1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
9 PP/02 Make -n do something, by making it not do something.
10 When combined with -bP, the name of an option is not output.
12 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
15 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
16 $sender_host_name and config options to manage this, and basic check
19 PP/05 DSCP support for outbound connections and control modifier for inbound.
21 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
22 (Only plugin which currently uses this is kerberos4, which nobody should
23 be using, but we should make it available and other future plugins might
24 conceivably use it, even though it would break NAT; stuff *should* be
25 using channel bindings instead).
27 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
28 name; added for Sendmail compatibility; requires admin caller.
29 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
30 just ignore it); requires trusted caller.
31 Also parse but ignore: -Ac -Am -X<logfile>
34 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
36 TL/02 Add +smtp_confirmation as a default logging option.
38 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
39 Patch by Magnus Holmgren from 2007-02-20.
41 TL/04 Bugzilla 1281 - Spec typo.
42 Bugzilla 1283 - Spec typo.
43 Bugzilla 1290 - Spec grammar fixes.
45 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
47 TL/06 Add Experimental DMARC support using libopendmarc libraries.
49 TL/07 Fix an out of order global option causing a segfault. Reported to dev
50 mailing list by by Dmitry Isaikin.
52 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
54 JH/02 Support "G" suffix to numbers in ${if comparisons.
56 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
58 NM/01 Bugzilla 1197 - Spec typo
59 Bugzilla 1196 - Spec examples corrections
61 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
63 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
64 gnutls_enable_pkcs11, but renamed to more accurately indicate its
67 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
68 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
70 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
71 "acl {{name}{arg}...}", and optional args on acl condition
74 JH/05 Permit multiple router/transport headers_add/remove lines.
76 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
78 JH/07 Avoid using a waiting database for a single-message-only transport.
79 Performance patch from Paul Fisher. Bugzilla 1262.
81 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
84 JH/09 Add $headers_added variable, with content from use of ACL modifier
85 add_header (but not yet added to the message). Bugzilla 199.
87 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
88 Pulled from Bugzilla 817 by Wolfgang Breyha.
90 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
92 (nb: this is the same fix as in Exim 4.80.1)
94 JH/11 Add A= logging on delivery lines, and a client_set_id option on
97 JH/12 Add optional authenticated_sender logging to A= and a log_selector
100 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
102 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
103 advertise SMTP AUTH mechanism to us, instead of a generic
104 protocol violation error. Also, make Exim more robust to bad
105 data from the Dovecot auth socket.
107 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
109 When a queue runner is handling a message, Exim first routes the
110 recipient addresses, during which it prunes them based on the retry
111 hints database. After that it attempts to deliver the message to
112 any remaining recipients. It then updates the hints database using
115 So if a recipient address works intermittently, it can get repeatedly
116 deferred at routing time. The retry hints record remains fresh so the
117 address never reaches the final cutoff time.
119 This is a fairly common occurrence when a user is bumping up against
120 their storage quota. Exim had some logic in its local delivery code
121 to deal with this. However it did not apply to per-recipient defers
122 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
124 This change adds a proper retry rule check during routing so that the
125 final cutoff time is checked against the message's age. We only do
126 this check if there is an address retry record and there is not a
127 domain retry record; this implies that previous attempts to handle
128 the address had the retry_use_local_parts option turned on. We use
129 this as an approximation for the destination being like a local
130 delivery, as in LMTP.
132 I suspect this new check makes the old local delivery cutoff check
133 redundant, but I have not verified this so I left the code in place.
135 TF/02 Correct gecos expansion when From: is a prefix of the username.
137 Test 0254 submits a message to Exim with the header
141 When I ran the test suite under the user fanf2, Exim expanded
142 the header to contain my full name, whereas it should have added
143 a Resent-Sender: header. It erroneously treats any prefix of the
144 username as equal to the username.
146 This change corrects that bug.
148 GF/01 DCC debug and logging tidyup
149 Error conditions log to paniclog rather than rejectlog.
150 Debug lines prefixed by "DCC: " to remove any ambiguity.
152 TF/03 Avoid unnecessary rebuilds of lookup-related code.
154 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
155 Bug spotted by Jeremy Harris; was flawed since initial commit.
156 Would have resulted in OCSP responses post-SNI triggering an Exim
157 NULL dereference and crash.
159 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
161 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
162 Bug detection, analysis and fix by Samuel Thibault.
163 Bugzilla 1331, Debian bug #698092.
165 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
167 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
168 Server implementation by Todd Lyons, client by JH.
169 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
170 config variable "prdr_enable" controls whether the server
171 advertises the facility. If the client requests PRDR a new
172 acl_data_smtp_prdr ACL is called once for each recipient, after
173 the body content is received and before the acl_smtp_data ACL.
174 The client is controlled by bolth of: a hosts_try_prdr option
175 on the smtp transport, and the server advertisement.
176 Default client logging of deliveries and rejections involving
177 PRDR are flagged with the string "PRDR".
179 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
180 fclose(). Diagnosis by Todd Lyons.
182 PP/17 Update configure.default to handle IPv6 localhost better.
183 Patch by Alain Williams (plus minor tweaks).
186 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
187 This is now consistent with GnuTLS, and is now documented: the
188 previous undocumented portable approach to treating the option as
189 unset was to force an expansion failure. That still works, and
190 an empty string is now equivalent.
192 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
193 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
194 not performing validation itself.
196 PP/20 Added force_command boolean option to pipe transport.
197 Patch from Nick Koston, of cPanel Inc.
199 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
202 TF/04 Added udpsend ACL modifer and hexquote expansion operator
204 PP/21 Fix eximon continuous updating with timestamped log-files.
205 Broken in a format-string cleanup in 4.80, missed when I repaired the
206 other false fix of the same issue.
207 Report and fix from Heiko Schlichting.
210 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
211 Report from Prashanth Katuri.
213 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
214 It's SecureTransport, so affects any MacOS clients which use the
215 system-integrated TLS libraries, including email clients.
217 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
218 using a MIME ACL for non-SMTP local injection.
219 Report and assistance in diagnosis by Warren Baker.
221 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
223 JH/16 Fix comparisons for 64b. Bugzilla 1385.
225 TL/09 Add expansion variable $authenticated_fail_id to keep track of
226 last id that failed so it may be referenced in subsequent ACL's.
228 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
231 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
232 ldap library initialization, allowing self-signed CA's to be
233 used. Also properly sets require_cert option later in code by
234 using NULL (global ldap config) instead of ldap handle (per
235 session). Bug diagnosis and testing by alxgomz.
237 TL/12 Enhanced documentation in the ratelimit.pl script provided in
238 the src/util/ subdirectory.
240 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
241 renamed to Transport Post Delivery Action by Jeremy Harris, as
244 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
245 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
246 redis_servers = needs to be configured which will be used by the redis
247 lookup. Patch from Warren Baker, of The Packet Hub.
249 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
251 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
252 hostname or reverse DNS when processing a host list. Used suggestions
253 from multiple comments on this bug.
255 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
257 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
258 Missed a few lines, added it to make the runtest require no keyboard
261 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
262 contains upper case chars. Make router use caseful_local_part.
264 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
265 support when GnuTLS has been built with p11-kit.
271 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
273 This, or similar/improved, will also be change PP/11 of 4.82.
279 PP/01 Handle short writes when writing local log-files.
280 In practice, only affects FreeBSD (8 onwards).
281 Bugzilla 1053, with thanks to Dmitry Isaikin.
283 NM/01 Bugzilla 949 - Documentation tweak
285 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
288 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
290 PP/02 Implemented gsasl authenticator.
292 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
294 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
295 `pkg-config foo` for cflags/libs.
297 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
298 with rest of GSASL and with heimdal_gssapi.
300 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
301 `pkg-config foo` for cflags/libs for the TLS implementation.
303 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
304 properties get this fed in as external SSF. A number of robustness
305 and debugging improvements to the cyrus_sasl authenticator.
307 PP/08 cyrus_sasl server now expands the server_realm option.
309 PP/09 Bugzilla 1214 - Log authentication information in reject log.
310 Patch by Jeremy Harris.
312 PP/10 Added dbmjz lookup type.
314 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
316 PP/12 MAIL args handles TAB as well as SP, for better interop with
317 non-compliant senders.
318 Analysis and variant patch by Todd Lyons.
320 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
321 Bug report from Lars Müller <lars@samba.org> (via SUSE),
322 Patch from Dirk Mueller <dmueller@suse.com>
324 PP/13 tls_peerdn now print-escaped for spool files.
325 Observed some $tls_peerdn in wild which contained \n, which resulted
326 in spool file corruption.
328 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
329 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
330 or write after TLS renegotiation, which otherwise led to messages
333 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
334 as a tracking header (ie: a signed header comes before the signature).
335 Patch from Wolfgang Breyha.
337 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
338 comma-sep list; embedded commas doubled.
340 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
342 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
344 Report and patch from Dmitry Banschikov.
346 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
347 Removed SSL_clear() after SSL_new() which led to protocol negotiation
348 failures. We appear to now support TLS1.1+ with Exim.
350 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
351 lets Exim select keys and certificates based upon TLS SNI from client.
352 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
353 before an outbound SMTP session. New log_selector, +tls_sni.
355 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
356 NULL dereference. Report and patch from Alun Jones.
358 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
359 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
360 Not seeing resolver debug output on NetBSD, but suspect this is a
361 resolver implementation change.
363 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
364 Left warnings. Added "eximon gdb" invocation mode.
366 PP/21 Defaulting "accept_8bitmime" to true, not false.
368 PP/22 Added -bw for inetd wait mode support.
370 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
371 locate the relevant includes and libraries. Made this the default.
373 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
374 Bugzilla 1246, report and most of solution from Tomasz Kusy.
376 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
377 This may cause build issues on older platforms.
379 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
380 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
381 gnutls_require_mac & gnutls_require_protocols (no longer supported).
382 Added SNI support via GnuTLS too.
383 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
385 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
387 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
388 Applied second patch from Janne, implementing suggestion to default
389 multiple-strings-in-record handling to match SPF spec.
391 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
393 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
394 read-only, out of scope).
395 Patch from Wolfgang Breyha, report from Stuart Northfield.
397 PP/29 Fix three issues highlighted by clang analyser static analysis.
398 Only crash-plausible issue would require the Cambridge-specific
399 iplookup router and a misconfiguration.
400 Report from Marcin Mirosław.
402 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
404 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
405 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
406 As part of this, removing so much warning spew let me fix some minor
407 real issues in debug logging.
409 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
410 assignment on my part. Fixed.
412 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
413 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
414 Janne Snabb (who went above and beyond: thank you).
416 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
417 string otherwise requires a connection and a bunch more work and it's
418 relatively easy to get wrong. Should also expose TLS library linkage
421 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
422 64-bit ${eval} (JH/03).
424 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
425 GNU libc to support some of the 64-bit stuff, should not lead to
426 conflicts. Defined before os.h is pulled in, so if a given platform
427 needs to override this, it can.
429 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
430 protection layer was required, which is not implemented.
431 Bugzilla 1254, patch from Wolfgang Breyha.
433 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
434 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
435 tls_dhparam take prime identifiers. Also unbreak combination of
436 OpenSSL+DH_params+TLSSNI.
438 PP/39 Disable SSLv2 by default in OpenSSL support.
444 PP/01 Solaris build fix for Oracle's LDAP libraries.
445 Bugzilla 1109, patch from Stephen Usher.
447 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
449 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
452 TF/02 Fix a couple more cases where we did not log the error message
453 when unlink() failed. See also change 4.74-TF/03.
455 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
456 lock up or crash if it happened to be inside a call to libc when it
457 got a SIGUSR1 from exiwhat.
459 The SIGUSR1 handler appends the current process status to the process
460 log which is later printed by exiwhat. It used to use the general
461 purpose logging code to do this, but several functions it calls are
462 not safe for signals.
464 The new output code in the SIGUSR1 handler is specific to the process
465 log, and simple enough that it's easy to inspect for signal safety.
466 Removing some special cases also simplifies the general logging code.
467 Removing the spurious timestamps from the process log simplifies
470 TF/04 Improved ratelimit ACL condition.
472 The /noupdate option has been deprecated in favour of /readonly which
473 has clearer semantics. The /leaky, /strict, and /readonly update modes
474 are mutually exclusive. The update mode is no longer included in the
475 database key; it just determines when the database is updated. (This
476 means that when you upgrde Exim will forget old rate measurements.)
478 Exim now checks that the per_* options are used with an update mode that
479 makes sense for the current ACL. For example, when Exim is processing a
480 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
481 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
482 must specify per_mail/readonly. If you omit the update mode it defaults to
483 /leaky where that makes sense (as before) or /readonly where required.
485 The /noupdate option is now undocumented but still supported for
486 backwards compatibility. It is equivalent to /readonly except that in
487 ACLs where /readonly is required you may specify /leaky/noupdate or
488 /strict/noupdate which are treated the same as /readonly.
490 A useful new feature is the /count= option. This is a generalization
491 of the per_byte option, so that you can measure the throughput of other
492 aggregate values. For example, the per_byte option is now equivalent
493 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
495 The per_rcpt option has been generalized using the /count= mechanism
496 (though it's more complicated than the per_byte equivalence). When it is
497 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
498 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
499 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
500 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
501 Note that using per_rcpt with a non-readonly update mode in more than
502 one ACL will cause the recipients to be double-counted. (The per_mail
503 and per_byte options don't have this problem.)
505 The handling of very low rates has changed slightly. If the computed rate
506 is less than the event's count (usually one) then this event is the first
507 after a long gap. In this case the rate is set to the same as this event's
508 count, so that the first message of a spam run is counted properly.
510 The major new feature is a mechanism for counting the rate of unique
511 events. The new per_addr option counts the number of different
512 recipients that someone has sent messages to in the last time period. It
513 behaves like per_rcpt if all the recipient addresses are different, but
514 duplicate recipient addresses do not increase the measured rate. Like
515 the /count= option this is a general mechanism, so the per_addr option
516 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
517 example, measure the rate that a client uses different sender addresses
518 with the options per_mail/unique=$sender_address. There are further
519 details in the main documentation.
521 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
523 TF/06 Removed a few PCRE remnants.
525 TF/07 Automatically extract Exim's version number from tags in the git
526 repository when doing development or release builds.
528 PP/02 Raise smtp_cmd_buffer_size to 16kB.
529 Bugzilla 879. Patch from Paul Fisher.
531 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
532 Heavily based on revision 40f9a89a from Simon Arlott's tree.
535 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
537 PP/05 Variable $av_failed, true if the AV scanner deferred.
538 Bugzilla 1078. Patch from John Horne.
540 PP/06 Stop make process more reliably on build failure.
541 Bugzilla 1087. Patch from Heiko Schlittermann.
543 PP/07 Make maildir_use_size_file an _expandable_ boolean.
544 Bugzilla 1089. Patch from Heiko Schlittermann.
546 PP/08 Handle ${run} returning more data than OS pipe buffer size.
547 Bugzilla 1131. Patch from Holger Weiß.
549 PP/09 Handle IPv6 addresses with SPF.
550 Bugzilla 860. Patch from Wolfgang Breyha.
552 PP/10 GnuTLS: support TLS 1.2 & 1.1.
554 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
557 PP/11 match_* no longer expand right-hand-side by default.
558 New compile-time build option, EXPAND_LISTMATCH_RHS.
559 New expansion conditions, "inlist", "inlisti".
561 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
563 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
565 PP/14 fix log_write() format string regression from TF/03.
566 Bugzilla 1152. Patch from Dmitry Isaikin.
572 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
574 PP/02 Harmonised TLS library version reporting; only show if debugging.
575 Layout now matches that introduced for other libraries in 4.74 PP/03.
577 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
579 PP/04 New "dns_use_edns0" global option.
581 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
584 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
585 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
587 TK/01 Updated PolarSSL code to 0.14.2.
588 Bugzilla 1097. Patch from Andreas Metzler.
590 PP/07 Catch divide-by-zero in ${eval:...}.
593 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
596 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
597 format-string attack -- SECURITY: remote arbitrary code execution.
599 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
600 time unintentionally subject to list matching rules, letting the header
601 cause arbitrary Exim lookups (of items which can occur in lists, *not*
602 arbitrary string expansion). This allowed for information disclosure.
604 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
605 INT_MIN/-1 -- value coerced to INT_MAX.
611 NM/01 Workround for PCRE version dependency in version reporting
614 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
615 This fixes portability to compilers other than gcc, notably
616 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
618 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
619 makefiles for portability to HP-UX and POSIX correctness.
621 PP/01 Permit LOOKUP_foo enabling on the make command-line.
622 Also via indented variable definition in the Makefile.
623 (Debugging by Oliver Heesakkers).
625 PP/02 Restore caching of spamd results with expanded spamd_address.
626 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
628 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
629 Improves build reliability. Fix from: Frank Elsner
631 NM/02 Fix wide character breakage in the rfc2047 coding
632 Fixes bug 1064. Patch from Andrey N. Oktyabrski
634 NM/03 Allow underscore in dnslist lookups
635 Fixes bug 1026. Patch from Graeme Fowler
637 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
638 Code patches from Adam Ciarcinski of NetBSD.
640 NM/04 Fixed exiqgrep to cope with mailq missing size issue
643 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
644 is logged, to avoid truncation. Patch from John Horne.
646 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
647 Patch from Jakob Hirsch.
649 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
650 SQL string expansion failure details.
651 Patch from Andrey Oktyabrski.
653 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
654 Patch from Simon Arlott.
656 PP/09 New lookups functionality failed to compile on old gcc which rejects
657 extern declarations in function scope.
658 Patch from Oliver Fleischmann
660 PP/10 Use sig_atomic_t for flags set from signal handlers.
661 Check getgroups() return and improve debugging.
662 Fixed developed for diagnosis in bug 927 (which turned out to be
665 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
666 Patch from Mark Zealey.
668 PP/12 Bugzilla 1056: Improved spamd server selection.
669 Patch from Mark Zealey.
671 PP/13 Bugzilla 1086: Deal with maildir quota file races.
672 Based on patch from Heiko Schlittermann.
674 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
675 Patch from Uwe Doering, sign-off by Michael Haardt.
677 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
678 variable declaration deep within a block. Bug and patch from
681 PP/15 lookups-Makefile IRIX compatibilty coercion.
683 PP/16 Make DISABLE_DKIM build knob functional.
685 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
686 Patch by Simon Arlott
688 TF/03 Fix valgrind.h portability to C89 compilers that do not support
689 variable argument macros. Our copy now differs from upstream.
695 TF/01 Failure to get a lock on a hints database can have serious
696 consequences so log it to the panic log.
698 TF/02 Log LMTP confirmation messages in the same way as SMTP,
699 controlled using the smtp_confirmation log selector.
701 TF/03 Include the error message when we fail to unlink a spool file.
703 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
704 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
705 for maintaining out-of-tree patches for some time.
707 PP/01 Bugzilla 139: Documentation and portability issues.
708 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
709 Handle per-OS dynamic-module compilation flags.
711 PP/02 Let /dev/null have normal permissions.
712 The 4.73 fixes were a little too stringent and complained about the
713 permissions on /dev/null. Exempt it from some checks.
714 Reported by Andreas M. Kirchwitz.
716 PP/03 Report version information for many libraries, including
717 Exim version information for dynamically loaded libraries. Created
718 version.h, now support a version extension string for distributors
719 who patch heavily. Dynamic module ABI change.
721 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
722 privilege escalation vulnerability whereby the Exim run-time user
723 can cause root to append content of the attacker's choosing to
726 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
729 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
730 If dropping privileges for untrusted macros, we disabled normal logging
731 on the basis that it would fail; for the Exim run-time user, this is not
732 the case, and it resulted in successful deliveries going unlogged.
733 Fixed. Reported by Andreas Metzler.
739 PP/01 Date: & Message-Id: revert to normally being appended to a message,
740 only prepend for the Resent-* case. Fixes regression introduced in
741 Exim 4.70 by NM/22 for Bugzilla 607.
743 PP/02 Include check_rfc2047_length in configure.default because we're seeing
744 increasing numbers of administrators be bitten by this.
746 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
748 PP/03 Bugzilla 994: added openssl_options main configuration option.
750 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
752 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
754 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
756 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
757 without a peer certificate, leading to a segfault because of an
758 assumption that peers always have certificates. Be a little more
759 paranoid. Problem reported by Martin Tscholak.
761 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
762 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
763 NB: ClamAV planning to remove STREAM in "middle of 2010".
764 CL also introduces -bmalware, various -d+acl logging additions and
765 more caution in buffer sizes.
767 PP/09 Implemented reverse_ip expansion operator.
769 PP/10 Bugzilla 937: provide a "debug" ACL control.
771 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
773 PP/12 Bugzilla 973: Implement --version.
775 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
777 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
779 PP/15 Bugzilla 816: support multiple condition rules on Routers.
781 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
782 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
783 ignore trailing whitespace.
785 JJ/02 prevent non-panic DKIM error from being sent to paniclog
787 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
790 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
791 Notification from Dr Andrew Aitchison.
793 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
794 ExtendedDetectionInfo response format.
795 Notification from John Horne.
797 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
800 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
801 XSL and documented dependency on system catalogs, with examples of how
804 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
807 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
808 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
809 configuration file which is writeable by the Exim user or group.
811 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
812 of configuration files to cover files specified with the -C option if
813 they are going to be used with root privileges, not just the default
816 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
817 option (effectively making it always true).
819 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
820 files to be used while preserving root privileges.
822 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
823 that rogue child processes cannot use them.
825 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
826 run-time user, instead of root.
828 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
829 Exim run-time user without dropping privileges.
831 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
832 result string, instead of calling string_vformat() twice with the same
835 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
836 for other users. Others should always drop root privileges if they use
837 -C on the command line, even for a whitelisted configure file.
839 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
841 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
847 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
848 $data_path, and $header_path variables; fixed documentation bugs and
851 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
852 exipick to access non-standard spools, including the "frozen" queue
855 NM/01 Bugzilla 965: Support mysql stored procedures.
856 Patch from Alain Williams
858 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
860 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
861 Patch from Andreas Metzler
863 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
864 Patch from Kirill Miazine
866 NM/05 Bugzilla 671: Added umask to procmail example.
868 JJ/03 installed exipick 20100323.0, fixing doc bug
870 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
871 directory. Notification and patch from Dan Rosenberg.
873 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
875 TK/02 Improve log output when DKIM signing operation fails.
877 MH/01 Treat the transport option dkim_domain as a colon separated
878 list, not as a single string, and sign the message with each element,
879 omitting multiple occurences of the same signer.
881 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
882 Bugzilla 985, 986. Patch by Simon Arlott
884 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
885 Patch by Simon Arlott
887 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
888 MBX locking. Notification from Dan Rosenberg.
894 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
896 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
898 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
900 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
902 NM/04 Bugzilla 915: Flag broken perl installation during build.
908 TK/01 Added patch by Johannes Berg that expands the main option
909 "spamd_address" if it starts with a dollar sign.
911 TK/02 Write list of recipients to X-Envelope-Sender header when building
912 the mbox-format spool file for content scanning (suggested by Jakob
915 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
916 (http://www.dcc-servers.net/) support via dccifd. Activated by
917 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
919 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
920 by Mark Daniel Reidel <mr@df.eu>.
922 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
923 When building exim an external PCRE library is now needed -
924 PCRE is a system library on the majority of modern systems.
925 See entry on PCRE_LIBS in EDITME file.
927 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
928 conversation. Added nologin parameter to request.
929 Patch contributed by Kirill Miazine.
931 TF/01 Do not log submission mode rewrites if they do not change the address.
933 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
935 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
936 log files in place. Contributed by Roberto Lima.
938 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
940 TF/03 Bugzilla 615: When checking the local_parts router precondition
941 after a local_part_suffix or local_part_prefix option, Exim now
942 does not use the address's named list lookup cache, since this
943 contains cached lookups for the whole local part.
945 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
946 Robert Millan. Documentation is in experimental-spec.txt.
948 TF/04 Bugzilla 668: Fix parallel build (make -j).
950 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
952 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
953 Patch provided by Jan Srzednicki.
955 TF/05 Leading white space used to be stripped from $spam_report which
956 wrecked the formatting. Now it is preserved.
958 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
959 that they are available at delivery time.
961 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
963 TF/08 TLS error reporting now respects the incoming_interface and
964 incoming_port log selectors.
966 TF/09 Produce a more useful error message if an SMTP transport's hosts
967 setting expands to an empty string.
969 NM/06 Bugzilla 744: EXPN did not work under TLS.
970 Patch provided by Phil Pennock.
972 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
973 Patch provided by Richard Godbee.
975 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
976 acl_smtp_notquit, added index entry.
978 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
979 Patch provided by Eugene Bujak.
981 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
982 accept(). Patch provided by Maxim Dounin.
984 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
985 Patch provided by Phil Pennock.
987 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
989 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
990 Patch provided by Brad "anomie" Jorsch.
992 NM/14 Bugzilla 622: Added timeout setting to transport filter.
993 Patch provided by Dean Brooks.
995 TK/05 Add native DKIM support (does not depend on external libraries).
997 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
998 Patch provided by Graeme Fowler.
1000 NM/16 Bugzilla 851: Documentation example syntax fix.
1002 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1004 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1007 NM/19 Bugzilla 745: TLS version reporting.
1008 Patch provided by Phil Pennock.
1010 NM/20 Bugzilla 167: bool: condition support.
1011 Patch provided by Phil Pennock.
1013 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1014 clients. Patch provided by Phil Pennock.
1016 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1017 Patch provided by Brad "anomie" Jorsch.
1019 NM/23 Bugzilla 687: Fix misparses in eximstats.
1020 Patch provided by Heiko Schlittermann.
1022 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1023 Patch provided by Heiko Schlittermann.
1025 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1026 plus update to original patch.
1028 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1030 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1031 Patch provided by David Brownlee.
1033 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1035 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1037 NM/30 Bugzilla 888: TLS documentation bugfixes.
1039 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1041 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1042 Unlike the original bugzilla I have changed all shell scripts in src tree.
1044 NM/33 Bugzilla 898: Transport filter timeout fix.
1045 Patch by Todd Rinaldo.
1047 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1048 Patch by Serge Demonchaux.
1050 NM/35 Bugzilla 39: Base64 decode bug fixes.
1051 Patch by Jakob Hirsch.
1053 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1055 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1057 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1059 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1065 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1066 ALT-N's libdkim that I have put here:
1067 http://duncanthrax.net/exim-experimental/
1069 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1070 were called 'true' and it seems that C99 defines that as a reserved
1071 keyword to be used with 'bool' variable types. That means you could
1072 not include C99-style headers which use bools without triggering
1073 build errors in sieve.c.
1075 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1076 as mailq or other aliases. Changed the --help handling significantly
1077 to do whats expected. exim_usage() emits usage/help information.
1079 SC/01 Added the -bylocaldomain option to eximstats.
1081 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1083 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1085 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1091 PH/01 Another patch from the Sieve maintainer.
1093 PH/02 When an IPv6 address is converted to a string for single-key lookup
1094 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1095 dots are used instead of colons so that keys in lsearch files need not
1096 contain colons. This was done some time before quoting was made available
1097 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1098 (notated using the quote facility) so as to distinguish them from IPv4
1099 keys. This meant that lookups for IP addresses in host lists did not work
1100 for iplsearch lookups.
1102 This has been fixed by arranging for IPv6 addresses to be expressed with
1103 colons if the lookup type is iplsearch. This is not incompatible, because
1104 previously such lookups could never work.
1106 The situation is now rather anomolous, since one *can* have colons in
1107 ordinary lsearch keys. However, making the change in all cases is
1108 incompatible and would probably break a number of configurations.
1110 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1113 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1114 conversion specification without a maximum field width, thereby enabling
1115 a rogue spamd server to cause a buffer overflow. While nobody in their
1116 right mind would setup Exim to query an untrusted spamd server, an
1117 attacker that gains access to a server running spamd could potentially
1118 exploit this vulnerability to run arbitrary code as the Exim user.
1120 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1121 $primary_hostname instead of what libspf2 thinks the hosts name is.
1123 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1124 a directory entry by the name of the lookup key. Previously, if a
1125 symlink pointed to a non-existing file or a file in a directory that
1126 Exim lacked permissions to read, a lookup for a key matching that
1127 symlink would fail. Now it is enough that a matching directory entry
1128 exists, symlink or not. (Bugzilla 503.)
1130 PH/03 The body_linecount and body_zerocount variables are now exported in the
1133 PH/04 Added the $dnslist_matched variable.
1135 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1136 This means they are set thereafter only if the connection becomes
1139 PH/06 Added the client_condition to authenticators so that some can be skipped
1140 by clients under certain conditions.
1142 PH/07 The error message for a badly-placed control=no_multiline_responses left
1143 "_responses" off the end of the name.
1145 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1147 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1148 (without spaces) instead of just copying the configuration text.
1150 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1152 PH/11 Added $max_received_linelength.
1154 PH/12 Added +ignore_defer and +include_defer to host lists.
1156 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1157 way in which PCRE > 7.0 is built.
1159 PH/14 Implemented queue_only_load_latch.
1161 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1162 MAIL command. The effect was to mangle the value on 64-bit systems.
1164 PH/16 Another patch from the Sieve maintainer.
1166 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1168 PH/18 If a system quota error occurred while trying to create the file for
1169 a maildir delivery, the message "Mailbox is full" was not appended to the
1170 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1171 only to a quota excession during the actual writing of the file.
1173 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1174 characters?) which causes problems in log lines. The DN values are now
1175 passed through string_printing() before being added to log lines.
1177 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1178 and InterBase are left for another time.)
1180 PH/21 Added message_body_newlines option.
1182 PH/22 Guard against possible overflow in moan_check_errorcopy().
1184 PH/23 POSIX allows open() to be a macro; guard against that.
1186 PH/24 If the recipient of an error message contained an @ in the local part
1187 (suitably quoted, of course), incorrect values were put in $domain and
1188 $local_part during the evaluation of errors_copy.
1194 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1195 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1198 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1199 issue a MAIL command.
1201 PH/02 In an ACL statement such as
1203 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1205 if a client was not listed at all, or was listed with a value other than
1206 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1207 the condition was not true (as it should be), so access was not denied.
1208 The bug was that the ! inversion was incorrectly passed on to the second
1209 item. This has been fixed.
1211 PH/03 Added additional dnslists conditions == and =& which are different from
1212 = and & when the dns lookup returns more than one IP address.
1214 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1215 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1217 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1218 FSYNC, which compiles an option called disable_fsync that allows for
1219 bypassing fsync(). The documentation is heavily laced with warnings.
1221 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1223 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1224 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1225 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1226 including adding "make clean"; (3) Added -fPIC when compiling the test
1227 dynamically loaded module, to get rid of a warning.
1229 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1230 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1231 = 0s. The bug is otherwise harmless.
1233 PH/07 There was a bug in the dovecot authenticator such that the value of
1234 $auth1 could be overwritten, and so not correctly preserved, after a
1235 successful authentication. This usually meant that the value preserved by
1236 the server_setid option was incorrect.
1238 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1240 PH/09 Installed PCRE release 7.0.
1242 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1243 run for batched SMTP input. It is now run at the start of every message
1244 in the batch. While fixing this I discovered that the process information
1245 (output by running exiwhat) was not always getting set for -bs and -bS
1246 input. This is fixed, and it now also says "batched" for BSMTP.
1248 PH/11 Added control=no_pipelining.
1250 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1251 patch, slightly modified), and move the expansion of helo_data till after
1252 the connection is made in the smtp transport (so it can use these
1255 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1257 PH/14 Added log_selector = +pid.
1259 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1261 PH/16 Add ${if forany and ${if forall.
1263 PH/17 Added dsn_from option to vary the From: line in DSNs.
1265 PH/18 Flush SMTP output before performing a callout, unless control =
1266 no_callout_flush is set.
1268 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1269 was true (the default) a successful delivery failed to delete the retry
1270 item, thus causing premature timeout of the address. The bug is now
1273 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1275 PH/21 Long custom messages for fakedefer and fakereject are now split up
1276 into multiline reponses in the same way that messages for "deny" and
1277 other ACL rejections are.
1279 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1280 with slight modification.
1282 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1283 draft, changing the syntax and factoring some duplicate code".
1285 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1286 for deliveries of the second and subsequent messages over the same SMTP
1289 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1290 ${reduce, with only minor "tidies".
1292 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1294 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1295 expansion side effects.
1297 PH/27 When a message times out after an over-quota error from an Exim-imposed
1298 quota, the bounce message says "mailbox is full". This message was not
1299 being given when it was a system quota that was exceeded. It now should
1302 MH/03 Made $recipients available in local_scan(). local_scan() already has
1303 better access to the recipient list through recipients_list[], but
1304 $recipients can be useful in postmaster-provided expansion strings.
1306 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1307 in the case of a MAIL command with additional options following the
1308 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1309 were accidentally chopped off.
1311 PH/29 SMTP synchronization checks are implemented when a command is read -
1312 there is a check that no more input is waiting when there shouldn't be
1313 any. However, for some commands, a delay in an ACL can mean that it is
1314 some time before the response is written. In this time, more input might
1315 arrive, invalidly. So now there are extra checks after an ACL has run for
1316 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1317 pipelining has not been advertised.
1319 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1321 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1322 returned in the SMTP response when smtp_return_error_details was false.
1323 This has been fixed.
1325 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1326 instead of the C I/O that was originally supplied, because problems were
1327 reported on Solaris.
1329 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1330 Exim which did not show up earlier: it was assuming that a call to
1331 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1332 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1333 was a macro that became an assignment, so it seemed to work. This has
1334 changed to a proper function call with a void return, hence the compile
1335 error. Exim's code has been fixed.
1337 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1340 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1342 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1344 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1345 because it turns out to be too big or there is a timeout) there is no
1346 "Completed" line in the log. When some messages of this type were
1347 selected by exigrep, they were listed as "not completed". Others were
1348 picked up by some special patterns. I have improved the selection
1349 criteria to be more general.
1351 PH/38 The host_find_failed option in the manualroute router can now be set
1352 to "ignore", to completely ignore a host whose IP address cannot be
1353 found. If all hosts are ignored, the behaviour is controlled by the new
1354 host_all_ignored option.
1356 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1357 homing or because of multiple MX records with /mx) generated more than
1358 one IP address, and the following item turned out to be the local host,
1359 all the secondary addresses of the first item were incorrectly removed
1360 from the list, along with the local host and any following hosts (which
1361 is what is supposed to happen).
1363 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1364 whoever called Exim into the -H file. In the case of the daemon it was
1365 behaving confusingly. When first started, it used values for whoever
1366 started the daemon, but after a SIGHUP it used the Exim user (because it
1367 calls itself on a restart). I have changed the code so that it now always
1370 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1371 message are rejected with the same error (e.g. no authentication or bad
1372 sender address), and a DATA command is nevertheless sent (as can happen
1373 with PIPELINING or a stupid MUA), the error message that was given to the
1374 RCPT commands is included in the rejection of the DATA command. This is
1375 intended to be helpful for MUAs that show only the final error to their
1378 PH/42 Another patch from the Sieve maintainer.
1380 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1381 Eximstats - Fixed some broken HTML links and added missing column headers
1383 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1384 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1386 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1388 PH/43 Yet another patch from the Sieve maintainer.
1390 PH/44 I found a way to check for a TCP/IP connection going away before sending
1391 the response to the final '.' that terminates a message, but only in the
1392 case where the client has not sent further data following the '.'
1393 (unfortunately, this is allowed). However, in many cases there won't be
1394 any further data because there won't be any more messages to send. A call
1395 to select() can be used: if it shows that the input is "ready", there is
1396 either input waiting, or the socket has been closed. An attempt to read
1397 the next input character can distinguish the two cases. Previously, Exim
1398 would have sent an OK response which the client would never have see.
1399 This could lead to message repetition. This fix should cure that, at
1400 least in a lot of common cases.
1402 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1403 advertised in response to EHLO.
1409 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1410 fixed by 4.65/MH/01 (is this a record?) are fixed:
1412 (i) An empty string was always treated as zero by the numeric comparison
1413 operators. This behaviour has been restored.
1415 (ii) It is documented that the numeric comparison operators always treat
1416 their arguments as decimal numbers. This was broken in that numbers
1417 starting with 0 were being interpreted as octal.
1419 While fixing these problems I realized that there was another issue that
1420 hadn't been noticed. Values of message_size_limit (both the global option
1421 and the transport option) were treated as octal if they started with 0.
1422 The documentation was vague. These values are now always treated as
1423 decimal, and I will make that clear in the documentation.
1429 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1430 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1433 MH/01 Don't check that the operands of numeric comparison operators are
1434 integers when their expansion is in "skipping" mode (fixes bug
1435 introduced by 4.64-PH/07).
1437 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1438 child addresses, Exim now panics and dies. Previously, because the count
1439 is held in a short int, deliveries were likely to be lost. As such a
1440 large number of recipients for a single message is ridiculous
1441 (performance will be very, very poor), I have chosen to impose a limit
1442 rather than extend the field.
1448 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1449 leftover -K file (the existence of which was triggered by #402).
1450 While we were at it, introduced process PID as part of the -K
1451 filename. This should rule out race conditions when creating
1454 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1455 processing considerably. Previous code took too long for large mails,
1456 triggering a timeout which in turn triggers #401.
1458 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1459 in the DK code in transports.c. sendfile() is not really portable,
1460 hence the _LINUX specificness.
1462 TF/01 In the add_headers option to the mail command in an Exim filter,
1463 there was a bug that Exim would claim a syntax error in any
1464 header after the first one which had an odd number of characters
1467 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1468 callout verification, Exim cached a "reject" for the entire domain. This
1469 is correct for most verifications, but it is not correct for a recipient
1470 verification with use_sender or use_postmaster set, because in that case
1471 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1472 case of MAIL FROM:<> rejection from other early rejections (e.g.
1473 rejection of HELO). When verifying a recipient using a non-null MAIL
1474 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1475 Whatever the result of the callout, the value of the domain cache is
1476 left unchanged (for any other kind of callout, getting as far as trying
1477 RCPT means that the domain itself is ok).
1479 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1482 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1483 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1484 session, the daemon ignores ECONNECT errors and logs others; it now
1485 ignores EPIPE as well.
1487 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1488 (quoted-printable decoding).
1490 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1491 later the small subsequent patch to fix an introduced bug.
1493 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1495 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1497 PH/08 An error is now given if message_size_limit is specified negative.
1499 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1500 to be given (somewhat) arbitrary names.
1502 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1505 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1506 miscellaneous code fixes
1508 PH/10 Added the log_reject_target ACL modifier to specify where to log
1511 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1512 hostname. This is wrong, because it relates to the incoming message (and
1513 probably the interface on which it is arriving) and not to the outgoing
1514 callout (which could be using a different interface). This has been
1515 changed to use the value of the helo_data option from the smtp transport
1516 instead - this is what is used when a message is actually being sent. If
1517 there is no remote transport (possible with a router that sets up host
1518 addresses), $smtp_active_hostname is used.
1520 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1521 tweaks were necessary in order to get it to work (see also 21 below):
1522 (a) The code assumed that strncpy() returns a negative number on buffer
1523 overflow, which isn't the case. Replaced with Exim's string_format()
1525 (b) There were several signed/unsigned issues. I just did the minimum
1526 hacking in of casts. There is scope for a larger refactoring.
1527 (c) The code used strcasecmp() which is not a standard C function.
1528 Replaced with Exim's strcmpic() function.
1529 (d) The code set only $1; it now sets $auth1 as well.
1530 (e) A simple test gave the error "authentication client didn't specify
1531 service in request". It would seem that Dovecot has changed its
1532 interface. Fortunately there's a specification; I followed it and
1533 changed what the client sends and it appears to be working now.
1535 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1538 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1539 address A is aliased to B and C, where B exists and C does not. Without
1540 -v the output is "A verified" because verification stops after a
1541 successful redirection if more than one address is generated. However,
1542 with -v the child addresses are also verified. Exim was outputting "A
1543 failed to verify" and then showing the successful verification for C,
1544 with its parentage. It now outputs "B failed to verify", showing B's
1545 parentage before showing the successful verification of C.
1547 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1548 look up a TXT record in a specific list after matching in a combined
1551 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1552 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1553 they consult the DNS. I had assumed they would set it the way they
1554 wanted; and indeed my experiments on Linux seem to show that in some
1555 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1556 To be on the safe side, however, I have now made the interface to
1557 host_find_byname() similar to host_find_bydns(), with an argument
1558 containing the DNS resolver options. The host_find_byname() function now
1559 sets these options at its start, just as host_find_bydns() does. The smtp
1560 transport options dns_qualify_single and dns_search_parents are passed to
1561 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1562 of host_find_byname() use the default settings of RES_DEFNAMES
1563 (qualify_single) but not RES_DNSRCH (search_parents).
1565 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1566 spool_read_header() do less string testing, by means of a preliminary
1567 switch on the second character of optional "-foo" lines. (This is
1568 overdue, caused by the large number of possibilities that now exist.
1569 Originally there were few.) While I was there, I also converted the
1570 str(n)cmp tests so they don't re-test the leading "-" and the first
1571 character, in the hope this might squeeze out yet more improvement.
1573 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1574 flag allowing group syntax was set by the header_syntax check but not
1575 turned off, possible causing trouble later; (2) The flag was not being
1576 set at all for the header_verify test, causing "group"-style headers to
1577 be rejected. I have now set it in this case, and also caused header_
1578 verify to ignore an empty address taken from a group. While doing this, I
1579 came across some other cases where the code for allowing group syntax
1580 while scanning a header line wasn't quite right (mostly, not resetting
1581 the flag correctly in the right place). These bugs could have caused
1582 trouble for malformed header lines. I hope it is now all correct.
1584 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1585 with the "reply" argument non-NULL. The code, however (which originally
1586 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1587 but it didn't always do it. This confused somebody who was copying the
1588 code for some other use. I have removed all the tests.
1590 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1591 feature that was used to support insecure browsers during the U.S. crypto
1592 embargo. It requires special client support, and Exim is probably the
1593 only MTA that supported it -- and would never use it because real RSA is
1594 always available. This code has been removed, because it had the bad
1595 effect of slowing Exim down by computing (never used) parameters for the
1596 RSA_EXPORT functionality.
1598 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1599 authenticator to fail if there's a tab character in the incoming data
1600 (there should never be unless someone is messing about, as it's supposed
1601 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1602 if the connection is using TLS or if the remote IP is the same as the
1603 local IP, and the "valid-client-cert option" if a client certificate has
1606 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1607 authenticators. This can be used for authorization after authentication
1608 succeeds. (In the case of plaintext, it servers for both authentication
1611 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1612 if any retry times were supplied.
1614 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1615 connection, where there is no client IP address to check. In this
1616 situation, the verify now always succeeds.
1618 PH/25 Applied John Jetmore's -Mset patch.
1620 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1622 PH/27 In a string expansion for a processed (not raw) header when multiple
1623 headers of the same name were present, leading whitespace was being
1624 removed from all of them, but trailing whitespace was being removed only
1625 from the last one. Now trailing whitespace is removed from each header
1626 before concatenation. Completely empty headers in a concatenation (as
1627 before) are ignored.
1629 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1630 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1632 PH/29 [Removed. This was a change that I later backed out, and forgot to
1633 correct the ChangeLog entry (that I had efficiently created) before
1634 committing the later change.]
1636 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1637 address errors (4xx response to RCPT) over the same connection as other
1638 messages routed to the same hosts. Such deliveries are always "forced",
1639 so retry times are not inspected. This resulted in far too many retries
1640 for the affected addresses. The effect occurred only when there were more
1641 hosts than the hosts_max_try setting in the smtp transport when it had
1642 the 4xx errors. Those hosts that it had tried were not added to the list
1643 of hosts for which the message was waiting, so if all were tried, there
1644 was no problem. Two fixes have been applied:
1646 (i) If there are any address or message errors in an SMTP delivery, none
1647 of the hosts (tried or untried) are now added to the list of hosts
1648 for which the message is waiting, so the message should not be a
1649 candidate for sending over the same connection that was used for a
1650 successful delivery of some other message. This seems entirely
1651 reasonable: after all the message is NOT "waiting for some host".
1652 This is so "obvious" that I'm not sure why it wasn't done
1653 previously. Hope I haven't missed anything, but it can't do any
1654 harm, as the worst effect is to miss an optimization.
1656 (ii) If, despite (i), such a delivery is accidentally attempted, the
1657 routing retry time is respected, so at least it doesn't keep
1658 hammering the server.
1660 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1661 in ${readsocket because some servers need this prod.
1663 PH/32 Added some extra debug output when updating a wait-xxx database.
1665 PH/33 The hint "could be header name not terminated by colon", which has been
1666 given for certain expansion errors for a long time, was not being given
1667 for the ${if def:h_colon_omitted{... case.
1669 PH/34 The spec says: "With one important exception, whenever a domain list is
1670 being scanned, $domain contains the subject domain." There was at least
1671 one case where this was not true.
1673 PH/35 The error "getsockname() failed: connection reset by peer" was being
1674 written to the panic log as well as the main log, but it isn't really
1675 panic-worthy as it just means the connection died rather early on. I have
1676 removed the panic log writing for the ECONNRESET error when getsockname()
1679 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1680 runs only) independently of the message's sender address. This meant
1681 that, if the 4xx error was in fact related to the sender, a different
1682 message to the same recipient with a different sender could confuse
1683 things. In particualar, this can happen when sending to a greylisting
1684 server, but other circumstances could also provoke similar problems.
1685 I have changed the default so that the retry time for these errors is now
1686 based a combination of the sender and recipient addresses. This change
1687 can be overridden by setting address_retry_include_sender=false in the
1690 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1691 remote server are returned as part of bounce messages. This was not
1692 happening for LMTP over a pipe (the lmtp transport), but now it is the
1693 same for both kinds of LMTP.
1695 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1696 in header lines that were in fact CNAMEs. This is no longer the case.
1698 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1699 and queue runs started by the daemon processed all messages. This has
1700 been fixed so that -R and -S can now usefully be given with -q<time>.
1702 PH/40 Import PCRE release 6.7 (fixes some bugs).
1704 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1706 PH/42 Give an error if -q is specified more than once.
1708 PH/43 Renamed the variables $interface_address and $interface_port as
1709 $received_ip_address and $received_port, to make it clear that these
1710 values apply to message reception, and not to the outgoing interface when
1711 a message is delivered. (The old names remain recognized, of course.)
1713 PH/44 There was no timeout on the connect() call when using a Unix domain
1714 socket in the ${readsocket expansion. There now is.
1716 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1717 be meaningful with "accept".
1719 SC/01 Eximstats V1.43
1720 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1722 SC/02 Eximstats V1.44
1723 Use a glob alias rather than an array ref in the generated
1724 parser. This improves both readability and performance.
1726 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1727 Collect SpamAssassin and rejection statistics.
1728 Don't display local sender or destination tables unless
1729 there is data to show.
1730 Added average volumes into the top table text output.
1732 SC/04 Eximstats V1.46
1733 Collect data on the number of addresses (recipients)
1734 as well as the number of messages.
1736 SC/05 Eximstats V1.47
1737 Added 'Message too big' to the list of mail rejection
1738 reasons (thanks to Marco Gaiarin).
1740 SC/06 Eximstats V1.48
1741 Mainlog lines which have GMT offsets and are too short to
1742 have a flag are now skipped.
1744 SC/07 Eximstats V1.49 (Alain Williams)
1745 Added the -emptyok flag.
1747 SC/08 Eximstats V1.50
1748 Fixes for obtaining the IP address from reject messages.
1750 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1751 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1752 whitesspace changes from 4.64-PH/27
1754 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1757 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1758 are found, allow negative numbers in numeric criteria)
1760 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1762 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1763 to match changes made in 4.64-PH/43
1765 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1767 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1768 "domain" is actually an IP address, and force a failure. This locks out
1769 those revolvers/nameservers that support "A-for-A" lookups, in
1770 contravention of the specifications.
1772 PH/48 When a host name was looked up from an IP address, and the subsequent
1773 forward lookup of the name timed out, the host name was left in
1774 $sender_host_name, contrary to the specification.
1776 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1777 restricted to single-key lookups, Exim was not diagnosing an error if
1778 * or *@ was used with a query-style lookup.
1780 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1782 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1783 long ago, but noone interested enough thought of it. Let's just say that
1784 the "1.1" means that there are some new functions that weren't there at
1785 some point in the past.
1787 PH/51 Error processing for expansion failure of helo_data from an smtp
1788 transport during callout processing was broken.
1790 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1791 tested/used via the -bh/-bhc/-bs options.
1793 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1794 bug, fixed in subsequent PCRE releases).
1796 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1797 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1799 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1805 SC/01 Use a glob alias rather than an array ref in eximstats generated
1806 parser. This improves both readability and performance.
1808 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1809 Don't display local sender or destination tables in eximstats unless
1810 there is data to show.
1811 Added average volumes into the eximstats top table text output.
1813 SC/03 Collect data on the number of addresses (recipients) as well
1814 as the number of messages in eximstats.
1816 TF/01 Correct an error in the documentation for the redirect router. Exim
1817 does (usually) call initgroups() when daemonizing.
1819 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1820 with consistent privilege compared to when running as a daemon.
1822 TF/03 Note in the spec that $authenticated_id is not set for local
1823 submissions from trusted users.
1825 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1826 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1828 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1829 by adding some example configuration directives to the default
1830 configuration file. A little bit of work is required to uncomment the
1831 directives and define how usernames and passwords are checked, but
1832 there is now a framework to start from.
1834 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1835 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1836 without this. I don't know how relevant this is to other LDAP libraries.
1838 PH/02 Add the verb name to the "unknown ACL verb" error.
1840 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1842 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1844 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1845 directory not expanded when it should be if an expanded home directory
1846 was set for the address (which is overridden by the transport).
1848 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1851 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1852 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1853 because it is too late at that time, and has no effect.
1855 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1856 security issue with \' (bugzilla #107). I could not use the
1857 PQescapeStringConn() function, because it needs a PGconn value as one of
1860 PH/08 When testing addresses using -bt, indicate those final addresses that
1861 are duplicates that would not cause an additional delivery. At least one
1862 person was confused, thinking that -bt output corresponded to deliveries.
1863 (Suppressing duplicates isn't a good idea as you lose the information
1864 about possibly different redirections that led to the duplicates.)
1866 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1867 systems where poll() doesn't work, in particular OS X.
1869 PH/10 Added more information to debugging output for retry time not reached.
1871 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1872 operations in malware.c.
1874 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1877 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1878 syslog_duplication set false, log lines that would normally be written
1879 both the the main log and to the reject log were not written to syslog at
1882 PH/14 In the default configuration, change the use of "message" in ACL warn
1883 statements to "add_header".
1885 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1886 not followed by a command (e.g. "seen endif").
1888 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1889 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1892 PH/17 Added extra conditions to the default value of delay_warning_condition
1896 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1897 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1898 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1901 The Auto-Submitted: and various List- headers are standardised, whereas I
1902 don't think Precedence: ever was.
1904 PH/18 Refactored debugging code in route_finduser() to show more information,
1905 in particular, the error code if getpwnam() issues one.
1907 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1908 This is apparently needed in addition to the PH/07 change above to avoid
1909 any possible encoding problems.
1911 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1912 but not after initializing Perl.
1914 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1915 output them only if debugging. By default they are written stderr,
1916 apparently, which is not desirable.
1918 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1921 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1924 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1926 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1927 authenticated or an ident call has been made. Suppress the default
1928 values for $authenticated_id and $authenticated_sender (but permit -oMai
1929 and -oMas) when testing with -bh.
1931 PH/24 Re-jigged the order of the tests in the default configuration so that the
1932 tests for valid domains and recipients precede the DNS black list and CSA
1933 tests, on the grounds that those ones are more expensive.
1935 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1936 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1937 This bug exists in every version of Exim that I still have, right back to
1940 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1941 However, an attempt to turn on case-sensitivity in a regex key by
1942 including (?-i) didn't work because the subject string was already
1943 lowercased, and the effects were non-intuitive. It turns out that a
1944 one-line patch can be used to allow (?-i) to work as expected.
1950 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1951 other effects) broke the use of negated acl sub-conditions.
1953 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1956 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1957 "Deny" causes Exim to reject the incoming connection with a 554 error.
1958 Unfortunately, if there is a major crisis, such as a disk failure,
1959 tcp-wrappers gives "deny", whereas what one would like would be some
1960 kind of temporary error. A kludge has been added to help with this.
1961 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1962 554 error is used if errno is still zero or contains ENOENT (which occurs
1963 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1966 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1968 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1969 errors. Otherwise a message that provokes a temporary error (when other
1970 messages do not) can cause a whole host to time out.
1972 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1973 addresses were routed directly to files or pipes from a redirect router.
1974 File deliveries just didn't batch; pipe deliveries might have suffered
1977 PH/06 A failure to get a lock for a hints database would erroneously always say
1978 "Failed to get write lock", even when it was really a read lock.
1980 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1981 of 0600. This has been changed to use the value of the lockfile_mode
1982 option (which defaults to 0600).
1984 PH/08 Applied small patch from the Sieve maintainer.
1986 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1987 folder from quota calculations, a direct delivery into this folder messed
1988 up the contents of the maildirsize file. This was because the regex was
1989 used only to exclude .Trash (or whatever) when the size of the mailbox
1990 was calculated. There was no check that a delivery was happening into an
1991 excluded directory. This bug has been fixed by ignoring all quota
1992 processing for deliveries into excluded directories.
1994 PH/10 Added the maildirfolder_create_regex option to appendfile.
2000 PH/01 The code for finding all the local interface addresses on a FreeBSD
2001 system running IPv6 was broken. This may well have applied to all BSD
2002 systems, as well as to others that have similar system calls. The broken
2003 code found IPv4 interfaces correctly, but gave incorrect values for the
2004 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2005 that it would not match correctly against @[] and not recognize the IPv6
2008 PH/02 The ipliteral router was not recognizing addresses of the form user@
2009 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2011 PH/03 Added disable_ipv6.
2013 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2014 decoded form, because it is most often used to construct To: headers
2015 lines in autoreplies, and the decoded form may well be syntactically
2016 invalid. However, $reply_address has leading white space removed, and all
2017 newlines turned into spaces so that the autoreply transport does not
2020 PH/05 If group was specified without a user on a router, and no group or user
2021 was specified on a transport, the group from the router was ignored.
2023 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2024 for visible compile-time settings that can be used to change these
2025 numbers, for those that want even more. Backwards compatibility with old
2026 spool files has been maintained. However, going back to a previous Exim
2027 release will lost any variables that are in spool files.
2029 PH/07 Two small changes when running in the test harness: increase delay when
2030 passing a TCP/IP connection to a new process, in case the original
2031 process has to generate a bounce, and remove special handling of
2032 127.0.0.2 (sic), which is no longer necessary.
2034 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2035 be the same on different OS.
2037 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2040 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2041 whether --show-vars was specified or not
2043 JJ/02 exipick: Added support for new ACL variable spool format introduced
2046 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2047 syntactically invalid From: or Reply-to: line, and a filter used this to
2048 generate an autoreply, and therefore failed to obtain an address for the
2049 autoreply, Exim could try to deliver to a non-existent relative file
2050 name, causing unrelated and misleading errors. What now happens is that
2051 it logs this as a hard delivery error, but does not attempt to create a
2054 PH/11 The exinext utility has a -C option for testing purposes, but although
2055 the given file was scanned by exinext itself; it wasn't being passed on
2056 when Exim was called.
2058 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2059 an end-of-file indication when reading a command response.
2061 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2062 compiled. In many other places in Exim, IPv6 addresses are always
2063 recognized, so I have changed this. It also means that IPv4 domain
2064 literals of the form [IPV4:n.n.n.n] are now always recognized.
2066 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2067 used if the router is not running as root, for example, when verifying at
2068 ACL time, or when using -bh. The debugging output from this situation was
2069 non-existent - all you got was a failure to exec. I have made two
2072 (a) Failures to set uid/gid, the current directory, or a process leader
2073 in a subprocess such as that created by queryprogram now generate
2074 suitable debugging ouput when -d is set.
2076 (b) The queryprogram router detects when it is not running as root,
2077 outputs suitable debugging information if -d is set, and then runs
2078 the subprocess without attempting to change uid/gid.
2080 PH/15 Minor change to Makefile for building test_host (undocumented testing
2083 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2084 additional section of a DNS packet that returns MX or SRV records.
2085 Instead, it always explicitly searches for A/AAAA records. This avoids
2086 major problems that occur when a DNS server includes only records of one
2087 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2088 fixed another bug: if SRV records were looked up and the corresponding
2089 address records were *not* found in the additional section, the port
2090 values from the SRV records were lost.
2092 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2093 using the correct key (the original address) when searching the retry
2094 rules in order to find which one to use for generating the retry hint.
2096 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2097 adding the default one. Similarly, if it contains a Reply-To: header, the
2098 errors_reply_to option, if set, is not used.
2100 PH/19 When calculating a retry time, Exim used to measure the "time since
2101 failure" by looking at the "first failed" field in the retry record. Now
2102 it does not use this if it is later than than the arrival time of the
2103 message. Instead it uses the arrival time. This makes for better
2104 behaviour in cases where some deliveries succeed, thus re-setting the
2105 "first failed" field. An example is a quota failure for a huge message
2106 when small messages continue to be delivered. Without this change, the
2107 "time since failure" will always be short, possible causing more frequent
2108 delivery attempts for the huge message than are intended.
2109 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2111 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2112 $1, $2, $3) because the numerical variables can be reset during some
2113 expansion items (e.g. "match"), thereby losing the authentication data.
2115 PH/21 Make -bV show the size of off_t variables so that the test suite can
2116 decide whether to run tests for quotas > 2G.
2118 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2119 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2120 filecount value is greater than 2G or if a quota value is greater than 2G
2121 on a system where the size of off_t is not greater than 4, a panic error
2124 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2125 never match. The debug and -bh output now contains an explicit error
2126 message indicating a malformed IPv4 address or mask.
2128 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2129 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2130 PH/23 above applies.
2132 PH/25 Do not write to syslog when running in the test harness. The only
2133 occasion when this arises is a failure to open the main or panic logs
2134 (for which there is an explicit test).
2136 PH/26 Added the /no_tell option to "control=freeze".
2138 PH/27 If a host name lookup failed very early in a connection, for example, if
2139 the IP address matched host_lookup and the reverse lookup yielded a name
2140 that did not have a forward lookup, an error message of the form "no IP
2141 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2142 could be logged. Now it outputs the IP address instead of "NULL".
2144 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2145 allows the sender and the authenticated sender to be set when
2146 submitting a message from within Exim. Since child_open_exim() is
2147 documented for local_scan(), the new function should be too.
2149 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2150 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2151 results in an empty string is now treated as unset.
2153 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2155 PH/31 Added sender_verify_fail logging option.
2157 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2158 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2159 tidied the source and removed it altogether.
2161 PH/33 When a queue run was abandoned because the load average was too high, a
2162 log line was always written; now it is written only if the queue_run log
2163 selector is set. In addition, the log line for abandonment now contains
2164 information about the queue run such as the pid. This is always present
2165 in "start" and "stop" lines but was omitted from the "abandon" line.
2167 PH/34 Omit spaces between a header name and the colon in the error message that
2168 is given when verify = headers_syntax fails (if there are lots of them,
2169 the message gets confusing).
2171 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2172 names, as there are now some PTR records that contain slashes. This check
2173 is only to protect against broken name servers that fall over on strange
2174 characters, so the fact that it applies to all lookups doesn't matter.
2176 PH/36 Now that the new test suite is complete, we can remove some of the
2177 special code in Exim that was needed for the old test suite. For example,
2178 sorting DNS records because real resolvers return them in an arbitrary
2179 order. The new test suite's fake resolver always returns records in the
2182 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2183 bounces) except when queue_only is set, to avoid logging races between
2184 the different processes.
2186 PH/38 Panic-die if .include specifies a non-absolute path.
2188 PH/39 A tweak to the "H" retry rule from its user.
2190 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2191 a label. They prevented compilation on older perls.
2193 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2194 a warning to be raised on newish perls.
2196 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2197 on queue. Changes to match documented behaviour of showing count of
2198 messages matching specified criteria.
2200 PH/40 Changed the default ident timeout from 30s to 5s.
2202 PH/41 Added support for the use of login_cap features, on those BSD systems
2203 that have them, for controlling the resources used by pipe deliveries.
2205 PH/42 The content-scanning code uses fopen() to create files in which to put
2206 message data. Previously it was not paying any attention to the mode of
2207 the files. Exim runs with umask(0) because the rest of the code creates
2208 files with open(), and sets the required mode explicitly. Thus, these
2209 files were ending up world-writeable. This was not a big issue, because,
2210 being within the spool directory, they were not world-accessible. I have
2211 created a function called modefopen, which takes an additional mode
2212 argument. It sets umask(777), creates the file, chmods it to the required
2213 mode, then resets the umask. All the relevant calls to fopen() in the
2214 content scanning code have been changed to use this function.
2216 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2217 to 24 hours. This avoids potential overflow problems when processing G
2218 and H retry rules. I suspect nobody ever tinkers with this value.
2220 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2222 PH/45 When the plaintext authenticator is running as a client, the server's
2223 challenges are checked to ensure they are valid base64 strings. By
2224 default, the authentication attempt is cancelled if an invalid string is
2225 received. Setting client_ignore_invalid_base64 true ignores these errors.
2226 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2227 they are received. Thus, the responses can be made to depend on the
2228 challenges. If an invalid string is ignored, an empty string is placed in
2231 PH/46 Messages that are created by the autoreply transport now contains a
2232 References: header, in accordance with RFCs 2822 and 3834.
2234 PH/47 Added authenticated_sender_force to the smtp transport.
2236 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2238 PH/49 Installed latest patch from the Sieve maintainer.
2240 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2241 was also set, the appendfile transport was unnecessarily scanning a
2242 directory of message files (e.g. for maildir delivery) to find the count
2243 of files (along with the size), even though it did not need this
2244 information. It now does the scan only if it needs to find either the
2245 size of the count of files.
2247 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2249 PH/52 Two bugs concerned with error handling when the smtp transport is
2252 (i) Exim was not creating retry information for temporary errors given
2253 for individual recipients after the DATA command when the smtp transport
2254 was used in LMTP mode. This meant that they could be retried too
2255 frequently, and not timed out correctly.
2257 (ii) Exim was setting the flag that allows error details to be returned
2258 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2259 recipients that were returned after the DATA command.
2261 PH/53 This is related to PH/52, but is more general: for any failing address,
2262 when detailed error information was permitted to be returned to the
2263 sender, but the error was temporary, then after the final timeout, only
2264 "retry timeout exceeded" was returned. Now it returns the full error as
2265 well as "retry timeout exceeded".
2267 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2268 do this, and (what is worse) MTAs that accept it.
2270 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2271 will now be deprecated.
2273 PH/56 New os.c-cygwin from the Cygwin maintainer.
2275 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2276 formats (previously only available in exim formats via -bpr, -bpru,
2277 and -bpra. Now also available in native and exiqgrep formats)
2279 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2280 with very large, slow to parse queues
2282 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2284 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2286 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2287 responses to authentication challenges, though it was showing the
2288 challenges; (ii) I've removed the CR characters from the debug output for
2291 PH/58 Allow for the insertion of a newline as well as a space when a string
2292 is turned into more than one encoded-word during RFC 2047 encoding. The
2293 Sieve code now uses this.
2295 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2296 data_4xx, lost_connection, tls_required.
2298 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2299 message was being sent as an SMTP response.
2301 PH/61 Add -l and -k options to exicyclog.
2303 PH/62 When verifying, if an address was redirected to one new address, so that
2304 verification continued, and the new address failed or deferred after
2305 having set something in $address_data, the value of $address_data was not
2306 passed back to the ACL. This was different to the case when no
2307 redirection occurred. The value is now passed back in both cases.
2309 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2310 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2311 use login_cap.h, so on its own it isn't the distinguishing feature. The
2312 new name refers directly to the setclassresources() function.
2314 PH/65 Added configuration files for NetBSD3.
2316 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2318 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2319 is preferred over IPv4.
2321 PH/68 The bounce_return_message and bounce_return_body options were not being
2322 honoured for bounces generated during the reception of non-SMTP messages.
2323 In particular, this applied to messages rejected by the ACL. This bug has
2324 been fixed. However, if bounce_return_message is true and bounce_return_
2325 body is false, the headers that are returned for a non-SMTP message
2326 include only those that have been read before the error was detected.
2327 (In the case of an ACL rejection, they have all been read.)
2329 PH/69 The HTML version of the specification is now built in a directory called
2330 spec_html instead of spec.html, because the latter looks like a path with
2331 a MIME-type, and this confuses some software.
2333 PH/70 Catch two compiler warnings in sieve.c.
2335 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2336 function verify_get_ident() calls ip_connect() to connect a socket, but
2337 if the "connect()" function timed out, ip_connect() used to close the
2338 socket. However, verify_get_ident() also closes the socket later, and in
2339 between Exim writes to the log, which may get opened at this point. When
2340 the socket was closed in ip_connect(), the log could get the same file
2341 descriptor number as the socket. This naturally causes chaos. The fix is
2342 not to close the socket in ip_connect(); the socket should be closed by
2343 the function that creates it. There was only one place in the code where
2344 this was missing, in the iplookup router, which I don't think anybody now
2345 uses, but I've fixed it anyway.
2347 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2348 well as to direct DNS lookups. Otherwise the handling of names in host
2349 lists is inconsistent and therefore confusing.
2355 PH/01 Two changes to the default runtime configuration:
2357 (1) Move the checks for relay_from_hosts and authenticated clients from
2358 after to before the (commented out) DNS black list checks.
2360 (2) Add control=submission to the relay_from_hosts and authenticated
2361 clients checks, on the grounds that messages accepted by these
2362 statements are most likely to be submissions.
2364 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2366 (1) Generate an error if the third argument for the ${prvs expansion is
2369 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2372 (3) Reset the variables that are obtained from the first argument of
2373 ${prvscheck and used in the second argument before leaving the code,
2374 because their memory is reclaimed, so using them afterwards may do
2377 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2378 one (it's much easier than Tom thought :-).
2380 (5) Because of (4), we can now allow for the use of $prvscheck_result
2381 inside the third argument.
2383 PH/03 For some reason, the default setting of PATH when running a command from
2384 a pipe transport was just "/usr/bin". I have changed it to
2387 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2388 anything to be listed in the output from -bV.
2390 PH/05 When a filter generated an autoreply, the entire To: header line was
2391 quoted in the delivery log line, like this:
2393 => >A.N.Other <ano@some.domain> <original@ddress> ...
2395 This has been changed so that it extracts the operative address. There
2396 may be more than one such address. If so, they are comma-separated, like
2399 => >ano@some.domain,ona@other.domain <original@ddress> ...
2401 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2402 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2403 not being looked up in the rDNS to get a host name, Exim was showing the
2404 IP address twice in Received: lines, even though the IP addresses were
2405 identical. For example:
2407 Received: from [1.2.3.4] (helo=[1.2.3.4])
2409 However, if the real host name was known, it was omitting the HELO data
2410 if it matched the actual IP address. This has been tidied up so that it
2411 doesn't show the same IP address twice.
2413 PH/07 When both +timestamp and +memory debugging was on, the value given by
2414 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2415 called by the debug printing, thereby overwriting the timestamp buffer.
2416 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2418 PH/08 When the original message was included in an autoreply transport, it
2419 always said "this is a copy of the message, including all the headers",
2420 even if body_only or headers_only was set. It now gives an appropriate
2423 PH/09 Applied a patch from the Sieve maintainer which:
2425 o fixes some comments
2426 o adds the (disabled) notify extension core
2427 o adds some debug output for the result of if/elsif tests
2428 o points to the current vacation draft in the documentation
2429 and documents the missing references header update
2433 o fixes a bug in processing the envelope test (when testing
2434 multiple envelope elements, the last element determinted the
2437 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2438 Electronic Mail") by including:
2440 Auto-submitted: auto-generated
2442 in the messages that it generates (bounce messages and others, such as
2443 warnings). In the case of bounce messages for non-SMTP mesages, there was
2444 also a typo: it was using "Auto_submitted" (underscore instead of
2445 hyphen). Since every message generated by Exim is necessarily in response
2446 to another message, thes have all been changed to:
2448 Auto-Submitted: auto-replied
2450 in accordance with these statements in the RFC:
2452 The auto-replied keyword:
2454 - SHOULD be used on messages sent in direct response to another
2455 message by an automatic process,
2457 - MUST NOT be used on manually-generated messages,
2459 - MAY be used on Delivery Status Notifications (DSNs) and Message
2460 Disposition Notifications (MDNs),
2462 - MUST NOT be used on messages generated by automatic or periodic
2463 processes, except for messages which are automatic responses to
2466 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2467 to the default Received: header definition.
2469 PH/12 Added log selector acl_warn_skipped (default on).
2471 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2472 variables because (a) they are in the wrong storage pool and (b) even if
2473 they were copied, it wouldn't work properly because of the caching.
2475 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2476 checking when decoding. Apparently there are clients that generate
2477 overlong encoded strings. Why am I not surprised?
2479 PH/15 If the first argument of "${if match_address" was not empty, but did not
2480 contain an "@" character, Exim crashed. Now it writes a panic log message
2481 and treats the condition as false.
2483 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2485 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2486 extension "envelope-auth". The code is finished and in agreement with
2487 other implementations, but there is no documentation so far and in fact,
2488 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2489 not changing the active code.
2491 Print executed "if" and "elsif" statements when debugging is used. This
2492 helps a great deal to understand what a filter does.
2494 Document more things not specified clearly in RFC3028. I had all this
2495 sorted out, when out of a sudden new issues came to my mind. Oops."
2497 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2500 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2501 canonical form (as documented). However, after a host name lookup from
2502 the IP address, check_host() was doing a simple string comparison with
2503 addresses acquired from the DNS when checking that the found name did
2504 have the original IP as one of its addresses. Since any found IPv6
2505 addresses are likely to be in abbreviated form, the comparison could
2506 fail. Luckily, there already exists a function for doing the comparison
2507 by converting both addresses to binary, so now that is used instead of
2508 the text comparison.
2510 PH/20 There was another similar case to PH/19, when a complete host name was
2511 given in a host list; looking up its IP address could give an abbreviated
2512 form, whereas the current host's name might or might not be abbreviated.
2513 The same fix has been applied.
2519 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2520 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2523 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2524 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2526 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2528 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2529 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2530 Also, for "Auto-Submitted", treat anything other than "no" as
2531 non-personal, in accordance with RFC 3834. (Previously it treated
2532 anything starting "auto-" as non-personal.)
2534 TF/01 The control=submission/name=... option had a problem with syntax
2535 errors if the name included a slash character. The /name= option
2536 now slurps the rest of the string, so it can include any characters
2537 but it must come last in the list of options (after /sender_retain
2540 PH/05 Some modifications to the interface to the fake nameserver for the new
2548 TK/01 Added the "success_on_redirect" address verification option. See
2549 NewStuff for rationale and an example.
2551 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2553 PH/02 Patch to exigrep to allow it to work on syslog lines.
2555 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2556 fread() to skip over the body file's header line, because in Cygwin the
2557 header line is locked and is inaccessible.
2559 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2560 co-exist for some time) to make it clear that it is the Exim ID that is
2561 referenced, not the Message-ID: header line.
2563 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2564 string_format() function, because snprintf() does not exist on all
2567 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2568 ${stat: expansion item.
2570 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2571 protocol synchronization error", to keep the pedants happy.
2573 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2574 well as for IRIX systems, when gcc is being used. See the host.c source
2577 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2579 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2582 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2583 transport and to the smtp transport in LMTP mode.
2585 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2587 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2588 run a filter in a subprocess. This could lead to confusion in subsequent
2589 lookups in the parent process. There should also be a search_tidyup() at
2590 the end of the subprocess.
2592 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2593 only if the host matched helo_try_verify_hosts, which caused the
2594 verification to occur when the EHLO/HELO command was issued. The ACL just
2595 tested the remembered result. Now, if a previous verification attempt has
2596 not happened, "verify = helo" does it there and then.
2598 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2600 TK/03 Fix log output including CR from clamd.
2602 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2603 exist provoked a memory error which could cause a segfault.
2605 PH/15 Installed PCRE 6.2
2607 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2609 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2610 of the problem. Specifically, suggested +O2 rather than +O1 for the
2613 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2615 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2616 record, the port was not correctly logged when the outgoing_port log
2617 selector was set (it logged the transort's default port).
2619 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2620 fallback_hosts, and "hosts" in the smtp transport.
2622 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2623 host errors such as "Connection refused".
2625 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2626 authentication with radiusclient 0.4.9:
2628 - Error returned from rc_read_config was caught wrongly
2629 - Username/password not passed on to radius server due to wrong length.
2631 The presumption is that some radiusclient API changes for 4.51/PH/17
2632 were not taken care of correctly. The code is still untested by me (my
2633 Linux distribution still has 0.3.2 of radiusclient), but it was
2634 contributed by a Radius user.
2636 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2637 expanding the "port" option of the smtp transport.
2639 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2640 while reading a MIME header. Thanks to Tom Hughes for a patch.
2642 PH/24 Include config.h inside local_scan.h so that configuration settings are
2645 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2646 that in an ACL for RCPT (for example), you can examine exactly what was
2649 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2650 commands, but it was not correctly comparing the address with the actual
2651 client host address. Thus, it would show the EHLO address in Received:
2652 header lines when this was not necessary.
2654 PH/27 Added the % operator to ${eval:}.
2656 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2657 it should be ignoring failures (because with -C, for example, it has lost
2658 privilege). It wasn't ignoring creation failures other than "already
2661 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2664 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2665 because an input line was too long, either on its own, or by virtue of
2666 too many continuations, the temporary file was not being removed, and the
2667 return code was incorrect.
2669 PH/31 Missing "BOOL" in function definition in filtertest.c.
2671 PH/32 Applied Sieve patches from the maintainer.
2673 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2675 PH/33 Added "verify = not_blind".
2677 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2678 Local/Makefile (with some defaults set). These are used in built scripts
2679 such as exicyclog, but they have never been used in the exim_install
2680 script (though there are many overriding facilities there). I have
2681 arranged that the exim_install script now takes note of these two
2684 PH/35 Installed configuration files for Dragonfly.
2686 PH/36 When a locally submitted message by a trusted user did not contain a
2687 From: header, and the sender address was obtained from -f or from an SMTP
2688 MAIL command, and the trusted user did not use -F to supply a sender
2689 name, $originator_name was incorrectly used when constructing a From:
2690 header. Furthermore, $originator_name was used for submission mode
2691 messages from external hosts without From: headers in a similar way,
2692 which is clearly wrong.
2694 PH/37 Added control=suppress_local_fixups.
2696 PH/38 When log_selector = +received_sender was set, and the addition of the
2697 sender made the log line's construction buffer exactly full, or one byte
2698 less than full, an overflow happened when the terminating "\n" was
2701 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2702 when the result of a list match is failure because a DNS lookup failed.
2704 PH/40 RM_COMMAND is now used in the building process.
2706 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2707 the "build-* directories that it finds.
2709 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2710 address in a domain literal was a prefix of an interface address.
2712 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2713 when verifying a sender address, unless rewrite_headers is false.
2715 PH/44 Wrote a long comment about why errors_to addresses are verified as
2716 recipients, not senders.
2718 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2719 the ratelimit ACL was added.
2721 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2723 PH/46 Added extra information about PostgreSQL errors to the error string.
2725 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2726 avoiding the need to install special zones in a real server. This is
2727 backwards compatible; if it can't find the fake resolver, it drops back.
2728 Thus, both old and new test suites can be run.
2730 TF/02 Added util/ratelimit.pl
2732 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2733 clock is set back in time.
2735 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2736 Candler <B.Candler@pobox.com>.
2738 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2739 OK for addresses that are the result of redirections.
2741 PH/48 A number of further additions for the benefit of the new test suite,
2742 including a fake gethostbyname() that interfaces to the fake DNS resolver
2745 TF/06 The fix for widen_domains has also been applied to qualify_single and
2746 search_parents which are the other dnslookup options that can cause
2749 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2752 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2754 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2755 service" if it encountered an MX record with an empty target hostname.
2756 The message is now "an MX or SRV record indicated no SMTP service".
2758 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2759 if the DNS of the sending site is misconfigured. This is quite a
2760 common situation. This change restores the behaviour of treating a
2761 helo verification defer as a failure.
2763 PH/51 If self=fail was set on a router, the bounce message did not include the
2764 actual error message.
2770 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2772 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2773 command itself ended in error, the underlying message about the transport
2774 filter timeout was being overwritten with the pipe command error. Now the
2775 underlying error message should be appended to the second error message.
2777 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2779 PH/02 Reduce the amount of output that "make" produces by default. Full output
2780 can still be requested.
2782 PH/03 The warning log line about a condition test deferring for a "warn" verb
2783 was being output only once per connection, rather than after each
2784 occurrence (because it was using the same function as for successful
2785 "warn" verbs). This seems wrong, so I have changed it.
2787 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2788 it should not have, which might have caused a crash in the right
2789 circumstances, but probably never did.
2791 PH/04 Installed a modified version of Tony Finch's patch to make submission
2792 mode fix the return path as well as the Sender: header line, and to
2793 add a /name= option so that you can make the user's friendly name appear
2796 TF/03 Added the control = fakedefer ACL modifier.
2798 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2799 Mark Lowes for thorough testing.
2801 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2803 TK/03 Merged latest SRS patch from Miles Wilton.
2805 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2806 with the definition in sysexits.h (which is #included earlier).
2807 Fortunately, Exim does not actually use EX_OK. The code used to try to
2808 preserve the sysexits.h value, by assumimg that macro definitions were
2809 scanned for macro replacements. I have been disabused of this notion,
2810 so now the code just undefines EX_OK before #including unistd.h.
2812 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2813 in the smtp transport. When a block could not be written in a single
2814 write() function, the timeout was being re-applied to each part-write.
2815 This seems wrong - if the receiver was accepting one byte at a time it
2816 would take for ever. The timeout is now adjusted when this happens. It
2817 doesn't have to be particularly precise.
2819 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2820 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2822 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2823 without a domain if the check to <postmaster@domain> fails.
2825 SC/01 Eximstats: added -xls and the ability to specify output files
2826 (patch written by Frank Heydlauf).
2828 SC/02 Eximstats: use FileHandles for outputing results.
2830 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2832 SC/04 Eximstats: fixed display of large numbers with -nvr option
2834 SC/05 Eximstats: fixed merging of reports with empty tables.
2836 SC/06 Eximstats: added the -include_original_destination flag
2838 SC/07 Eximstats: removed tabs and trailing whitespace.
2840 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2842 TK/06 MBOX spool code: Add real "From " MBOX separator line
2843 so the .eml file is really in mbox format (even though
2844 most programs do not really care). Patch from Alex Miller.
2846 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2847 The latter is generated from $received_to and is only set if the
2848 message has one envelope recipient. SA can use these headers,
2849 obviously out-of-the-box. Patch from Alex Miller.
2851 PH/08 The ${def test on a variable was returning false if the variable's
2852 value was "0", contrary to what the specification has always said!
2853 The result should be true unless the variable is empty.
2855 PH/09 The syntax error of a character other than { following "${if
2856 def:variable_name" (after optional whitespace) was not being diagnosed.
2857 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2858 accidental colon was present, for example, could give incorrect results.
2860 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2861 result is used (not including appendfile, where other changes are about
2864 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2865 This involved changing a lot of size variables from int to off_t. It
2866 should work with maildirs and everything.
2868 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2869 spamd dying while we are connected to it.
2871 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2874 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2875 function for BDB 4.3.
2877 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2879 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2880 the sources for PCRE have been re-arranged and more files are now
2883 PH/15 The code I had for printing potentially long long variables in PH/11
2884 above was not the best (it lost precision). The length of off_t variables
2885 is now inspected at build time, and an appropriate printing format (%ld
2886 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2887 to be "long long int" or "long int". This is needed for the internal
2888 formatting function string_vformat().
2890 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2891 the configuration file to be ":syslog", then the script "guesses" where
2892 the logs files are, rather than using the compiled in default. In our
2893 case the guess is not the same as the compiled default, so the script
2894 suddenly stopped working when I started to use syslog. The patch checks
2895 to see if log_file_path is "". If so, it attempts to read it from exim
2896 with no configuration file to get the compiled in version, before it
2897 falls back to the previous guessing code."
2899 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2900 implementing BATV in an Exim configuration. See NewStuff for the gory
2903 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2904 Makefile that are specific to HP-UX.
2906 PH/18 If the "use_postmaster" option was set for a recipient callout together
2907 with the "random" option, the postmaster address was used as the MAIL
2908 FROM address for the random test, but not for the subsequent recipient
2909 test. It is now used for both.
2911 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2912 patch removes a few documentation additions to RFC 3028, because the
2913 latest draft now contains them. It adds the new en;ascii-case comparator
2914 and a new error check for 8bit text in MIME parts. Comparator and
2915 require names are now matched exactly. I enabled the subaddress
2916 extension, but it is not well tested yet (read: it works for me)."
2918 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2919 rework some of the code of TK/09 above to avoid the hardwired use of
2920 "%lld" and "long long". Replaced the call to snprintf() with a call to
2923 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2924 records point to non-existent hosts", "retry timeout exceeded", and
2925 "retry time not reached for any host after a long failure period".
2927 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2928 experimental DomainKeys support:
2930 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2931 (2) On an error such as an illegally used "control", the wrong name for
2932 the control was given.
2934 These problems did NOT occur unless DomainKeys support was compiled.
2936 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2938 PH/24 Added ${if match_ip condition.
2940 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2941 databases so that it will be absolutely obvious if a crash occurs in the
2942 DB library. This is a regular occurrence (often caused by mis-matched
2945 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2946 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2947 picked up on a user's system that detects such things. There doesn't seem
2948 to be a gcc warning option for this - only an attribute that has to be
2949 put on the function's prototype. It seems that in Fedora Core 4 they have
2950 set this on a number of new functions. No doubt there will be more in due
2953 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2954 specify a transport. However, if an address that was verified by such a
2955 router was the subject of a callout, Exim crashed because it tried to
2956 read the rcpt_include_affixes from the non-existent transport. Now it
2957 just assumes that the setting of that option is false. This bug was
2958 introduced by 4.51/PH/31.
2960 PH/28 Changed -d+all to exclude +memory, because that information is very
2961 rarely of interest, but it makes the output a lot bigger. People tend to
2962 do -d+all out of habit.
2964 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2965 code in os-type was giving problems when libc.so lives in lib64, like on
2968 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2969 aren't the modern standard, and it seems that some systems' include files
2970 don't always have them. Exim was already checking for some of the newer
2971 ones like T_AAAA, and defining it itself. I've added checks for all the
2972 record types that Exim uses.
2974 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2975 not automatically generating a new one, as it is supposed to. This
2976 prevented TLS from working. If the file did exist, but contained invalid
2977 data, a new version was generated, as expected. It was only the case of a
2978 non-existent file that was broken.
2980 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2981 with a change in libdomainkeys > 0.64.
2983 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2984 from DNS. If the selector record carries the flag, it now has
2985 precedence over the domain-wide flag.
2987 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2989 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2990 the use of a port name that isn't defined in /etc/services) occurred, the
2991 message was deferred as in a normal delivery, and thus remained on the
2992 spool, instead of being failed because of the mua_wrapper setting. This
2993 is now fixed, and I tidied up some of the mua_wrapper messages at the
2996 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2997 the messages in a hash of arrays rather than using individual hashes.
2998 This is a bit cleaner and results in dramatic memory savings, albeit
2999 at a slight CPU cost.
3001 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3002 as requested by Marc Sherman.
3004 SC/10 Eximstats: added histograms for user specified patterns as requested
3007 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3009 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3010 fopen() in the content-scanning modules that did not already have it.
3016 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3017 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3019 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3021 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3023 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3024 to test Sieve filters that use "vacation".
3026 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3027 that changes the way the GnuTLS parameters are stored in the cache file.
3028 The new format can be generated externally. For backward compatibility,
3029 if the data in the cache doesn't make sense, Exim assumes it has read an
3030 old-format file, and it generates new data and writes a new file. This
3031 means that you can't go back to an older release without removing the
3034 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3035 work if there are any delivery delays because "one_time" forces the
3036 parent to be marked "delivered", so its unseen clone is never tried
3037 again. For this reason, Exim now forbids the simultaneous setting of
3040 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3041 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3042 entry if you want to know the details. The fix, however, neglected to
3043 consider the case when local delivery batching is involved. The test for
3044 "previously delivered" was not happening when checking to see if an
3045 address could be batched with a previous (undelivered) one; under
3046 certain circumstances this could lead to multiple deliveries to the same
3049 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3050 in its include files, and this causes problems building Exim.
3052 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3053 header_syntax) but Exim was just ignoring anything given after a slash.
3054 In particular, this caused confusion with an attempt to use "verify =
3055 reverse_host_lookup/defer_ok". An error is now given when options are
3056 supplied for verify items that do not have them. (Maybe reverse_host_
3057 lookup should have a defer_ok option, but that's a different point.)
3059 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3060 defined by RFC 821) to 2048, because there were problems with some AUTH
3061 commands, and RFC 1869 says the size should be increased for extended
3062 SMTP commands that take arguments.
3064 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3067 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3068 "unknown" error; now it says that the functionality isn't in the binary.
3070 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3071 an address' error message when a string expansion fails (syntax or
3072 whatever). Otherwise the password may appear in the log. Following change
3073 PH/42 below, there is no longer a chance of it appearing in a bounce
3076 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3078 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3079 of its addresses was ever tried. (Bugzilla bug #2.)
3081 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3082 the result incorrectly in the debug output. (It correctly added a newline
3083 to what was transported.)
3085 TF/01 Added $received_time.
3087 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3088 commented out examples of how to interface to a virus scanner and to
3089 SpamAssassin. Also added commented examples of av_scanner and
3090 spamd_address settings.
3092 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3093 and controls are allowed in which ACLs. There were a couple of minor
3094 errors. Some of the entries in the conditions table (which is a table of
3095 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3096 negation of where the condition IS allowed.
3098 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3100 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3101 header file does not have a version number, so I've had to invent a new
3102 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3103 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3104 radiusclient), but it was contributed by a Radius user.
3106 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3107 files or directories, for OpenSSL.
3109 PH/19 When an Exim process that is running as root has to create an Exim log
3110 file, it does so in a subprocess that runs as exim:exim so as to get the
3111 ownership right at creation (otherwise, other Exim processes might see
3112 the file with the wrong ownership). There was no test for failure of this
3113 fork() call, which would lead to the process getting stuck as it waited
3114 for a non-existent subprocess. Forks do occasionally fail when resources
3115 run out. I reviewed all the other calls to fork(); they all seem to check
3118 PH/20 When checking for unexpected SMTP input at connect time (before writing
3119 the banner), Exim was not dealing correctly with a non-positive return
3120 from the read() function. If the client had disconnected by this time,
3121 the result was a log entry for a synchronization error with an empty
3122 string after "input=" when read() returned zero. If read() returned -1
3123 (an event I could not check), uninitialized data bytes were printed.
3124 There were reports of junk text (parts of files, etc) appearing after
3127 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3129 PH/22 Added support for macro redefinition, and (re)definition in between
3130 driver and ACL definitions.
3132 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3133 forgetting to use the resulting value; it was using the unexpanded value.
3135 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3136 hadn't been configured. The fix is from Juergen Kreileder, who
3137 understands it better than I do:
3139 "Here's what I see happening with three configured cyrus_sasl
3140 authenticators configured (plain, login, cram-md5):
3142 On startup auth_cyrus_sasl_init() gets called for each of these.
3143 This means three calls to sasl_listmech() without a specified mech_list.
3144 => SASL tests which mechs of all available mechs actually work
3145 => three warnings about OTP not working
3146 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3148 With the patch, sasl_listmech() also gets called three times. But now
3149 SASL's mech_list option is set to the server_mech specified in the the
3150 authenticator. Or in other words, the answer from sasl_listmech()
3151 gets limited to just the mech you're testing for (which is different
3153 => the return list contains just 'plain' or 'login', 'cram-md5' or
3154 nothing depending on the value of ob->server_mech.
3156 I've just tested the patch: Authentication still works fine,
3157 unavailable mechs specified in the exim configuration are still
3158 caught, and the auth.log warnings about OTP are gone."
3160 PH/25 When debugging is enabled, the contents of the command line are added
3161 to the debugging output, even when log_selector=+arguments is not
3164 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3165 answer is "GNU", and only if the return is "GNU/something" is the answer
3168 PH/27 $acl_verify_message is now set immediately after the failure of a
3169 verification in an ACL, and so is available in subsequent modifiers. In
3170 particular, the message can be preserved by coding like this:
3172 warn !verify = sender
3173 set acl_m0 = $acl_verify_message
3175 Previously, $acl_verify_message was set only while expanding "message"
3176 and "log_message" when a very denied access.
3178 PH/28 Modified OS/os.c-Linux with
3180 -#ifndef OS_LOAD_AVERAGE
3181 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3183 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3184 nomenclature these days.)
3186 PH/29 Installed patch from the Sieve maintainer that adds the options
3187 sieve_useraddress and sieve_subaddress to the redirect router.
3189 PH/30 In these circumstances:
3190 . Two addresses routed to the same list of hosts;
3191 . First host does not offer TLS;
3192 . First host accepts first address;
3193 . First host gives temporary error to second address;
3194 . Second host offers TLS and a TLS session is established;
3195 . Second host accepts second address.
3196 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3197 and peerdn, if requested) that were in fact used only for the second
3200 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3201 attention to any local part prefix or suffix that was matched by the
3202 router that accepted the address. It now behaves in the same way as it
3203 does for delivery: the affixes are removed from the local part unless
3204 rcpt_include_affixes is set on the transport.
3206 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3207 timeout during the DATA phase of an incoming message.
3209 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3210 applied a patch sanctioned by the Sieve maintainer.
3212 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3213 the uid or gid is negative. A case of a negative gid caused this to be
3214 noticed. The fix allows for either to be negative.
3216 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3217 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3220 PH/36 The change PH/12 above was broken. Fixed it.
3222 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3223 the grounds that routing the same address twice would always produce the
3224 same answer. This might have been true once, but it is certainly no
3225 longer true now. Routing a child address may depend on the previous
3226 routing that produced that child. Some complicated redirection strategies
3227 went wrong when messages had multiple recipients, and made Exim's
3228 behaviour dependent on the order in which the addresses were given.
3230 I have moved the duplicate checking until after the routing is complete.
3231 Exim scans the addresses that are assigned to local and remote
3232 transports, and removes any duplicates. This means that more work will be
3233 done, as duplicates will always all be routed, but duplicates are
3234 presumably rare, so I don't expect this is of any significance.
3236 For deliveries to pipes, files, and autoreplies, the duplicate checking
3237 still happens during the routing process, since they are not going to be
3240 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3241 It corrects a timeout issue with spamd. This is Ian's comment: "The
3242 background is that sometimes spamd either never reads data from a
3243 connection it has accepted, or it never writes response data. The exiscan
3244 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3245 blindly assumes that writes won't block so it may never time out."
3247 PH/39 Allow G after quota size as well as K and M.
3249 PH/40 The value set for $authenticated_id in an authenticator may not contain
3250 binary zeroes or newlines because the value is written to log lines and
3251 to spool files. There was no check on this. Now the value is run through
3252 the string_printing() function so that such characters are converted to
3253 printable escape sequences.
3255 PH/41 $message_linecount is a new variable that contains the total number of
3256 lines in the message. Compare $body_linecount, which is the count for the
3259 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3260 bounce and delay warning messages, except in certain special cases, which
3263 (a) An SMTP error message from a remote host;
3264 (b) A message specified in a :fail: redirection;
3265 (c) A message specified in a "fail" command in a system filter;
3266 (d) A message specified in a FAIL return from the queryprogram router;
3267 (e) A message specified by the cannot_route_message router option.
3269 In these cases only, Exim does include the error details in bounce and
3270 warning messages. There are also a few cases where bland messages such
3271 as "unrouteable address" or "local delivery error" are given.
3273 PH/43 $value is now also set for the "else" part of a ${run expansion.
3275 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3276 being worked on, but at least Exim now implements the latest version to
3279 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3280 process to complete was treated as a delivery failure, a timeout while
3281 writing the message to the pipe was logged, but erroneously treated as a
3282 successful delivery. Such timeouts include transport filter timeouts. For
3283 consistency with the overall process timeout, these timeouts are now
3284 treated as errors, giving rise to delivery failures by default. However,
3285 there is now a new Boolean option for the pipe transport called
3286 timeout_defer, which, if set TRUE, converts the failures into defers for
3287 both kinds of timeout. A transport filter timeout is now identified in
3290 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3291 systems where "make" and "gmake" are different, calling "gmake" at top
3292 level broke things. I've arranged for the value of $(MAKE) to be passed
3293 from the Makefile to this script so that it can call the same version of
3297 A note about Exim versions 4.44 and 4.50
3298 ----------------------------------------
3300 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3301 changes of various kinds. As a consequence, a big documentation update was
3302 needed. This delayed the release for rather longer than seemed good, especially
3303 in the light of a couple of (minor) security issues. Therefore, the changes
3304 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3305 release. So 4.44 and 4.50 are in effect two different branches that both start
3308 I have left the 4.50 change log unchanged; it contains all the changes since
3309 4.43. The change log for 4.44 is below; many of its items are identical to
3310 those for 4.50. This seems to be the most sensible way to preserve the
3311 historical information.
3317 1. Minor wording change to the doc/README.SIEVE file.
3319 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3320 computation of the current number of files was incorrect.
3322 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3323 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3326 4. Give more explanation in the error message when the command for a transport
3327 filter fails to execute.
3329 5. There are several places where Exim runs a non-Exim command in a
3330 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3331 was being done only for the command run by the queryprogram router. It is
3332 now done for all such subprocesses. The other cases are: ${run, transport
3333 filters, and the commands run by the lmtp and pipe transports.
3335 6. Added CONFIGURE_GROUP build-time option.
3337 7. Some older OS have a limit of 256 on the maximum number of file
3338 descriptors. Exim was using setrlimit() to set 1000 as a large value
3339 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3340 systems. I've change it so that if it can't get 1000, it tries for 256.
3342 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3343 was an oversight, and furthermore, ever since the addition of extra
3344 controls (e.g. 4.43/32), the checks on when to allow different forms of
3345 "control" were broken. There should now be diagnostics for all cases when a
3346 control that does not make sense is encountered.
3348 9. Added the /retain_sender option to "control=submission".
3350 10. $recipients is now available in the predata ACL (oversight).
3352 11. Tidy the search cache before the fork to do a delivery from a message
3353 received from the command line. Otherwise the child will trigger a lookup
3354 failure and thereby defer the delivery if it tries to use (for example) a
3355 cached ldap connection that the parent has called unbind on.
3357 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3358 of $address_data from the recipient verification was clobbered by the
3359 sender verification.
3361 13. The value of address_data from a sender verification is now available in
3362 $sender_address_data in subsequent conditions in the ACL statement.
3364 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3366 15. Added a new option "connect=<time>" to callout options, to set a different
3369 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3370 was its contents. (It was OK if the option was not defined at all.)
3372 17. A "Completed" log line is now written for messages that are removed from
3373 the spool by the -Mrm option.
3375 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3376 information about exactly what failed.
3378 19. Added -dd to debug only the daemon process.
3380 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3381 handles timeouts, both on the server side and network timeouts. Renamed the
3382 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3384 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3385 It is now set to "smtps".
3387 22. $host_address is now set to the target address during the checking of
3388 ignore_target_hosts.
3390 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3391 being passed; this would have caused $sender_host_name to have been used if
3392 matching the list had actually called for a host name (not very likely,
3393 since this list is usually IP addresses). A host name is now passed as
3396 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3397 code that specifies a non-restarting handler (typically sigaction() in
3398 modern systems) in an attempt to fix a rare and obscure crash bug.
3400 25. Narrowed the window for a race in the daemon that could cause it to ignore
3401 SIGCHLD signals. This is not a major problem, because they are used only to
3402 wake it up if nothing else does.
3404 26. A malformed maildirsize file could cause Exim to calculate negative values
3405 for the mailbox size or file count. Odd effects could occur as a result.
3406 The maildirsize information is now recalculated if the size or filecount
3409 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3410 support for a long time. Removed HAVE_SYS_VFS_H.
3412 28. Installed the latest version of exipick from John Jetmore.
3414 29. In an address list, if the pattern was not a regular expression, an empty
3415 subject address (from a bounce message) matched only if the pattern was an
3416 empty string. Non-empty patterns were not even tested. This was the wrong
3417 because it is perfectly reasonable to use an empty address as part of a
3418 database query. An empty address is now tested by patterns that are
3419 lookups. However, all the other forms of pattern expect the subject to
3420 contain a local part and a domain, and therefore, for them, an empty
3421 address still always fails if the pattern is not itself empty.
3423 30. Exim went into a mad DNS loop when attempting to do a callout where the
3424 host was specified on an smtp transport, and looking it up yielded more
3425 than one IP address.
3427 31. Re-factored the code for checking spool and log partition space into a
3428 function that finds that data and another that does the check. The former
3429 is then used to implement four new variables: $spool_space, $log_space,
3430 $spool_inodes, and $log_inodes.
3432 32. The RFC2047 encoding function was originally intended for short strings
3433 such as real names; it was not keeping to the 75-character limit for
3434 encoded words that the RFC imposes. It now respects the limit, and
3435 generates multiple encoded words if necessary. To be on the safe side, I
3436 have increased the buffer size for the ${rfc2047: expansion operator from
3439 33. It is now permitted to omit both strings after an "if" condition; if the
3440 condition is true, the result is "true". As before, when the second string
3441 is omitted, a false condition yields an empty string. This makes it less
3442 cumbersome to write custom ACL and router conditions.
3444 34. Failure to deliver a bounce message always caused it to be frozen, even if
3445 there was an errors_to setting on the router. The errors_to setting is now
3448 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3449 canonical form (fully expanded) before being placed in
3450 $sender_host_address.
3452 36. The table in the code that translates DNS record types into text (T_A to
3453 "A" for instance) was missing entries for NS and CNAME. It is just possible
3454 that this could have caused confusion if both these types were looked up
3455 for the same domain, because the text type is used as part of Exim's
3456 per-process caching. But the chance of anyone hitting this buglet seems
3459 37. The dnsdb lookup has been extended in a number of ways.
3461 (1) There is a new type, "zns", which walks up the domain tree until it
3462 finds some nameserver records. It should be used with care.
3464 (2) There is a new type, "mxh", which is like "mx" except that it returns
3465 just the host names, not the priorities.
3467 (3) It is now possible to give a list of domains (or IP addresses) to be
3468 looked up. The behaviour when one of the lookups defers can be
3469 controlled by a keyword.
3471 (4) It is now possible to specify the separator character for use when
3472 multiple records are returned.
3474 38. The dnslists ACL condition has been extended: it is now possible to supply
3475 a list of IP addresses and/or domains to be looked up in a particular DNS
3478 39. Added log_selector=+queue_time_overall.
3480 40. When running the queue in the test harness, wait just a tad after forking a
3481 delivery process, to get repeatability of debugging output.
3483 41. Include certificate and key file names in error message when GnuTLS fails
3484 to set them up, because the GnuTLS error message doesn't include the name
3485 of the failing file when there is a problem reading it.
3487 42. Allow both -bf and -bF in the same test run.
3489 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3491 44. The "Exiscan patch" is now merged into the mainline Exim source.
3493 45. Sometimes the final signoff response after QUIT could fail to get
3494 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3495 < 0 before doing a fflush(). This bug looks as though it goes back to the
3496 introduction of TLS in release 3.20, but "sometimes" must have been rare
3497 because the tests only now provoked it.
3499 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3500 (this can affect the format of dates).
3502 47. exim_tidydb, when checking for the continued existence of a message for
3503 which it has found a message-specific retry record, was not finding
3504 messages that were in split spool directories. Consequently, it was
3505 deleting retry records that should have stayed in existence.
3507 48. Steve fixed some bugs in eximstats.
3509 49. The SPA authentication driver was not abandoning authentication and moving
3510 on to the next authenticator when an expansion was forced to fail,
3511 contradicting the general specification for all authenticators. Instead it
3512 was generating a temporary error. It now behaves as specified.
3514 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3515 (the order specifies the preference for clients). The order is now AES256,
3516 AES128, 3DES, ARCFOUR128.
3518 51. Small patch to Sieve code - explicitly set From: when generating an
3521 52. Exim crashed if a remote delivery caused a very long error message to be
3522 recorded - for instance if somebody sent an entire SpamAssassin report back
3523 as a large number of 550 error lines. This bug was coincidentally fixed by
3524 increasing the size of one of Exim's internal buffers (big_buffer) that
3525 happened as part of the Exiscan merge. However, to be on the safe side, I
3526 have made the code more robust (and fixed the comments that describe what
3529 53. Now that there can be additional text after "Completed" in log lines (if
3530 the queue_time_overall log selector is set), a one-byte patch to exigrep
3531 was needed to allow it to recognize "Completed" as not the last thing in
3534 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3535 patch that reportedly fixes this has been added. I am not expert enough to
3536 create a test for it. This is what the patch creator wrote:
3538 "I found a little strange behaviour of ldap code when working with
3539 Windows 2003 AD Domain, where users was placed in more than one
3540 Organization Units. When I tried to give exim partial DN, the exit code
3541 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3542 But simultaneously result of request was absolutely normal ldap result,
3543 so I produce this patch..."
3545 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3546 REFERENCE, so I have modified the code to exclude the patch when that macro
3549 55. Some experimental protocols are using DNS PTR records for new purposes. The
3550 keys for these records are domain names, not reversed IP addresses. The
3551 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3552 leaves it alone. Component reversal etc. now happens only for IP addresses.
3555 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3557 57. Double the size of the debug message buffer (to 2048) so that more of very
3558 long debug lines gets shown.
3560 58. The exicyclog utility now does better if the number of log files to keep
3561 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3563 59. Two changes related to the smtp_active_hostname option:
3565 (1) $smtp_active_hostname is now available as a variable.
3566 (2) The default for smtp_banner uses $smtp_active_hostname instead
3567 of $primary_hostname.
3569 60. The host_aton() function is supposed to be passed a string that is known
3570 to be a valid IP address. However, in the case of IPv6 addresses, it was
3571 not checking this. This is a hostage to fortune. Exim now panics and dies
3572 if the condition is not met. A case was found where this could be provoked
3573 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3574 components; fortuitously, this particular loophole had already been fixed
3575 by change 4.50/55 above.
3577 If there are any other similar loopholes, the new check in host_aton()
3578 itself should stop them being exploited. The report I received stated that
3579 data on the command line could provoke the exploit when Exim was running as
3580 exim, but did not say which command line option was involved. All I could
3581 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3582 running as the user.
3585 61. There was a buffer overflow vulnerability in the SPA authentication code
3586 (which came originally from the Samba project). I have added a test to the
3587 spa_base64_to_bits() function which I hope fixes it.
3590 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3591 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3593 63. The daemon start-up calls getloadavg() while still root for those OS that
3594 need the first call to be done as root, but it missed one case: when
3595 deliver_queue_load_max is set with deliver_drop_privilege. This is
3596 necessary for the benefit of the queue runner, because there is no re-exec
3597 when deliver_drop_privilege is set.
3599 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3600 This has been fixed.
3602 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3603 was in use, was not putting the data itself into the right store pool;
3604 consequently, it could be overwritten for a subsequent message in the same
3605 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3608 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3610 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3611 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3612 the function were treating the return as a boolean value, which happened to
3613 work because 0=false and not-0=true, but is not correct code.
3615 68. The host_aton() function was not handling scoped IPv6 addresses (those
3616 with, for example, "%eth0" on the end) correctly.
3618 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3619 negated items (that is, ~something) in unsigned ints. Some compilers
3620 apparently mutter when there is no cast.
3622 70. If an address verification called from an ACL failed, and did not produce a
3623 user-specific message (i.e. there was only a "system" message), nothing was
3624 put in $acl_verify_message. In this situation, it now puts the system
3627 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3628 session; change 4.31/43 added the unwanted input to the log line - except
3629 that it did not do this in the start of session case. It now does.
3631 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3632 This is wrong and can cause the other end to generate a synchronization
3633 error if it is another Exim or anything else that does the synchronization
3634 check. A QUIT command is no longer sent after a timeout.
3636 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3637 during host lookups.
3639 74. The defer_ok option of callout verification was not working if it was used
3640 when verifying addresses in header lines, that is, for this case:
3642 verify = header_sender/callout=defer_ok
3644 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3645 those file descriptors could be used for SMTP connections. If anything
3646 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3647 could be sent to the SMTP client, causing chaos. The daemon now opens
3648 stdin, stdout, and stderr to /dev/null when it puts itself into the
3651 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3652 log by default. The user can override this with suitable Perl magic.
3654 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3655 the log message when discard triggers, was not working for the DATA ACL or
3656 for the non-SMTP ACL.
3658 78. Error message wording change in sieve.c.
3660 79. If smtp_accept_max_per_host was set, the number of connections could be
3661 restricted to fewer than expected, because the daemon was trying to set up
3662 a new connection before checking whether the processes handling previous
3663 connections had finished. The check for completed processes is now done
3664 earlier. On busy systems, this bug wouldn't be noticed because something
3665 else would have woken the daemon, and it would have reaped the completed
3668 80. If a message was submitted locally by a user whose login name contained one
3669 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3670 It caused a spool format error. I have fixed the spool reading code. A
3671 related problem was that the "from" clause in the Received: line became
3672 illegal because of the space(s). It is now covered by ${quote_local_part.
3674 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3677 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3679 83. Patch from Sieve maintainer to fix unterminated string problem in
3680 "vacation" handling.
3682 84. Some minor changes to the Linux configuration files to help with other
3683 OS variants using glibc.
3685 85. One more patch for Sieve to update vacation handling to latest spec.
3688 ----------------------------------------------------
3689 See the note above about the 4.44 and 4.50 releases.
3690 ----------------------------------------------------
3696 1. Change 4.43/35 introduced a bug that caused file counts to be
3697 incorrectly computed when quota_filecount was set in an appendfile
3700 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3701 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3704 3. Give more explanation in the error message when the command for a transport
3705 filter fails to execute.
3707 4. There are several places where Exim runs a non-Exim command in a
3708 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3709 was being done only for the command run by the queryprogram router. It is
3710 now done for all such subprocesses. The other cases are: ${run, transport
3711 filters, and the commands run by the lmtp and pipe transports.
3713 5. Some older OS have a limit of 256 on the maximum number of file
3714 descriptors. Exim was using setrlimit() to set 1000 as a large value
3715 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3716 systems. I've change it so that if it can't get 1000, it tries for 256.
3718 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3719 was an oversight, and furthermore, ever since the addition of extra
3720 controls (e.g. 4.43/32), the checks on when to allow different forms of
3721 "control" were broken. There should now be diagnostics for all cases when a
3722 control that does not make sense is encountered.
3724 7. $recipients is now available in the predata ACL (oversight).
3726 8. Tidy the search cache before the fork to do a delivery from a message
3727 received from the command line. Otherwise the child will trigger a lookup
3728 failure and thereby defer the delivery if it tries to use (for example) a
3729 cached ldap connection that the parent has called unbind on.
3731 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3732 of $address_data from the recipient verification was clobbered by the
3733 sender verification.
3735 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3736 was its contents. (It was OK if the option was not defined at all.)
3738 11. A "Completed" log line is now written for messages that are removed from
3739 the spool by the -Mrm option.
3741 12. $host_address is now set to the target address during the checking of
3742 ignore_target_hosts.
3744 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3745 being passed; this would have caused $sender_host_name to have been used if
3746 matching the list had actually called for a host name (not very likely,
3747 since this list is usually IP addresses). A host name is now passed as
3750 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3751 code that specifies a non-restarting handler (typically sigaction() in
3752 modern systems) in an attempt to fix a rare and obscure crash bug.
3754 15. Narrowed the window for a race in the daemon that could cause it to ignore
3755 SIGCHLD signals. This is not a major problem, because they are used only to
3756 wake it up if nothing else does.
3758 16. A malformed maildirsize file could cause Exim to calculate negative values
3759 for the mailbox size or file count. Odd effects could occur as a result.
3760 The maildirsize information is now recalculated if the size or filecount
3763 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3764 support for a long time. Removed HAVE_SYS_VFS_H.
3766 18. Updated exipick to current release from John Jetmore.
3768 19. Allow an empty sender to be matched against a lookup in an address list.
3769 Previously the only cases considered were a regular expression, or an
3772 20. Exim went into a mad DNS lookup loop when doing a callout where the
3773 host was specified on the transport, if the DNS lookup yielded more than
3776 21. The RFC2047 encoding function was originally intended for short strings
3777 such as real names; it was not keeping to the 75-character limit for
3778 encoded words that the RFC imposes. It now respects the limit, and
3779 generates multiple encoded words if necessary. To be on the safe side, I
3780 have increased the buffer size for the ${rfc2047: expansion operator from
3783 22. Failure to deliver a bounce message always caused it to be frozen, even if
3784 there was an errors_to setting on the router. The errors_to setting is now
3787 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3788 canonical form (fully expanded) before being placed in
3789 $sender_host_address.
3791 24. Updated eximstats to version 1.33
3793 25. Include certificate and key file names in error message when GnuTLS fails
3794 to set them up, because the GnuTLS error message doesn't include the name
3795 of the failing file when there is a problem reading it.
3797 26. Expand error message when OpenSSL has problems setting up cert/key files.
3800 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3801 (this can affect the format of dates).
3803 28. exim_tidydb, when checking for the continued existence of a message for
3804 which it has found a message-specific retry record, was not finding
3805 messages that were in split spool directories. Consequently, it was
3806 deleting retry records that should have stayed in existence.
3808 29. eximstats updated to version 1.35
3809 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3810 1.35 - bugfix such that pie charts by volume are generated correctly
3812 30. The SPA authentication driver was not abandoning authentication and moving
3813 on to the next authenticator when an expansion was forced to fail,
3814 contradicting the general specification for all authenticators. Instead it
3815 was generating a temporary error. It now behaves as specified.
3817 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3818 (the order specifies the preference for clients). The order is now AES256,
3819 AES128, 3DES, ARCFOUR128.
3821 31. Small patch to Sieve code - explicitly set From: when generating an
3824 32. Exim crashed if a remote delivery caused a very long error message to be
3825 recorded - for instance if somebody sent an entire SpamAssassin report back
3826 as a large number of 550 error lines. This bug was coincidentally fixed by
3827 increasing the size of one of Exim's internal buffers (big_buffer) that
3828 happened as part of the Exiscan merge. However, to be on the safe side, I
3829 have made the code more robust (and fixed the comments that describe what
3832 33. Some experimental protocols are using DNS PTR records for new purposes. The
3833 keys for these records are domain names, not reversed IP addresses. The
3834 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3835 leaves it alone. Component reversal etc. now happens only for IP addresses.
3838 34. The host_aton() function is supposed to be passed a string that is known
3839 to be a valid IP address. However, in the case of IPv6 addresses, it was
3840 not checking this. This is a hostage to fortune. Exim now panics and dies
3841 if the condition is not met. A case was found where this could be provoked
3842 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3843 components; fortuitously, this particular loophole had already been fixed
3844 by change 4.50/55 or 4.44/33 above.
3846 If there are any other similar loopholes, the new check in host_aton()
3847 itself should stop them being exploited. The report I received stated that
3848 data on the command line could provoke the exploit when Exim was running as
3849 exim, but did not say which command line option was involved. All I could
3850 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3851 running as the user.
3854 35. There was a buffer overflow vulnerability in the SPA authentication code
3855 (which came originally from the Samba project). I have added a test to the
3856 spa_base64_to_bits() function which I hope fixes it.
3859 36. The daemon start-up calls getloadavg() while still root for those OS that
3860 need the first call to be done as root, but it missed one case: when
3861 deliver_queue_load_max is set with deliver_drop_privilege. This is
3862 necessary for the benefit of the queue runner, because there is no re-exec
3863 when deliver_drop_privilege is set.
3865 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3866 was in use, was not putting the data itself into the right store pool;
3867 consequently, it could be overwritten for a subsequent message in the same
3868 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3871 38. Sometimes the final signoff response after QUIT could fail to get
3872 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3873 < 0 before doing a fflush(). This bug looks as though it goes back to the
3874 introduction of TLS in release 3.20, but "sometimes" must have been rare
3875 because the tests only now provoked it.
3881 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3882 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3883 Now it returns an integer. A number of calls were still expecting a T/F
3884 return. Fortuitously, in all cases, the tests worked in OK situations,
3885 which is the norm. However, things would have gone wrong on any write
3886 failures on the smtp file descriptor. This function is used when sending
3887 messages over SMTP and also when doing verify callouts.
3889 2. When Exim is called to do synchronous delivery of a locally submitted
3890 message (the -odf or -odi options), it no longer closes stderr before doing
3893 3. Implemented the mua_wrapper option.
3895 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3897 5. Implemented the functions header_remove(), header_testname(),
3898 header_add_at_position(), and receive_remove_recipient(), and exported them
3901 6. If an ACL "warn" statement specified the addition of headers, Exim already
3902 inserted X-ACL-Warn: at the start if there was no header name. However, it
3903 was not making this test for the second and subsequent header lines if
3904 there were newlines in the string. This meant that an invalid header could
3905 be inserted if Exim was badly configured.
3907 7. Allow an ACL "warn" statement to add header lines at the start or after all
3908 the Received: headers, as well as at the end.
3910 8. Added the rcpt_4xx retry error code.
3912 9. Added postmaster_mailfrom=xxx to callout verification option.
3914 10. Added mailfrom=xxxx to the callout verification option, for verify=
3917 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3918 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3920 12. Inserted some casts to stop certain compilers warning when using pointer
3921 differences as field lengths or precisions in printf-type calls (mostly
3922 affecting debugging statements).
3924 13. Added optional readline() support for -be (dynamically loaded).
3926 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3927 same clock tick as a message's arrival, so that its received time was the
3928 same as the "first fail" time on the retry record, and that message
3929 remained on the queue past the ultimate address timeout, every queue runner
3930 would try a delivery (because it was past the ultimate address timeout) but
3931 after another failure, the ultimate address timeout, which should have then
3932 bounced the address, did not kick in. This was a "< instead of <=" error;
3933 in most cases the first failure would have been in the next clock tick
3934 after the received time, and all would be well.
3936 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3937 being recognized when the domain list was tested by the match_domain
3938 condition in an expansion string.
3940 16. Added the ${str2b64: operator.
3942 17. Exim was always calling setrlimit() to set a large limit for the number of
3943 processes, without checking whether the existing limit was already
3944 adequate. (It did check for the limit on file descriptors.) Furthermore,
3945 errors from getrlimit() and setrlimit() were being ignored. Now they are
3946 logged to the main and panic logs, but Exim does carry on, to try to do its
3947 job under whatever limits there are.
3949 18. Imported PCRE 5.0.
3951 19. Trivial typo in log message " temporarily refused connection" (the leading
3954 20. If the log selector return_path_on_delivery was set and an address was
3955 redirected to /dev/null, the delivery process crashed because it assumed
3956 that a return path would always be set for a "successful" delivery. In this
3957 case, the whole delivery is bypassed as an optimization, and therefore no
3960 21. Internal re-arrangement: the function for sending a challenge and reading
3961 a response while authentication was assuming a zero-terminated challenge
3962 string. It's now changed to take a pointer and a length, to allow for
3963 binary data in such strings.
3965 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3967 23. Exim was not respecting finduser_retries when seeking the login of the
3968 uid under which it was called; it was always trying 10 times. (The default
3969 setting of finduser_retries is zero.) Also, it was sleeping after the final
3970 failure, which is pointless.
3972 24. Implemented tls_on_connect_ports.
3974 25. Implemented acl_smtp_predata.
3976 26. If the domain in control=submission is set empty, Exim assumes that the
3977 authenticated id is a complete email address when it generates From: or
3978 Sender: header lines.
3980 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3981 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3982 chown and chgrp in /bin and hostname in /usr/bin.
3984 28. Exim was keeping the "process log" file open after each use, just as it
3985 does for the main log. This opens the possibility of it remaining open for
3986 long periods when the USR1 signal hits a daemon. Occasional processlog
3987 errors were reported, that could have been caused by this. Anyway, it seems
3988 much more sensible not to leave this file open at all, so that is what now
3991 29. The long-running daemon process does not normally write to the log once it
3992 has entered its main loop, and it closes the log before doing so. This is
3993 so that log files can straightforwardly be renamed and moved. However,
3994 there are a couple of unusual error situations where the daemon does write
3995 log entries, and I had neglected to close the log afterwards.
3997 30. The text of an SMTP error response that was received during a remote
3998 delivery was being truncated at 512 bytes. This is too short for some of
3999 the long messages that one sometimes sees. I've increased the limit to
4002 31. It is now possible to make retry rules that apply only when a message has a
4003 specific sender, in particular, an empty sender.
4005 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4006 it possible to be selective about when SMTP synchronization is enforced.
4008 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4010 32. Implemented hosts_connection_nolog.
4012 33. Added an ACL for QUIT.
4014 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4017 35. Added mailbox_size and mailbox_filecount to appendfile.
4019 36. Added control = no_multiline_responses to ACLs.
4021 37. There was a bug in the logic of the code that waits for the clock to tick
4022 in the case where the clock went backwards by a substantial amount such
4023 that the microsecond fraction of "now" was more than the microsecond
4024 fraction of "then" (but the whole seconds number was less).
4026 38. Added support for the libradius Radius client library this is found on
4027 FreeBSD (previously only the radiusclient library was supported).
4033 1. When certain lookups returned multiple values in the form name=value, the
4034 quoting of the values was not always being done properly. Specifically:
4035 (a) If the value started with a double quote, but contained no whitespace,
4037 (b) If the value contained whitespace other than a space character (i.e.
4038 tabs or newlines or carriage returns) it was not quoted.
4039 This fix has been applied to the mysql and pgsql lookups by writing a
4040 separate quoting function and calling it from the lookup code. The fix
4041 should probably also be applied to nisplus, ibase and oracle lookups, but
4042 since I cannot test any of those, I have not disturbed their existing code.
4044 2. A hit in the callout cache for a specific address caused a log line with no
4045 reason for rejecting RCPT. Now it says "Previous (cached) callout
4046 verification failure".
4048 3. There was an off-by-one bug in the queryprogram router. An over-long
4049 return line was truncated at 256 instead of 255 characters, thereby
4050 overflowing its buffer with the terminating zero. As well as fixing this, I
4051 have increased the buffer size to 1024 (and made a note to document this).
4053 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4054 when Exim is waiting for an SMTP response from a remote server, Exim
4055 restarts its select() call on the socket, thereby resetting its timeout.
4056 This is not a problem when such interrupts are rare. Somebody set up a cron
4057 job to run exiwhat every 2 minutes, which is less than the normal select()
4058 timeout (5 or 10 minutes). This meant that the select() timeout never
4059 kicked in because it was always reset. I have fixed this by comparing the
4060 time when an interrupt arrives with the time at the start of the first call
4061 to select(). If more time than the timeout has elapsed, the interrupt is
4062 treated as a timeout.
4064 5. Some internal re-factoring in preparation for the addition of Sieve
4065 extensions (by MH). In particular, the "personal" test is moved to a
4066 separate function, and given an option for scanning Cc: and Bcc: (which is
4067 not set for Exim filters).
4069 6. When Exim created an email address using the login of the caller as the
4070 local part (e.g. when creating a From: or Sender: header line), it was not
4071 quoting the local part when it contained special characters such as @.
4073 7. Installed new OpenBSD configuration files.
4075 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4076 try to make them clearer.
4078 9. Callout options, other than the timeout value, were being ignored when
4079 verifying sender addresses in header lines. For example, when using
4081 verify = header_sender/callout=no_cache
4083 the cache was (incorrectly) being used.
4085 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4086 only the Cygwin environment.
4088 11. When return_path_on_delivery was set as a log selector, if different remote
4089 addresses in the same message used different return paths and parallel
4090 remote delivery occurred, the wrong values would sometimes be logged.
4091 (Whenever a remote delivery process finished, the return path value from
4092 the most recently started remote delivery process was logged.)
4094 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4095 lines when AUTH and/or TLS are in use. This is the "received protocol"
4096 field. Exim used to use "asmtp" for authenticated SMTP, without any
4097 indication (in the protocol name) for TLS use. Now it follows the RFC and
4098 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4099 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4100 names appear in log lines as well as in Received: header lines.
4102 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4103 extensions, and comparison tests, and to fix some bugs.
4105 14. Changes to the "personal" filter test:
4107 (1) The test was buggy in that it was just doing the equivalent of
4108 "contains" tests on header lines. For example, if a user's address was
4109 anne@some.where, the "personal" test would incorrectly be true for
4111 To: susanne@some.where
4113 This test is now done by extracting each address from the header in turn,
4114 and checking the entire address. Other tests that are part of "personal"
4115 are now done using regular expressions (for example, to check local parts
4116 of addresses in From: header lines).
4118 (2) The list of non-personal local parts in From: addresses has been
4119 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4120 taken from the Sieve specification recommendations.
4122 (3) If the message contains any header line starting with "List-" it is
4123 treated as non-personal.
4125 (4) The test for "circular" in the Subject: header line has been removed
4126 because it now seems ill-conceived.
4128 15. Minor typos in src/EDITME comments corrected.
4130 16. Installed latest exipick from John Jetmore.
4132 17. If headers_add on a router specified a text string that was too long for
4133 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4134 of string_sprintf() is now avoided.
4136 18. $message_body_size was not set (it was always zero) when running the DATA
4137 ACL and the local_scan() function.
4139 19. For the "mail" command in an Exim filter, no default was being set for
4140 the once_repeat time, causing a random time value to be used if "once" was
4141 specified. (If the value happened to be <= 0, no repeat happened.) The
4142 default is now 0s, meaning "never repeat". The "vacation" command was OK
4143 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4144 (I found it when inspecting the code).
4146 20. There is now an overall timeout for performing a callout verification. It
4147 defaults to 4 times the callout timeout, which applies to individual SMTP
4148 commands during the callout. The overall timeout applies when there is more
4149 than one host that can be tried. The timeout is checked before trying the
4150 next host. This prevents very long delays if there are a large number of
4151 hosts and all are timing out (e.g. when the network connections are timing
4152 out). The value of the overall timeout can be changed by specifying an
4153 additional sub-option for "callout", called "maxwait". For example:
4155 verify = sender/callout=5s,maxwait=20s
4157 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4158 to the end before writing, but this should make it even safer).
4160 22. Exim was forgetting that it had advertised PIPELINING for the second and
4161 subsequent messages on an SMTP connection. It was also not resetting its
4162 memory on STARTTLS and an internal HELO.
4164 23. When Exim logs an SMTP synchronization error within a session, it now
4165 records whether PIPELINING has been advertised or not.
4167 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4168 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4169 rather than long int.
4171 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4173 26. Added the never_mail option to autoreply.
4179 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4180 crash if the getsockname() call failed; this can happen if a connection is
4181 closed very soon after it is established. The problem was simply in the
4182 order in which certain operations were done, causing Exim to try to write
4183 to the SMTP stream before it had set up the file descriptor. The bug has
4184 been fixed by making things happen in the correct order.
4190 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4191 before the connection was closed, thus losing the rejection response.
4193 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4194 some early Solaris releases, but causes trouble in current releases where
4195 socklen_t is defined.
4197 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4200 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4203 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4204 that releasing the top of it at the end releases what was used for sub-
4205 expansions (unless the block got too big). However, discard this block if
4206 the first thing is a variable or header, so that we can use its block when
4207 it is dynamic (useful for very large $message_headers, for example).
4209 6. Lookups now cache *every* query, not just the most recent. A new, separate
4210 store pool is used for this. It can be recovered when all lookup caches are
4211 flushed. Lookups now release memory at the end of their result strings.
4212 This has involved some general refactoring of the lookup sources.
4214 7. Some code has been added to the store_xxx() functions to reduce the amount
4215 of flapping under certain conditions.
4217 8. log_incoming_interface used to affect only the <= reception log lines. Now
4218 it causes the local interface and port to be added to several more SMTP log
4219 lines, for example "SMTP connection from", and rejection lines.
4221 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4223 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4225 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4226 could be overwritten at the end of the current message (or the start of a
4227 new message if it was set in a HELO ACL). The value is now preserved for
4228 the duration of the SMTP connection.
4230 12. If a transport had a headers_rewrite setting, and a matching header line
4231 contained an unqualified address, that address was qualified, even if it
4232 did not match any rewriting rules. The underlying bug was that the values
4233 of the flags that permit the existence of unqualified sender and recipient
4234 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4235 non-local messages, and by -bnq for local messages) were not being
4236 preserved with the message after it was received.
4238 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4239 "next input=" as part of the text comprising the host identity instead of
4240 the correct text. The code was using the same buffer for two different
4241 strings. However, depending on which order the printing function evaluated
4242 its arguments, the bug did not always show up. Under Linux, for example, my
4243 test suite worked just fine.
4245 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4246 doesn't work with some older versions of Perl. It has been changed to "my",
4247 which in any case is probably the better facility to use.
4249 15. A really picky compiler found some instances of statements for creating
4250 error messages that either had too many or two few arguments for the format
4253 16. The size of the buffer for calls to the DNS resolver has been increased
4254 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4255 for addresses that have a lot of PTR records. This alleviates a problem; it
4256 does not fully solve it.
4258 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4259 buffer now truncates the list and logs the incident, which is the same
4260 action as happens when Exim is looking up a host name and its aliases.
4261 Previously in this situation something unpredictable would happen;
4262 sometimes it was "internal error: store_reset failed".
4264 18. If a server dropped the connection unexpectedly when an Exim client was
4265 using GnuTLS and trying to read a response, the client delivery process
4266 crashed while trying to generate an error log message.
4268 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4269 string, for example:
4271 warn message = H1: something\nH2: something
4273 the text was added as a single header line from Exim's point of view
4274 though it ended up OK in the delivered message. However, searching for the
4275 second and subsequent header lines using $h_h2: did not work. This has been
4276 fixed. Similarly, if a system filter added multiple headers in this way,
4277 the routers could not see them.
4279 20. Expanded the error message when iplsearch is called with an invalid key to
4280 suggest using net-iplsearch in a host list.
4282 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4283 ACLs are no longer actually imposed (and a message to that effect is
4286 22. If a "gecos" field in a passwd entry contained escaped characters, in
4287 particular, if it contained a \" sequence, Exim got it wrong when building
4288 a From: or a Sender: header from that name. A second bug also caused
4289 incorrect handling when an unquoted " was present following a character
4290 that needed quoting.
4292 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4293 was not being matched caselessly.
4295 24. Arranged for all hyphens in the exim.8 source to be escaped with
4298 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4299 properly. Recipient callout cache records were still being keyed to include
4300 the sender, even when use_sender was set false. This led to far more
4301 callouts that were necessary. The sender is no longer included in the key
4302 when use_sender is false.
4304 26. Added "control = submission" modifier to ACLs.
4306 27. Added the ${base62d: operator to decode base 62 numbers.
4308 28. dnsdb lookups can now access SRV records.
4310 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4311 the configuration file.
4313 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4314 (-v) mode. This makes the output for a verbose queue run more intelligible.
4316 31. Added a use_postmaster feature to recipient callouts.
4318 32. Added the $body_zerocount variable, containing the number of binary zero
4319 bytes in the message body.
4321 33. The time of last modification of the "new" subdirectory is now used as the
4322 "mailbox time last read" when there is a quota error for a maildir
4325 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4327 35. Added +ignore_unknown as a special item in host lists.
4329 36. Code for decoding IPv6 addresses in host lists is now included, even if
4330 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4331 address was recognized as an IP address, but was then not correctly decoded
4332 into binary, causing unexpected and incorrect effects when compared with
4339 1. Very minor rewording of debugging text in manualroute to say "list of
4340 hosts" instead of "hostlist".
4342 2. If verify=header_syntax was set, and a header line with an unqualified
4343 address (no domain) and a large number of spaces between the end of the
4344 name and the colon was received, the reception process suffered a buffer
4345 overflow, and (when I tested it) crashed. This was caused by some obsolete
4346 code that should have been removed. The fix is to remove it!
4348 3. When running in the test harness, delay a bit after writing a bounce
4349 message to get a bit more predictability in the log output.
4351 4. Added a call to search_tidyup() just before forking a reception process. In
4352 theory, someone could use a lookup in the expansion of smtp_accept_max_
4353 per_host which, without the tidyup, could leave open a database connection.
4355 5. Added the variables $recipient_data and $sender_data which get set from a
4356 lookup success in an ACL "recipients" or "senders" condition, or a router
4357 "senders" option, similar to $domain_data and $local_part_data.
4359 6. Moved the writing of debug_print from before to after the "senders" test
4362 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4363 problems for message scanning, either using a data ACL, or using
4364 local_scan() because the Received: header was not generated till after they
4365 were called (in order to set the time as the time of reception completion).
4366 I have revised the way this works. The header is now generated after the
4367 body is received, but before the ACL or local_scan() are called. After they
4368 are run, the timestamp in the header is updated.
4374 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4375 before starting a queue runner without re-exec. This happened only when
4376 deliver_drop_privilege was set or when the Exim user was set to root. The
4377 effect of the bug was that timeouts during subsequent deliveries caused
4378 crashes instead of being properly handled. The handler is now left at its
4379 default (and expected) setting.
4381 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4382 message, again when deliver_drop_privilege is set or Exim is run as root.
4383 The bug described in (1) was not present in this case, but the tidying up
4384 of the other signals was missing. I have made the two cases consistent.
4386 3. The ignore_target_hosts setting on a manualroute router was being ignored
4387 for hosts that were looked up using the /MX notation.
4389 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4392 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4393 operated on the sender address. After changing the $sender_address to <>
4394 for the sender address verify, Exim was re-instated it as the original
4395 (before rewriting) address, but remembering that it had rewritten it, so it
4396 wasn't rewriting it again. This bug also had the effect of breaking the
4397 sender address verification caching when the sender address was rewritten.
4399 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4400 This has been changed so that if the ip literal address matches
4401 ignore_target_hosts, the router declines.
4403 7. Added expansion conditions match_domain, match_address, and match_local_
4404 part (NOT match_host).
4406 8. The placeholder for the Received: header didn't have a length field set.
4408 9. Added code to Exim itself and to exim_lock to test for a specific race
4409 condition that could lead to file corruption when using MBX delivery. The
4410 issue is with the lockfile that is created in /tmp. If this file is removed
4411 after a process has opened it but before that process has acquired a lock,
4412 there is the potential for a second process to recreate the file and also
4413 acquire a lock. This could lead to two Exim processes writing to the file
4414 at the same time. The added code performs the same test as UW imapd; it
4415 checks after acquiring the lock that its file descriptor still refers to
4416 the same named file.
4418 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4419 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4422 11. Added the smtp_active_hostname option. If used, this will typically be made
4423 to depend on the incoming interface address. Because $interface_address is
4424 not set up until the daemon has forked a reception process, error responses
4425 that can happen earlier (such as "too many connections") no longer contain
4428 12. If an expansion in a condition on a "warn" statement fails because a lookup
4429 defers, the "warn" statement is abandoned, and the next ACL statement is
4430 processed. Previously this caused the whole ACL to be aborted.
4432 13. Added the iplsearch lookup type.
4434 14. Added ident_timeout as a log selector.
4436 15. Added tls_certificate_verified as a log selector.
4438 16. Added a global option tls_require_ciphers (compare the smtp transport
4439 option of the same name). This controls incoming TLS connections.
4441 17. I finally figured out how to make tls_require_ciphers do a similar thing
4442 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4443 before starting the TLS session.
4445 18. Tabs are now shown as \t in -bP output.
4447 19. If the log selector return_path_on_delivery was set, Exim crashed when
4448 bouncing a message because it had too many Received: header lines.
4450 20. If two routers both had headers_remove settings, and the first one included
4451 a superfluous trailing colon, the final name in the first list and the
4452 first name in the second list were incorrectly joined into one item (with a
4453 colon in the middle).
4459 1. Added -C and -D options to the exinext utility, mainly to make it easier
4460 to include in the automated testing, but these could be helpful when
4461 multiple configurations are in use.
4463 2. The exinext utility was not formatting the output nicely when there was
4464 an alternate port involved in the retry record key, nor when there was a
4465 message id as well (for retries that were specific to a specific message
4466 and a specific host). It was also confused by IPv6 addresses, because of
4467 the additional colons they contain. I have fixed the IPv4 problem, and
4468 patched it up to do a reasonable job for IPv6.
4470 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4471 delivery, the log line now contains "pipelined" if PIPELINING was used.
4473 4. An SMTP transport process used to panic and die if the bind() call to set
4474 an explicit outgoing interface failed. This has been changed; it is now
4475 treated in the same way as a connect() failure.
4477 5. A reference to $sender_host_name in the part of a conditional expansion
4478 that was being skipped was still causing a DNS lookup. This no longer
4481 6. The def: expansion condition was not recognizing references to header lines
4482 that used bh_ and bheader_.
4484 7. Added the _cache feature to named lists.
4486 8. The code for checking quota_filecount in the appendfile transport was
4487 allowing one more file than it should have been.
4489 9. For compatibility with Sendmail, the command line option
4497 and sets the incoming protocol and host name (for trusted callers). The
4498 host name and its colon can be omitted when only the protocol is to be set.
4499 Note the Exim already has two private options, -pd and -ps, that refer to
4500 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4501 "s", but I don't think that's a major issue.
4503 10. A number of refactoring changes to the code, none of which should affect
4506 (a) The number of logging options was getting close to filling up the
4507 32-bit word that was used as a bit map. I have split them into two classes:
4508 those that are passed in the argument to log_write(), and those that are
4509 only ever tested independently outside of that function. These are now in
4510 separate 32-bit words, so there is plenty of room for expansion again.
4511 There is no change in the user interface or the logging behaviour.
4513 (b) When building, for example, log lines, the code previously used a
4514 macro that called string_cat() twice, in order to add two strings. This is
4515 not really sufficiently general. Furthermore, there was one instance where
4516 it was actually wrong because one of the argument was used twice, and in
4517 one call a function was used. (As it happened, calling the function twice
4518 did not affect the overall behaviour.) The macro has been replaced by a
4519 function that can join an arbitrary number of extra strings onto a growing
4522 (c) The code for expansion conditions now uses a table and a binary chop
4523 instead of a serial search (which was left over from when there were very
4524 few conditions). Also, it now recognizes conditions like "pam" even when
4525 the relevant support is not compiled in: a suitably worded error message is
4526 given if an attempt is made to use such a condition.
4528 11. Added ${time_interval:xxxxx}.
4530 12. A bug was causing one of the ddress fields not to be passed back correctly
4531 from remote delivery subprocesses. The field in question was not being
4532 subsequently used, so this caused to problems in practice.
4534 13. Added new log selectors queue_time and deliver_time.
4536 14. Might have fixed a bug in maildirsizefile handling that threw up
4537 "unexpected character" debug warnings, and recalculated the data
4538 unnecessarily. In any case, I expanded the warning message to give more
4541 15. Added the message "Restricted characters in address" to the statements in
4542 the default ACL that block characters like @ and % in local parts.
4544 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4545 Three changes have been made:
4547 (a) There was a serious bug; a negative response to MAIL caused the whole
4548 recipient domain to be cached as invalid, thereby blocking all messages
4549 to all local parts at the same domain, from all senders. This bug has
4550 been fixed. The domain is no longer cached after a negative response to
4551 MAIL if the sender used is not empty.
4553 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4556 (c) A new callout option, "use_sender" has been added for people who want
4557 the modified behaviour.
4563 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4566 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4567 indeed breaks things for older releases.
4569 3. Added additional logging to the case where there is a problem reading data
4570 from a filter that is running in a subprocess using a pipe, in order to
4571 try to track down a specific problem.
4573 4. Testing facility fudge: when running in the test harness and attempting
4574 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4575 getting "No route to host". Convert this to a timeout.
4577 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4580 6. Some OS don't have socklen_t but use size_t instead. This affects the
4581 fifth argument of getsockopt() amongst other things. This is now
4582 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4583 can be set for individual OS. I have set it for SunOS5, OSF1, and
4584 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4585 some earlier ones do not.
4587 7. Change 4.30/15 was not doing the test caselessly.
4589 8. The standard form for an IPv6 address literal was being rejected by address
4590 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4591 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4592 this, as well as the form without the "IPv6" on the front (but only when
4593 address literals are enabled, of course).
4595 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4597 10. Exim crashed if a message with an empty sender address specified by -f
4598 encountered a router with an errors_to setting. This could be provoked only
4599 by a command such as
4603 where an empty string was supplied; "<>" did not hit this bug.
4605 11. Installed PCRE release 4.5.
4607 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4608 remained set. It is now erased.
4610 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4611 times from message ids (which are base 36 rather than the normal 62).
4613 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4614 were being counted as actual protocol errors, and logged if the log
4615 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4616 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4617 and DATA following a set of rejected RCPTs do not count as protocol errors.
4618 In other words, Exim assumes they were pipelined, though this may not
4619 actually be the case. Of course, in all cases the client gets an
4620 appropriate error code.
4622 15. If a lookup fails in an ACL condition, a message about the failure may
4623 be available; it is used if testing the ACL cannot continue, because most
4624 such messages specify what the cause of the deferral is. However, some
4625 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4626 that caused an old message to be retained and used if a later statement
4627 caused a defer, replacing the real cause of the deferral.
4629 16. If an IP address had so many PTR records that the DNS lookup buffer
4630 was not large enough to hold them, Exim could crash while trying to process
4631 the truncated data. It now detects and logs this case.
4633 17. Further to 4.21/58, another change has been made: if (and only if) the
4634 first line of a message (the first header line) ends with CRLF, a bare LF
4635 in a subsequent header line has a space inserted after it, so as not to
4636 terminate the header.
4638 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4639 unnecessarily, used atoi() instead of strtol(), and didn't check the
4640 termination when getting file sizes from file names by regex.
4642 19. Completely re-implemented the support for maildirsize files, in the light
4643 of a number of problems with the previous contributed implementation
4644 (4.30/29). In particular:
4646 . If the quota is zero, the maildirsize file is maintained, but no quota is
4649 . If the maildir directory does not exist, it is created before any attempt
4650 to write a maildirsize file.
4652 . The quota value in the file is just a cache; if the quota is changed in
4653 the transport, the new value overrides.
4655 . A regular expression is available for excluding directories from the
4658 20. The autoreply transport checks the characters in options that define the
4659 message's headers; it allows continued headers, but it was checking with
4660 isspace() after an embedded newline instead of explicitly looking for a
4663 21. If all the "regular" hosts to which an address was routed had passed their
4664 expiry times, and had not reached their retry times, the address was
4665 bounced, even if fallback hosts were defined. Now Exim should go on to try
4668 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4669 equivalent code in the SMTP transport. Some hosts send humungous responses
4670 to HELO/EHLO, more than 1024 it seems.
4672 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4673 gives compiler warnings in some environments. I've now done it "properly",
4676 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4677 (because of problems with the built-in one) was declared to return uschar *
4678 instead of char *, causing compiler failure.
4680 25. Fixed a file descriptor leak when processing alias/forward files.
4682 26. Fixed a minor format string issue in dbfn.c.
4684 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4686 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4687 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4690 29. When a sender address is verified, it is cached, to save repeating the test
4691 when there is more than one recipient in a message. However, when the
4692 verification involves a callout, it is possible for different callout
4693 options to be set for different recipients. It is too complicated to keep
4694 track of this in the cache, so now Exim always runs a verification when a
4695 callout is required, relying on the callout cache for the optimization.
4696 The overhead is duplication of the address routing, but this should not be
4699 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4700 to be verified with callout=postmaster, and the main callout worked but the
4701 postmaster check failed, the verification correctly failed. However, if a
4702 subsequent RCPT command asked for sender verification *without* the
4703 postmaster check, incorrect caching caused this verification also to fail,
4706 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4707 it was not caching the DNS options (qualify_single, search_parents) that
4708 were used when the lookup failed. A subsequent lookup with different
4709 options therefore always gave the same answer, though there were cases
4710 where it should not have. (Example: a "domains = !$mx_any" option on a
4711 dnslookup router: the "domains" option is always processed without any
4712 widening, but the router might have qualify_single set.) Now Exim uses the
4713 cached value only when the same options are set.
4715 32. Added John Jetmore's "exipick" utility to the distribution.
4717 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4718 than a timeout (e.g. a certificate is required, and is not provided), an
4719 Exim server now closes the connection immediately. Previously it waited for
4720 the client to close - but if the client is SSL, it seems that they each
4721 wait for each other, leading to a delay before one of them times out.
4723 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4724 maintained 0.8.x compatibility because I don't think many are using it, and
4725 it is clearly obsolete.
4727 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4730 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4731 client certificate was expired. A simple patch fixes this, though I don't
4732 understand the full logic of why the verify callback is called multiple
4735 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4736 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4737 which causes problems with some clients (such as the Certicom SSL Plus
4738 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4739 disables the coutermeasure allowing Eudora to connect."
4741 38. Exim was not checking that a write() to a log file succeeded. This could
4742 lead to Bad Things if a log got too big, in particular if it hit a file
4743 size limit. Exim now panics and dies if it cannot write to a log file, just
4744 as it does if it cannot open a log file.
4746 39. Modified OS/Makefile-Linux so that it now contains
4748 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4750 The two -D definitions ensure that Exim is compiled with large file
4751 support, which makes it possible to handle log files that are bigger than
4754 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4755 instance) a domain was checked against a named list that involved a lookup,
4756 causing $domain_data to be set, then another domain was checked against the
4757 same list, then the first domain was re-checked, the value of $domain_data
4758 after the final check could be wrong. In particular, if the second check
4759 failed, it could be set empty. This bug probably also applied to
4762 41. The strip_trailing_dot option was not being applied to the address given
4763 with the -f command-line option.
4765 42. The code for reading a message's header from the spool was incrementing
4766 $received_count, but never initializing it. This meant that the value was
4767 incorrect (doubled) while delivering a message in the same process in which
4768 it was received. In the most common configuration of Exim, this never
4769 happens - a fresh exec is done - but it can happen when
4770 deliver_drop_privilege is set.
4772 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4773 it now includes up to 150 characters of the unexpected data in the log
4776 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4777 and building data strings. The size of both of these buffers was 10 000
4778 bytes - far larger than anybody would *ever* want, thought I. Needless to
4779 say, somebody hit the limit. I have increased the maximum line length to
4780 20 000 and the maximum data length of concatenated lines to 100 000. I have
4781 also fixed two bugs, because there was no checking on these buffers. Tsk,
4782 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4783 buffer is too small.
4785 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4786 lsearch lookups. Now it does.
4788 46. When parsing a route_list item in a manualroute router, a fixed-length
4789 buffer was used for the list of hosts. I made this 1024 bytes long,
4790 thinking that nobody would ever have a list of hosts that long. Wrong.
4791 Somebody had a whole pile of complicated expansion conditions, and the
4792 string was silently truncated, leading to an expansion error. It turns out
4793 that it is easier to change to an unlimited length (owing to other changes
4794 that have happened since this code was originally written) than to build
4795 structure for giving a limitation error. The length of the item that
4796 expands into the list of hosts is now unlimited.
4798 47. The lsearch lookup could not handle data where the length of text line was
4799 more than 4095 characters. Such lines were truncated, leading to shortened
4800 data being returned. It should now handle lines of any length.
4802 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4803 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4806 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4807 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4808 no longer happens in comment lines. A list of replacements is now placed
4809 at the head of all of the source files, except those whose only change is
4810 to replace PERL_COMMAND in the very first #! line.
4812 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4813 messages on the queue, with a bottom-up merge sort, using code contributed
4814 by Michael Haardt. This should make operations like -bp somewhat faster on
4815 large queues. It won't affect queue runners, except when queue_run_in_order
4818 51. Installed eximstats 1.31 in the distribution.
4820 52. Added support for SRV lookups to the dnslookup router.
4822 53. If an ACL referred to $message_body or $message_body_end, the value was not
4823 reset for any messages that followed in the same SMTP session.
4825 54. The store-handling optimization for building very long strings was not
4826 differentiating between the different store pools. I don't think this
4827 actually made any difference in practice, but I've tidied it.
4829 55. While running the routers to verify a sender address, $sender_address
4830 was still set to the sender address. This is wrong, because when routing to
4831 send a bounce to the sender, it would be empty. Therefore, I have changed
4832 it so that, while verifying a sender address, $sender_address is set to <>.
4833 (There is no change to what happens when verifying a recipient address.)
4835 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4836 target A or AAAA records (if not already returned) without resetting the
4837 qualify_single or search_parents options of the DNS resolver. These are
4838 inappropriate in this case because the targets of MX and SRV records must
4839 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4840 target that, when qualified, matched something in the local domain. These
4841 two options are now turned off when doing these lookups.
4843 57. It seems that at least some releases of Reiserfs (which does not have the
4844 concept of a fixed number of inodes) returns zero and not -1 for the
4845 number of available inodes. This interacted badly with check_spool_inodes,
4846 which assumed that -1 was the "no such thing" setting. What I have done is
4847 to check that the total number of inodes is greater than zero before doing
4848 the test of how many are available.
4850 58. When a "warn" ACL statement has a log_message modifier, the message is
4851 remembered, and not repeated. This is to avoid a lot of repetition when a
4852 message has many recipients that cause the same warning to be written.
4853 Howewer, Exim was preserving the list of already written lines for an
4854 entire SMTP session, which doesn't seem right. The memory is now reset if a
4855 new message is started.
4857 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4858 debugging output unless log_rewrite was also set.
4860 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4861 of (int)(handle) when we know that handle contains (void *)(-1).
4863 61. The Exim daemon panic-logs an error return when it closes the incoming
4864 connection. However "connection reset by peer" seems to be common, and
4865 isn't really an error worthy of noting specially, so that particular error
4868 62. When Exim is trying to find all the local interfaces, it used to panic and
4869 die if the ioctl to get the interface flags failed. However, it seems that
4870 on at least one OS (Solaris 9) it is possible to have an interface that is
4871 included in the list of interfaces, but for which you get a failure error
4872 for this call. This happens when the interface is not "plumbed" into a
4873 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4874 failure of the "get flags" call assumes that the interface is down.
4876 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4877 makes life easier for people who are doing arithmetic on fields extracted
4878 from dates, where you often get leading zeros that should not be
4879 interpreted as octal.
4881 64. Added qualify_domain to the redirect router, to override the global
4884 65. If a pathologically long header line contained very many addresses (the
4885 report of this problem mentioned 10 000) and each of them was rewritten,
4886 Exim could use up a very large amount of memory. (It kept on making new
4887 copies of the header line as it rewrote, and never released the old ones.)
4888 At the expense of a bit more processing, the header rewriting function has
4889 been changed so that it no longer eats memory in this way.
4891 66. The generation of the Received: header has been moved from the time that a
4892 message starts to be received, to the time that it finishes. The timestamp
4893 in the Received: header should now be very close to that of the <= log
4894 line. There are two side-effects of this change:
4896 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4897 logged header lines no longer include the local Received: line, because
4898 it has not yet been created. The same applies to a copy of the message
4899 that is returned to a non-SMTP sender when a message is rejected.
4901 (b) When a filter file is tested using -bf, no additional Received: header
4902 is added to the test message. After some thought, I decided that this
4905 This change does not affect the value of $received_for. It is still set
4906 after address rewriting, but before local_scan() is called.
4908 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4910 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4911 gave an unhelpful panic error message, and a defer error. I have managed to
4912 change this behaviour so that it now rejects any supplied certificate,
4913 which seems right, as the list of acceptable certificates is empty.
4915 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4916 gave an unhelpful defer error. I have not managed to make this reject any
4917 supplied certificates, but the error message it gives is "no certificate
4918 supplied", which is not helpful.
4920 70. exigrep's output now also includes lines that are not associated with any
4921 message, but which match the given pattern. Implemented by a patch from
4922 Martin Sluka, which also tidied up the Perl a bit.
4924 71. Recipient callout verification, like sender verification, was using <> in
4925 the MAIL FROM command. This isn't really the right thing, since the actual
4926 sender may affect whether the remote host accepts the recipient or not. I
4927 have changed it to use the actual sender in the callout; this means that
4928 the cache record is now keyed on a recipient/sender pair, not just the
4929 recipient address. There doesn't seem to be a real danger of callout loops,
4930 since a callout by the remote host to check the sender would use <>.
4931 [SEE ABOVE: changed after hitting problems.]
4933 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4934 temporary errors. However, in the case of such a code being given after
4935 the end of a data transmission (i.e. after ".") Exim was failing to write
4936 a retry record for the message. (Yes, there was some broken host that was
4937 actually sending 8xx at this point.)
4939 73. An unknown lookup type in a host list could cause Exim to panic-die when
4940 the list was checked. (An example that provoked this was putting <; in the
4941 middle of a list instead of at the start.) If this happened during a DATA
4942 ACL check, a -D file could be left lying around. This kind of configuration
4943 error no longer causes Exim to die; instead it causes a defer errror. The
4944 incident is still logged to the main and panic logs.
4946 74. Buglet left over from Exim 3 conversion. The message "too many messages
4947 in one connection" was written to the rejectlog but not the mainlog, except
4948 when address rewriting (yes!) was being logged.
4950 75. Added write_rejectlog option.
4952 76. When a system filter was run not as root (that is, when system_filter_user
4953 was set), the values of the $n variables were not being returned to the
4954 main process; thus, they were not subsequently available in the $sn
4957 77. Added +return_path_on_delivery log selector.
4959 78. A connection timeout was being treated differently from recipients deferred
4960 when testing hosts_max_try with a message that was older than the host's
4961 retry timeout. (The host should not be counted, thus allowing all hosts to
4962 be tried at least once before bouncing.) This may have been the cause of an
4963 occasionally reported bug whereby a message would remain on the queue
4964 longer than the retry timeout, but would be bounced if a delivery was
4965 forced. I say "may" because I never totally pinned down the problem;
4966 setting up timeout/retry tests is difficult. See also the next item.
4968 79. The ultimate address timeout was not being applied to errors that involved
4969 a combination of host plus message (for example, a timeout on a MAIL
4970 command). When an address resolved to a number of possible hosts, and they
4971 were not all tried for each delivery (e.g. because of hosts_max_try), a
4972 message could remain on the queue longer than the retry timeout.
4974 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4977 81. Fixed an obscure SMTP outgoing bug which required at least the following
4978 conditions: (a) there was another message waiting for the same server;
4979 (b) the server returned 5xx to all RCPT commands in the first message so
4980 that the message was not completed; (c) the server dropped the connection
4981 or gave a negative response to the RSET that Exim sends to abort the
4982 transaction. The observed case was a dropped connection after DATA that had
4983 been sent in pipelining mode. That is, the server had advertised PIPELINING
4984 but was not implementing it correctly. The effect of the bug was incorrect
4985 behaviour, such as trying another host, and this could lead to a crash.
4991 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4992 and daemon.c were passed as pointers to ints; they should have been
4993 pointers to socklen_t variables (which are typically unsigned ints).
4995 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4998 3. Fixed a really odd bug that affected only the testing scheme; patching a
4999 certain fixed string in the binary changed the value of another string that
5000 happened to be identical to the end of the original first string.
5002 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5003 name", it returns that address as the IP address. On some operating
5004 systems (e.g. Solaris), it also passes back the IP address string as the
5005 "host name". However, on others (e.g. Linux), it passes back an empty
5006 string. Exim wasn't checking for this, and was changing the host name to an
5007 empty string, assuming it had been canonicized.
5009 5. Although rare, it is permitted to have more than one PTR record for a given
5010 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5011 all the names associated with an address, because they do in Solaris.
5012 However, it seems that they do not in Linux for data that comes from the
5013 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5014 I found this out when I moved to a new Linux workstation and tried to run
5015 the Exim test suite.
5017 To get round this problem I have changed the code so that it now does its
5018 own call to the DNS to look up PTR records when searching for a host name.
5019 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5020 addresses that are only in /etc/hosts are still found.
5022 This behaviour is, however, controlled by an option called host_lookup_
5023 order, which defaults to "bydns:byaddr". If people want to use the other
5024 order, or indeed, just use one or the other means of lookup, they can
5025 specify it in this variable.
5027 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5028 some operating systems, this comes back from gethostbyaddr() as an empty
5029 string, and this is what Exim used to test for. However, it seems that in
5030 other systems, "." is yielded. Exim now tests for this case too.
5032 7. The values of check_spool_space and check_log_space are now held internally
5033 as a number of kilobytes instead of an absolute number of bytes. If a
5034 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5035 kilobyte. This means that much larger values can be stored.
5037 8. Exim monitor: an attempt to get the action menu when not actually pointing
5038 at a message produces an empty menu entitled "No message selected". This
5039 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5040 no entries in it ("Shell widget menu has zero width and/or height"). So I
5041 have added a single, blank menu entry in this case.
5043 9. Added ${quote_local_part.
5045 10. MIME decoding is now applied to the contents of Subject: header lines when
5048 11. Now that a reference to $sender_host_address automatically causes a reverse
5049 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5050 host lookup before query-style lookups in lists that might use this
5051 variable. This has therefore been abolished, and the "net-" prefix is no
5052 longer necessary for query-style lookups.
5054 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5055 have been a typo for LFLAGS, so it has been changed.
5057 13. The install script calls Exim with "-C /dev/null" in order to find the
5058 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5059 to be output. Howeve, since Exim outputs its version number before the
5060 error, it didn't break the script. It just looked ugly. I fixed this by
5061 always allowing "-C /dev/null" if the caller is root.
5063 14. Ignore overlarge ACL variable number when reading spool file - insurance
5064 against a later release with more variables having written the file.
5066 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5067 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5068 this, as well as the form without the "IPv6" on the front.
5070 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5071 OS/Makefile-Darwin file.
5073 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5074 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5076 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5077 variable. However, some OS use ints for inodes. I've added cast to long int
5078 to get rid of the compiler warning.
5080 19. I had forgotten to lock out "/../" in configuration file names when
5081 ALT_CONFIG_PREFIX was set.
5083 20. Routers used for verification do not need to specify transports. However,
5084 if such a router generated a host list, and callout was configured, Exim
5085 crashed, because it could not find a port number from the (non-existent)
5086 transport. It now assumes port 25 in this circumstance.
5088 21. Added the -t option to exigrep.
5090 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5091 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5092 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5093 them is set, LOOKUP_LSEARCH is forced.
5095 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5097 24. Added sender and host information to the "rejected by local_scan()" log
5098 line; previously there was no indication of these.
5100 25. Added .include_if_exists.
5102 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5103 Linux. It turns out that not all file systems support this. Apparently some
5104 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5105 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5106 not supported on the file descriptor, is now ignored when Exim is trying to
5107 sync a directory. This applies only to Linux.
5109 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5111 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5112 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5115 29. Added support for maildirsize files from supplied patch (modified a bit).
5117 30. The use of :fail: followed by an empty string could lead Exim to respond to
5118 sender verification failures with (e.g.):
5120 550 Verification failed for <xxx>
5121 550 Sender verify failed
5123 where the first response line was missing the '-' that indicates it is not
5124 the final line of the response.
5126 31. The loop for finding the name of the user that called Exim had a hardwired
5127 limit of 10; it now uses the value of finduser_retries, which is used for
5128 all other user lookups.
5130 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5133 33. Exim was neglecting to zero errno before one call of strtol() when
5134 expanding a string and expecting an integer value. On some systems this
5135 resulted in spurious "integer overflow" errors. Also, it was casting the
5136 result into an int without checking.
5138 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5139 did not work. The code looks as if it has *never* worked, though it appears
5140 to have been documented since at least releast 1.62. I have made it work.
5142 35. The "timeout_DNS" error in retry rules, also documented since at least
5143 1.62, also never worked. As it isn't clear exactly what this means, and
5144 clearly it isn't a major issue, I have abolished the feature by treating it
5145 as "timeout", and writing a warning to the main and panic logs.
5147 36. The display of retry rules for -brt wasn't always showing the error code
5150 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5151 timeout_connect_A, timeout_connect_MX.
5153 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5154 to the empty sender.
5156 39. The daemon was not analysing the content of -oX till after it had closed
5157 stderr and disconnected from the controlling terminal. This meant that any
5158 syntax errors were only noted on the panic log, and the return code from
5159 the command was 0. By re-arranging the code a little, I've made the
5160 decoding happen first, so such errors now appear on stderr, and the return
5161 code is 1. However, the actual setting up of the sockets still happens in
5162 the disconnected process, so errors there are still only recorded on the
5165 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5166 connections (as happens on some IP stacks) was logged at start up time as
5167 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5168 it from "IPv6 and IPv4", which means that two separate sockets are being
5171 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5172 says whether AF_INET or AF_INET6 was passed as an argument.
5174 42. Exiwhat output was messed up when time zones were included in log
5177 43. Exiwhat now gives more information about the daemon's listening ports,
5178 and whether -tls-on-connect was used.
5180 44. The "port" option of the smtp transport is now expanded.
5182 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5183 silently ignored. Now an error message is written to the main and panic
5186 46. There's a new ACL modifier called "logwrite" which writes to a log file
5187 as soon as it is encountered.
5189 47. Added $local_user_uid and $local_user_gid at routing time.
5191 48. Exim crashed when trying to verify a sender address that was being
5194 49. Exim was recognizing only a space character after ".include". It now also
5195 recognizes a tab character.
5197 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5198 extracting the relevant information from the specification. The man page no
5199 longer contains scrambled data for the -d option, and I've added a section
5200 at the front about calling Exim under different names.
5202 51. Added "extra_headers" argument to the "mail" command in filter files.
5204 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5207 53. Installed eximstats 1.29.
5209 54. Added transport_filter_timeout as a generic transport option.
5211 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5212 Cc: header lines. This was required by RFC 822, but it not required by RFC
5215 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5216 incoming messages that did not have them. Now it does so only if the
5217 message originates locally, that is, if there is no associated remote host
5218 address. When Resent- header lines are present, this applies to the Resent-
5219 lines rather than the non-Resent- lines.
5221 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5222 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5224 58. Messages for configuration errors now include the name of the main
5225 configuration file - useful now that there may be more than one file in a
5226 list (.included file names were always shown).
5228 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5229 for those rare installations that do not start the daemon as root or run it
5230 setuid root. I've cut out the call to initgroups() if the daemon is not
5233 60. The Exim user and group can now be bound into the binary as text strings
5234 that are looked up at the start of Exim's processing.
5236 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5238 62. Added $mailstore_basename variable.
5240 63. Installed patch to sieve.c from Michael Haardt.
5242 64. When Exim failed to open the panic log after failing to open the main log,
5243 the original message it was trying to log was written to stderr and debug
5244 output, but if they were not available (the usual case in production), it
5245 was lost. Now it is written to syslog before the two lines that record the
5246 failures to open the logs.
5248 65. Users' Exim filters run in subprocesses under the user's uid. It is
5249 possible for a "deliver" command or an alias in a "personal" command to
5250 provoke an address rewrite. If logging of address rewriting is configured,
5251 this fails because the process is not running as root or exim. There may be
5252 a better way of dealing with this, but for the moment (because 4.30 needs
5253 to be released), I have disabled address rewrite logging when running a
5254 filter in a non-root, non-exim process.
5260 1. The buildconfig auxiliary program wasn't quoting the value set for
5261 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5262 not defined. This bug was masked in 4.22 by the effect that was fixed in
5265 2. Some messages that were rejected after a message id was allocated were
5266 shown as "incomplete" by exigrep. It no longer does this for messages that
5267 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5269 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5270 have allow_domain_literals set, the ID did not get logged in the <= line.
5271 Domain literals are now always recognized in Message-ID: header lines.
5273 4. The first argument for a ${extract expansion item is the key name or field
5274 number. Leading and trailing spaces in this item were not being ignored,
5275 causing some misleading effects.
5277 5. When deliver_drop_privilege was set, single queue runner processes started
5278 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5279 same command in the process it spins off) were not dropping privilege.
5281 6. When the daemon running as "exim" started a queue runner, it always
5282 re-executed Exim in the spun-off process. This is a waste of effort when
5283 deliver_drop_privilege is set. The new process now just calls the
5284 queue-runner function directly.
5290 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5293 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5294 ignored. Though the use of -oP was forcing the writing of a pid file, it
5295 was always written to the default place.
5297 3. If the message "no IP address found for host xxxx" is generated during
5298 incoming verification, it is now followed by identification of the incoming
5299 connection (so you can more easily find what provoked it).
5301 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5303 5. Added some features to "harden" Exim a bit more against certain attacks:
5305 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5306 be put in Local/Makefile. This is like the never_users runtime option,
5307 but it cannot be overridden. The default setting is "root".
5309 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5310 prefix string with which any file named in a -C command line option
5313 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5314 is retained for -C and -D only if the caller of Exim is root. Without
5315 it, the exim user may also use -C and -D and retain privilege.
5317 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5318 command line option is disabled.
5320 6. Macro names set by the -D option must start with an upper case letter, just
5321 like macro names defined in the configuration file.
5323 7. Added "dereference=" facility to LDAP.
5325 8. Two instances of the typo "uknown" in the source files are fixed.
5327 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5328 the Configure-Makefile script screwed up while processing it.
5330 10. Incorporated PCRE 4.4.
5332 11. The SMTP synchronization check was not operating right at the start of an
5333 SMTP session. For example, it could not catch a HELO sent before the client
5334 waited for the greeting. There is now a check for outstanding input at the
5335 point when the greeting is written. Because of the duplex, asynchronous
5336 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5337 way, but not yet received, when the check is performed.
5339 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5340 on TCP/IP sockets, because this apparently causes some broken clients to
5343 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5344 unchanged) from the Cygwin maintainer.
5346 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5347 was supported instead of testing for mbx. Effectively a typo.
5349 15. The spa authenticator server code was not checking that the input it
5350 received was valid base64.
5352 16. The debug output line for the "set" modifier in ACLs was not showing the
5353 name of the variable that was being set.
5355 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5357 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5358 lookup on its own. Something else was needed to trigger the lookup. For
5359 example, a match in host_lookup or the need for a host name in a host list.
5360 Now, if $sender_host_name is referenced and the host name has not yet been
5361 looked up, a lookup is performed. If the lookup fails, the variable remains
5362 empty, and $host_lookup_failed is set to "1".
5364 19. Added "eqi" as a case-independent comparison operator.
5366 20. The saslauthd authentication condition could segfault if neither service
5367 nor realm was specified.
5369 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5370 error message that was logged was misleading, and incoming SMTP
5371 connections were dropped. The message is now more accurate, and temporary
5372 errors are given to SMTP connections.
5374 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5375 (except RSET) with a 421 error, until QUIT is received. However, it was
5376 failing to send a response to QUIT.
5378 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5379 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5380 ACL is now run after the helo_try_verify_hosts code.
5382 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5383 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5384 case-independent, so other case variants are also recognized). Apparently
5385 some systems use these upper case variants.
5387 25. If more than two messages were waiting for the same host, and a transport
5388 filter was specified for the transport, Exim sent two messages over the
5389 same TCP/IP connection, and then failed with "socket operation on non-
5390 socket" when it tried to send the third.
5392 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5394 27. The extern definition of crypt16() in expand.c was not being excluded when
5395 the OS had its own crypt16() function.
5397 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5398 as a preferred synonym for return_size_limit, both as an option and as an
5401 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5403 30. Changed the default configuration ACL to relax the local part checking rule
5404 for addresses that are not in any local domains. For these addresses,
5405 slashes and pipe symbols are allowed within local parts, but the sequence
5406 /../ is explicitly forbidden.
5408 31. SPA server authentication was not clearing the challenge buffer before
5411 32. log_message in a "warn" ACL statement was writing to the reject log as
5412 well as to the main log, which contradicts the documentation and doesn't
5413 seem right (because no rejection is happening). So I have stopped it.
5415 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5416 However, I am unable to do any testing of this.
5418 34. Fixed an infelicity in the appendfile transport. When checking directories
5419 for a mailbox, to see if any needed to be created, it was accidentally
5420 using path names with one or more superfluous leading slashes; tracing
5421 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5423 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5424 contents are added to the log line that is written for every discarded
5425 recipient. (Previously a log_message setting was ignored.)
5427 36. The ${quote: operator now quotes the string if it is empty.
5429 37. The install script runs exim in order to find its version number. If for
5430 some reason other than non-existence or emptiness, which it checks, it
5431 could not run './exim', it was installing it with an empty version number,
5432 i.e. as "exim-". This error state is now caught, and the installation is
5435 38. An argument was missing from the function that creates an error message
5436 when Exim fails to connect to the socket for saslauthd authentication.
5437 This could cause Exim to crash, or give a corrupted message.
5439 39. Added isip, isip4, and isip6 to ${if conditions.
5441 40. The ACL variables $acl_xx are now saved with the message, and can be
5442 accessed later in routers, transports, and filters.
5444 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5445 strings in the file are not string-expanded.
5447 42. If a MAIL command specified a SIZE value that was too large to fit into an
5448 int variable, the check against message_size_limit failed. Such values are
5449 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5450 day this will have to be increased, but I don't think I want to be around
5451 when emails are that large.
5458 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5459 iconv() is not standard in FreeBSD.
5461 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5462 IPv6 enabled. The observed symptom was a segmentation fault on return from
5463 the function os_common_find_running_interfaces() in src/os.c.
5465 3. In the check_special_case() function in daemon.c I had used "errno" as an
5466 argument name, which causes warnings on some systems. This was basically a
5467 typo, since it was named "eno" in the comments!
5469 4. The code that waits for the clock to tick (at a resolution of some fraction
5470 of a second) so as to ensure message-id uniqueness was always waiting for
5471 at least one whole tick, when it could have waited for less. [This is
5472 almost certainly not relevant at current processor speeds, where it is
5473 unlikely to ever wait at all. But we try to future-proof.]
5475 5. The function that sleeps for a time interval that includes fractions of a
5476 second contained a race. It did not block SIGALRM between setting the
5477 timer, and suspending (a couple of lines later). If the interval was short
5478 and the sigsuspend() was delayed until after it had expired, the suspension
5479 never ended. On busy systems this could lead to processes getting stuck for
5482 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5483 process, before it forks any delivery processes. The open lookup caching
5484 mechanism meant that the open file or database connection was passed into
5485 the delivery process. The problem was that delivery processes always tidy
5486 up cached lookup data. This could cause a problem for the next delivery
5487 process started by the queue runner, because the external queue runner
5488 process does not know about the closure. So the next delivery process
5489 still has data in the lookup cache. In the case of a file lookup, there was
5490 no problem because closing a file descriptor in a subprocess doesn't affect
5491 the parent. However, if the lookup was caching a connection to a database,
5492 the connection was closed, and the second delivery process was likely to
5493 see errors such as "PGSQL: query failed: server closed the connection
5494 unexpectedly". The problem has been fixed by closing all cached lookups
5495 in a queue runner before running a delivery process.
5497 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5498 seem to have the "const" qualifier which it has on other OS. I've
5501 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5502 *of the same type* (client or server) with the same public name that an
5503 error should be diagnosed.
5505 9. When Exim looked up a host name for an IP address, but failed to find the
5506 original IP address when looking up the host name (a safety check), it
5507 output the message "<ip address> does not match any IP for NULL", which was
5508 confusing, to say the least. The bug was that the host name should have
5509 appeared instead of "NULL".
5511 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5512 user that is built into the binary, and the -C or -D options is used, root
5513 privilege is dropped before the configuration file is read. In addition,
5514 logging is switched to stderr instead of the normal log files. If the
5515 configuration then re-defines the Exim user, the unprivileged environment
5516 is probably not what is expected, so Exim logs a panic warning message (but
5519 However, if deliver_drop_privilege is set, the unprivileged state may well
5520 be exactly what is intended, so the warning has been cut out in that case,
5521 and Exim is allowed to try to write to its normal log files.
5527 1. smtp_return_error_details was not giving details for temporary sender
5528 or receiver verification errors.
5530 2. Diagnose a configuration error if two authenticators have the same public
5533 3. Exim used not to create the message log file for a message until the first
5534 delivery attempt. This could be confusing when incoming messages were held
5535 for policy or load reasons. The message log file is now created at the time
5536 the message is received, and an initial "Received" line is written to it.
5538 4. The automatically generated man page for command line options had a minor
5539 bug that caused no ill effects; however, a more serious problem was that
5540 the procedure for building the man page automatically didn't always
5541 operate. Consequently, release 4.20 contains an out-of-date version. This
5542 shouldn't happen again.
5544 5. When building Exim with embedded Perl support, the script that builds the
5545 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5546 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5548 6. The freeze_tell option was not being used for messages that were frozen on
5549 arrival, either by an ACL or by local_scan().
5551 7. Added the smtp_incomplete_transaction log selector.
5553 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5554 was accepting AUTH without a new EHLO.
5556 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5557 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5560 10. Logging of TCP/IP connections (when configured) now happens in the main
5561 daemon process instead of the child process, so that the TCP/IP connection
5562 count is more accurate (but it can never be perfect).
5564 11. The use of "drop" in a nested ACL was not being handled correctly in the
5565 outer ACL. Now, if condition failure induced by the nested "drop" causes
5566 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5567 or "require"), the connection is dropped.
5569 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5570 that yield "discard" can now be used with an "accept" or a "discard" verb,
5571 but an error is generated for any others (because I can't see a useful way
5572 to define what should happen).
5574 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5575 are now processed in the same way as lines in the Exim configuration file.
5576 In particular, continuation lines are supported.
5578 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5580 15. Added -ti meaning -t -i.
5582 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5583 domains, and warn by logging if others are found.
5585 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5586 returned from GIFCONF when Exim is trying to find the list of interfaces on
5587 a host. The code in os.c has been modified to copy each ifreq to an aligned
5588 structure in all cases.
5590 Also, in some cases, the returned ifreq's were being copied to a 'struct
5591 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5592 means the last couple of bytes of an IPv6 address could be chopped if the
5593 ifreq contained only a normal sockaddr (14 bytes storage).
5595 18. Named domain lists were not supported in the hosts_treat_as_local option.
5596 An entry such as +xxxx was not recognized, and was treated as a literal
5599 19. Ensure that header lines added by a DATA ACL are included in the reject log
5600 if the ACL subsequently rejects the message.
5602 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5603 MD5 (which is deprecated).
5605 21. When testing a filter file using -bf, Exim was writing a message when it
5606 took the sender from a "From " line in the message, but it was not doing so
5607 when it took $return_path from a Return-Path: header line. It now does.
5609 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5610 with a valid header line field name (a series of printing characters
5611 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5613 23. Changed "disc" in the source to "disk" to conform to the documentation and
5614 the book and for uniformity.
5616 24. Ignore Sendmail's -Ooption=value command line item.
5618 25. When execve() failed while trying to run a command in a pipe transport,
5619 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5620 could be confused with a return value of 69 from the command itself. This
5621 has been changed to 127, the value the shell returns if it is asked to run
5622 a non-existent command. The wording for the related log line suggests a
5623 non-existent command as the problem.
5625 26. If received_header_text expands to an empty string, do not add a Received:
5626 header line to the message. (Well, it adds a token one on the spool, but
5627 marks it "old" so that it doesn't get used or transmitted.)
5629 27. Installed eximstats 1.28 (addition of -nt option).
5631 28. There was no check for failure on the call to getsockname() in the daemon
5632 code. This can fail if there is a shortage of resources on the system, with
5633 ENOMEM, for example. A temporary error is now given on failure.
5635 29. Contrary to the C standard, it seems that in some environments, the
5636 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5637 program. Exim now does this explicitly; it affects the formatting of
5638 timestamps using strftime().
5640 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5641 complaints. I've now initialized all the variables, to avoid this.
5643 32. Header lines added by a system filter were not being "seen" during
5644 transport-time rewrites.
5646 33. The info_callback() function passed to OpenSSL is set up with type void
5647 (*)(SSL *, int, int), as described somewhere. However, when calling the
5648 function (actually a macro) that sets it up, the type void(*)() is
5649 expected. I've put in a cast to prevent warnings from picky compilers.
5651 34. If a DNS black list lookup found a CNAME record, but there were no A
5652 records associated with the domain it pointed at, Exim crashed.
5654 35. If a DNS black list lookup returned more than one A record, Exim ignored
5655 all but the first. It now scans all returned addresses if a particular IP
5656 value is being sought. In this situation, the contents of the
5657 $dnslist_value variable are a list of all the addresses, separated by a
5660 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5661 to accept a host name and all its aliases if the forward lookup for any of
5662 them yielded the IP address of the incoming connection. Now it accepts only
5663 those names whose forward lookup yields the correct IP address. Any other
5664 names are discarded. This closes a loophole whereby a rogue DNS
5665 administrator could create reverse DNS records to break through a
5666 wildcarded host restriction in an ACL.
5668 37. If a user filter or a system filter that ran in a subprocess used any of
5669 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5670 the wrong values were passed to the pipe command ($thisaddress had the
5671 value of $0, $0 had the value of $1, etc). This bug was introduced by
5672 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5674 38. Improved the line breaking for long SMTP error messages from ACLs.
5675 Previously, if there was no break point between 40 and 75 characters, Exim
5676 left the rest of the message alone. Two changes have been made: (a) I've
5677 reduced the minimum length to 35 characters; (b) if it can't find a break
5678 point between 35 and 75 characters, it looks ahead and uses the first one
5679 that it finds. This may give the occasional overlong line, but at least the
5680 remaining text gets split now.
5682 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5683 file descriptors might be low, and that setting 1000 would always raise it.
5684 It turns out that in some environments, the limit is already over 1000 and
5685 that lowering it causes trouble. So now Exim takes care not to decrease it.
5687 40. When delivering a message, the value of $return_path is set to $sender_
5688 address at the start of routing (routers may change the value). By an
5689 oversight, this default was not being set up when an address was tested by
5690 -bt or -bv, which affected the outcome if any router or filter referred to
5693 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5694 log only when "message" or "log_message" are set. However, if one of the
5695 conditions was an address verification, or a call to a nested ACL, the
5696 messages generated by the underlying test were being passed through. This
5697 no longer happens. The underlying message is available in $acl_verify_
5698 message for both "message" and "log_message" expansions, so it can be
5699 passed through if needed.
5701 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5702 new expansion $bh_ to give the encoded byte string without charset
5703 translation. Translation happens only if iconv() is available; HAVE_ICONV
5704 indicates this at build time. HEADERS_CHARSET gives the charset to
5705 translate to; headers_charset can change it in the configuration, and
5706 "headers charset" can change it in an individual filter file.
5708 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5709 that creates RFC 2047 encoded "words" labels them as that charset instead
5710 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5711 expansion operator; (ii) when Exim creates a From: line for a local
5712 message; (iii) when a header line is rewritten to include a "phrase" part.
5714 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5715 buggy, causing it to skip the first lines of messages whose message ID
5716 ended in 'D'. This would not have bitten before Exim release 4.14, because
5717 message IDs were unlikely to end in 'D' before then. The effect was to have
5718 incorrect size information for certain domains.
5720 45. #include "config.h" was missing at the start of the crypt16.c module. This
5721 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5724 46. If there was a timeout during a "random" callout check, Exim treated it as
5725 a failure of the random address, and carried on sending RSET and the real
5726 address. If the delay was just some slowness somewhere, the response to the
5727 original RCPT would be taken as a response to RSET and so on, causing
5728 mayhem of various kinds.
5730 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5731 when I implemented it. It didn't allow for the fact that some option values
5732 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5733 the right test for positive values.
5735 48. Domain names in DNS records are case-independent. Exim always looks them up
5736 in lower case. Some resolvers return domain names in exactly the case they
5737 appear in the zone file, that is, they may contain uppercase letters. Not
5738 all resolvers do this - some return always lower case. Exim was treating a
5739 change of case by a resolver as a change of domain, similar to a widening
5740 of a domain abbreviation. This triggered its re-routing code and so it was
5741 trying to route what was effectively the same domain again. This normally
5742 caused routing to fail (because the router wouldn't handle the domain
5743 twice). Now Exim checks for this case specially, and just changes the
5744 casing of the domain that it ultimately uses when it transmits the message
5747 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5750 50. If a filter generated a file delivery with a non-absolute name (possible if
5751 no home directory exists for the router), the forbid_file option was not
5754 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5755 the existing equality matching.
5757 52. Exim was using ints instead of ino_t variables in some places where it was
5758 dealing with inode numbers.
5760 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5761 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5762 called TMPDIR, and if it finds it is different, it changes its value.
5764 54. The smtp_printf() function is now made available to local_scan() so
5765 additional output lines can be written before returning. There is also an
5766 smtp_fflush() function to enable the detection of a dropped connection.
5767 The variables smtp_input and smtp_batched_input are exported to
5770 55. Changed the default runtime configuration: the message "Unknown user"
5771 has been removed from the ACL, and instead placed on the localuser router,
5772 using the cannot_route_message feature. This means that any verification
5773 failures that generate their own messages won't get overridden. Similarly,
5774 the "Unrouteable address" message that was in the ACL for unverifiable
5775 relay addresses has also been removed.
5777 56. Added hosts_avoid_esmtp to the smtp transport.
5779 57. The exicyclog script was not checking for the esoteric option
5780 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5781 will work only if exicyclog is run under the appropriate euid.
5783 58. Following a discussion on the list, the rules by which Exim recognises line
5784 endings on incoming messages have been changed. The -dropcr and drop_cr
5785 options are now no-ops, retained only for backwards compatibility. The
5786 following line terminators are recognized: LF CRLF CR. However, special
5787 processing applies to CR:
5789 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5790 nor a local message in the state where . is a terminator.
5792 (ii) If a bare CR is encountered in a header line, an extra space is added
5793 after the line terminator so as not to end the header. The reasoning
5794 behind this is that bare CRs in header lines are most likely either
5795 to be mistakes, or people trying to play silly games.
5797 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5798 was being incorrectly given as 18 bytes larger than it should have been.
5799 This is a VOB (very old bug).
5801 60. This may never have affected anything current, but just in case it has:
5802 When the local host is found other than at the start of a list of hosts,
5803 the local host, those with the same MX, and any that follow, are discarded.
5804 When the list in question was part of a longer list of hosts, the following
5805 hosts (not currently being processed) were also being discarded. This no
5806 longer happens. I'm not sure if this situation could ever has previously
5809 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5812 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5813 header. This is something that is recommended in a new Internet Draft, and
5814 is something that is documented as being done by Sendmail. There are two
5815 possible values. For messages generated by the autoreply transport, Exim
5818 Auto-Submitted: auto-replied
5820 whereas for all other generated messages (e.g. bounces) it adds
5822 Auto-Submitted: auto-generated
5824 63. The "personal" condition in filters now includes a test for the
5825 Auto-Submitted: header. If it contains the string "auto-" the message it
5826 not considered personal.
5828 64. Added rcpt_include_affixes as a generic transport option.
5830 65. Added queue_only_override (default true).
5832 66. Added the syslog_duplication option.
5834 67. If what should have been the first header line of a message consisted of
5835 a space followed by a colon, Exim was mis-interpreting it as a header line.
5836 It isn't of course - it is syntactically invalid and should therefore be
5837 treated as the start of the message body. The misbehaviour could have
5838 caused a number of strange effects, including loss of data in subsequent
5839 header lines, and spool format errors.
5841 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5842 client host had authenticated. This control can now be exercised by an ACL
5843 for more flexibility.
5845 69. By default, callouts do not happen when testing with -bh. There is now a
5846 variant, -bhc, which does actually run the callout code, including
5847 consulting and updating the callout cache.
5849 70. Added support for saslauthd authentication, courtesy of Alexander
5852 71. If statvfs() failed on the spool or log directories while checking their
5853 size for availability, Exim confusingly gave the error "space shortage".
5854 Furthermore, in debugging mode it crashed with a floating point exception.
5855 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5856 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5857 serious problem, Exim now writes to the main and panic logs when this
5858 happens, with details of the failure. It then refuses to accept the
5859 incoming message, giving the message "spool directory problem" or "log
5860 directory problem" with a 421 code for SMTP messages.
5862 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5863 0, 1, and 2 exist, because some OS complain for execs without them (see
5864 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5865 descriptors. However, the code omitted to check that the open succeeded,
5866 causing mysterious errors if for some reason the permissions on /dev/null
5867 got screwed. Now Exim writes a message to the main and panic logs, and
5868 bombs out if it can't open /dev/null.
5870 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5871 interact so that it is all more flexible. It is supposed to remain
5872 backwards compatible. Also added extra_local_interfaces.
5874 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5875 to bomb out with an assertion failure - to the client this appears as a
5876 connection drop. This problem occurs in the part of the code that was taken
5877 from the Samba project. Fortunately, the assertion is in a very simple
5878 function, so I have fixed this by reproducing the function inline in the
5879 one place where it is called, and arranging for authentication to fail
5880 instead of killing the process with assert().
5882 75. The SPA client code was not working when the server requested OEM rather
5883 than Unicode encoding.
5885 76. Added code to make require_files with a specific uid setting more usable in
5886 the case where statting the file as root fails - usually a non-root-mounted
5887 NFS file system. When this happens and the failure is EACCES, Exim now
5888 forks a subprocess and does the per-uid checking as the relevant uid.
5890 77. Added process_log_path.
5892 78. If log_file_path was not explicitly set, a setting of check_log_space or
5893 check_log_inodes was ignored.
5895 79. If a space check for the spool or log partitions fails, the incident is now
5896 logged. Of course, in the latter case the data may get lost...
5898 80. Added the %p formatting code to string_format() so that it can be used to
5899 print addresses in debug_print(). Adjusted all the address printing in the
5900 debugging in store.c to use %p rather than %d.
5902 81. There was a concern that a line of code in smtp_in.c could overflow a
5903 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5904 initially expressed, the concern was not well-founded, because trailing
5905 spaces are removed early. However, if the trailing spaces were followed by
5906 a NULL, they did not get removed, so the overflow was possible. Two fixes
5909 (a) I re-wrote the offending code in a cleaner fashion.
5910 (b) If an incoming SMTP command contains a NULL character, it is rejected
5913 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5914 runs initgroups(), so that if the Exim user is in any additional groups,
5915 they will be used during message reception.
5921 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob