problem was that
5417 the procedure for building the man page automatically didn't always
5418 operate. Consequently, release 4.20 contains an out-of-date version. This
5419 shouldn't happen again.
5421 5. When building Exim with embedded Perl support, the script that builds the
5422 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5423 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5425 6. The freeze_tell option was not being used for messages that were frozen on
5426 arrival, either by an ACL or by local_scan().
5428 7. Added the smtp_incomplete_transaction log selector.
5430 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5431 was accepting AUTH without a new EHLO.
5433 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5434 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5437 10. Logging of TCP/IP connections (when configured) now happens in the main
5438 daemon process instead of the child process, so that the TCP/IP connection
5439 count is more accurate (but it can never be perfect).
5441 11. The use of "drop" in a nested ACL was not being handled correctly in the
5442 outer ACL. Now, if condition failure induced by the nested "drop" causes
5443 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5444 or "require"), the connection is dropped.
5446 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5447 that yield "discard" can now be used with an "accept" or a "discard" verb,
5448 but an error is generated for any others (because I can't see a useful way
5449 to define what should happen).
5451 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5452 are now processed in the same way as lines in the Exim configuration file.
5453 In particular, continuation lines are supported.
5455 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5457 15. Added -ti meaning -t -i.
5459 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5460 domains, and warn by logging if others are found.
5462 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5463 returned from GIFCONF when Exim is trying to find the list of interfaces on
5464 a host. The code in os.c has been modified to copy each ifreq to an aligned
5465 structure in all cases.
5467 Also, in some cases, the returned ifreq's were being copied to a 'struct
5468 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5469 means the last couple of bytes of an IPv6 address could be chopped if the
5470 ifreq contained only a normal sockaddr (14 bytes storage).
5472 18. Named domain lists were not supported in the hosts_treat_as_local option.
5473 An entry such as +xxxx was not recognized, and was treated as a literal
5476 19. Ensure that header lines added by a DATA ACL are included in the reject log
5477 if the ACL subsequently rejects the message.
5479 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5480 MD5 (which is deprecated).
5482 21. When testing a filter file using -bf, Exim was writing a message when it
5483 took the sender from a "From " line in the message, but it was not doing so
5484 when it took $return_path from a Return-Path: header line. It now does.
5486 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5487 with a valid header line field name (a series of printing characters
5488 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5490 23. Changed "disc" in the source to "disk" to conform to the documentation and
5491 the book and for uniformity.
5493 24. Ignore Sendmail's -Ooption=value command line item.
5495 25. When execve() failed while trying to run a command in a pipe transport,
5496 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5497 could be confused with a return value of 69 from the command itself. This
5498 has been changed to 127, the value the shell returns if it is asked to run
5499 a non-existent command. The wording for the related log line suggests a
5500 non-existent command as the problem.
5502 26. If received_header_text expands to an empty string, do not add a Received:
5503 header line to the message. (Well, it adds a token one on the spool, but
5504 marks it "old" so that it doesn't get used or transmitted.)
5506 27. Installed eximstats 1.28 (addition of -nt option).
5508 28. There was no check for failure on the call to getsockname() in the daemon
5509 code. This can fail if there is a shortage of resources on the system, with
5510 ENOMEM, for example. A temporary error is now given on failure.
5512 29. Contrary to the C standard, it seems that in some environments, the
5513 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5514 program. Exim now does this explicitly; it affects the formatting of
5515 timestamps using strftime().
5517 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5518 complaints. I've now initialized all the variables, to avoid this.
5520 32. Header lines added by a system filter were not being "seen" during
5521 transport-time rewrites.
5523 33. The info_callback() function passed to OpenSSL is set up with type void
5524 (*)(SSL *, int, int), as described somewhere. However, when calling the
5525 function (actually a macro) that sets it up, the type void(*)() is
5526 expected. I've put in a cast to prevent warnings from picky compilers.
5528 34. If a DNS black list lookup found a CNAME record, but there were no A
5529 records associated with the domain it pointed at, Exim crashed.
5531 35. If a DNS black list lookup returned more than one A record, Exim ignored
5532 all but the first. It now scans all returned addresses if a particular IP
5533 value is being sought. In this situation, the contents of the
5534 $dnslist_value variable are a list of all the addresses, separated by a
5537 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5538 to accept a host name and all its aliases if the forward lookup for any of
5539 them yielded the IP address of the incoming connection. Now it accepts only
5540 those names whose forward lookup yields the correct IP address. Any other
5541 names are discarded. This closes a loophole whereby a rogue DNS
5542 administrator could create reverse DNS records to break through a
5543 wildcarded host restriction in an ACL.
5545 37. If a user filter or a system filter that ran in a subprocess used any of
5546 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5547 the wrong values were passed to the pipe command ($thisaddress had the
5548 value of $0, $0 had the value of $1, etc). This bug was introduced by
5549 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5551 38. Improved the line breaking for long SMTP error messages from ACLs.
5552 Previously, if there was no break point between 40 and 75 characters, Exim
5553 left the rest of the message alone. Two changes have been made: (a) I've
5554 reduced the minimum length to 35 characters; (b) if it can't find a break
5555 point between 35 and 75 characters, it looks ahead and uses the first one
5556 that it finds. This may give the occasional overlong line, but at least the
5557 remaining text gets split now.
5559 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5560 file descriptors might be low, and that setting 1000 would always raise it.
5561 It turns out that in some environments, the limit is already over 1000 and
5562 that lowering it causes trouble. So now Exim takes care not to decrease it.
5564 40. When delivering a message, the value of $return_path is set to $sender_
5565 address at the start of routing (routers may change the value). By an
5566 oversight, this default was not being set up when an address was tested by
5567 -bt or -bv, which affected the outcome if any router or filter referred to
5570 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5571 log only when "message" or "log_message" are set. However, if one of the
5572 conditions was an address verification, or a call to a nested ACL, the
5573 messages generated by the underlying test were being passed through. This
5574 no longer happens. The underlying message is available in $acl_verify_
5575 message for both "message" and "log_message" expansions, so it can be
5576 passed through if needed.
5578 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5579 new expansion $bh_ to give the encoded byte string without charset
5580 translation. Translation happens only if iconv() is available; HAVE_ICONV
5581 indicates this at build time. HEADERS_CHARSET gives the charset to
5582 translate to; headers_charset can change it in the configuration, and
5583 "headers charset" can change it in an individual filter file.
5585 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5586 that creates RFC 2047 encoded "words" labels them as that charset instead
5587 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5588 expansion operator; (ii) when Exim creates a From: line for a local
5589 message; (iii) when a header line is rewritten to include a "phrase" part.
5591 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5592 buggy, causing it to skip the first lines of messages whose message ID
5593 ended in 'D'. This would not have bitten before Exim release 4.14, because
5594 message IDs were unlikely to end in 'D' before then. The effect was to have
5595 incorrect size information for certain domains.
5597 45. #include "config.h" was missing at the start of the crypt16.c module. This
5598 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5601 46. If there was a timeout during a "random" callout check, Exim treated it as
5602 a failure of the random address, and carried on sending RSET and the real
5603 address. If the delay was just some slowness somewhere, the response to the
5604 original RCPT would be taken as a response to RSET and so on, causing
5605 mayhem of various kinds.
5607 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5608 when I implemented it. It didn't allow for the fact that some option values
5609 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5610 the right test for positive values.
5612 48. Domain names in DNS records are case-independent. Exim always looks them up
5613 in lower case. Some resolvers return domain names in exactly the case they
5614 appear in the zone file, that is, they may contain uppercase letters. Not
5615 all resolvers do this - some return always lower case. Exim was treating a
5616 change of case by a resolver as a change of domain, similar to a widening
5617 of a domain abbreviation. This triggered its re-routing code and so it was
5618 trying to route what was effectively the same domain again. This normally
5619 caused routing to fail (because the router wouldn't handle the domain
5620 twice). Now Exim checks for this case specially, and just changes the
5621 casing of the domain that it ultimately uses when it transmits the message
5624 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5627 50. If a filter generated a file delivery with a non-absolute name (possible if
5628 no home directory exists for the router), the forbid_file option was not
5631 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5632 the existing equality matching.
5634 52. Exim was using ints instead of ino_t variables in some places where it was
5635 dealing with inode numbers.
5637 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5638 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5639 called TMPDIR, and if it finds it is different, it changes its value.
5641 54. The smtp_printf() function is now made available to local_scan() so
5642 additional output lines can be written before returning. There is also an
5643 smtp_fflush() function to enable the detection of a dropped connection.
5644 The variables smtp_input and smtp_batched_input are exported to
5647 55. Changed the default runtime configuration: the message "Unknown user"
5648 has been removed from the ACL, and instead placed on the localuser router,
5649 using the cannot_route_message feature. This means that any verification
5650 failures that generate their own messages won't get overridden. Similarly,
5651 the "Unrouteable address" message that was in the ACL for unverifiable
5652 relay addresses has also been removed.
5654 56. Added hosts_avoid_esmtp to the smtp transport.
5656 57. The exicyclog script was not checking for the esoteric option
5657 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5658 will work only if exicyclog is run under the appropriate euid.
5660 58. Following a discussion on the list, the rules by which Exim recognises line
5661 endings on incoming messages have been changed. The -dropcr and drop_cr
5662 options are now no-ops, retained only for backwards compatibility. The
5663 following line terminators are recognized: LF CRLF CR. However, special
5664 processing applies to CR:
5666 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5667 nor a local message in the state where . is a terminator.
5669 (ii) If a bare CR is encountered in a header line, an extra space is added
5670 after the line terminator so as not to end the header. The reasoning
5671 behind this is that bare CRs in header lines are most likely either
5672 to be mistakes, or people trying to play silly games.
5674 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5675 was being incorrectly given as 18 bytes larger than it should have been.
5676 This is a VOB (very old bug).
5678 60. This may never have affected anything current, but just in case it has:
5679 When the local host is found other than at the start of a list of hosts,
5680 the local host, those with the same MX, and any that follow, are discarded.
5681 When the list in question was part of a longer list of hosts, the following
5682 hosts (not currently being processed) were also being discarded. This no
5683 longer happens. I'm not sure if this situation could ever has previously
5686 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5689 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5690 header. This is something that is recommended in a new Internet Draft, and
5691 is something that is documented as being done by Sendmail. There are two
5692 possible values. For messages generated by the autoreply transport, Exim
5695 Auto-Submitted: auto-replied
5697 whereas for all other generated messages (e.g. bounces) it adds
5699 Auto-Submitted: auto-generated
5701 63. The "personal" condition in filters now includes a test for the
5702 Auto-Submitted: header. If it contains the string "auto-" the message it
5703 not considered personal.
5705 64. Added rcpt_include_affixes as a generic transport option.
5707 65. Added queue_only_override (default true).
5709 66. Added the syslog_duplication option.
5711 67. If what should have been the first header line of a message consisted of
5712 a space followed by a colon, Exim was mis-interpreting it as a header line.
5713 It isn't of course - it is syntactically invalid and should therefore be
5714 treated as the start of the message body. The misbehaviour could have
5715 caused a number of strange effects, including loss of data in subsequent
5716 header lines, and spool format errors.
5718 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5719 client host had authenticated. This control can now be exercised by an ACL
5720 for more flexibility.
5722 69. By default, callouts do not happen when testing with -bh. There is now a
5723 variant, -bhc, which does actually run the callout code, including
5724 consulting and updating the callout cache.
5726 70. Added support for saslauthd authentication, courtesy of Alexander
5729 71. If statvfs() failed on the spool or log directories while checking their
5730 size for availability, Exim confusingly gave the error "space shortage".
5731 Furthermore, in debugging mode it crashed with a floating point exception.
5732 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5733 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5734 serious problem, Exim now writes to the main and panic logs when this
5735 happens, with details of the failure. It then refuses to accept the
5736 incoming message, giving the message "spool directory problem" or "log
5737 directory problem" with a 421 code for SMTP messages.
5739 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5740 0, 1, and 2 exist, because some OS complain for execs without them (see
5741 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5742 descriptors. However, the code omitted to check that the open succeeded,
5743 causing mysterious errors if for some reason the permissions on /dev/null
5744 got screwed. Now Exim writes a message to the main and panic logs, and
5745 bombs out if it can't open /dev/null.
5747 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5748 interact so that it is all more flexible. It is supposed to remain
5749 backwards compatible. Also added extra_local_interfaces.
5751 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5752 to bomb out with an assertion failure - to the client this appears as a
5753 connection drop. This problem occurs in the part of the code that was taken
5754 from the Samba project. Fortunately, the assertion is in a very simple
5755 function, so I have fixed this by reproducing the function inline in the
5756 one place where it is called, and arranging for authentication to fail
5757 instead of killing the process with assert().
5759 75. The SPA client code was not working when the server requested OEM rather
5760 than Unicode encoding.
5762 76. Added code to make require_files with a specific uid setting more usable in
5763 the case where statting the file as root fails - usually a non-root-mounted
5764 NFS file system. When this happens and the failure is EACCES, Exim now
5765 forks a subprocess and does the per-uid checking as the relevant uid.
5767 77. Added process_log_path.
5769 78. If log_file_path was not explicitly set, a setting of check_log_space or
5770 check_log_inodes was ignored.
5772 79. If a space check for the spool or log partitions fails, the incident is now
5773 logged. Of course, in the latter case the data may get lost...
5775 80. Added the %p formatting code to string_format() so that it can be used to
5776 print addresses in debug_print(). Adjusted all the address printing in the
5777 debugging in store.c to use %p rather than %d.
5779 81. There was a concern that a line of code in smtp_in.c could overflow a
5780 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5781 initially expressed, the concern was not well-founded, because trailing
5782 spaces are removed early. However, if the trailing spaces were followed by
5783 a NULL, they did not get removed, so the overflow was possible. Two fixes
5786 (a) I re-wrote the offending code in a cleaner fashion.
5787 (b) If an incoming SMTP command contains a NULL character, it is rejected
5790 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5791 runs initgroups(), so that if the Exim user is in any additional groups,
5792 they will be used during message reception.
5798 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob