1 Change log file for Exim from version 4.21
2 -------------------------------------------
4 HS/04 Add support for keep_environment and add_environment options.
8 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
10 PP/02 Make -n do something, by making it not do something.
11 When combined with -bP, the name of an option is not output.
13 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
16 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
17 $sender_host_name and config options to manage this, and basic check
20 PP/05 DSCP support for outbound connections and control modifier for inbound.
22 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
23 (Only plugin which currently uses this is kerberos4, which nobody should
24 be using, but we should make it available and other future plugins might
25 conceivably use it, even though it would break NAT; stuff *should* be
26 using channel bindings instead).
28 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
29 name; added for Sendmail compatibility; requires admin caller.
30 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
31 just ignore it); requires trusted caller.
32 Also parse but ignore: -Ac -Am -X<logfile>
35 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
37 TL/02 Add +smtp_confirmation as a default logging option.
39 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
40 Patch by Magnus Holmgren from 2007-02-20.
42 TL/04 Bugzilla 1281 - Spec typo.
43 Bugzilla 1283 - Spec typo.
44 Bugzilla 1290 - Spec grammar fixes.
46 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
48 TL/06 Add Experimental DMARC support using libopendmarc libraries.
50 TL/07 Fix an out of order global option causing a segfault. Reported to dev
51 mailing list by by Dmitry Isaikin.
53 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
55 JH/02 Support "G" suffix to numbers in ${if comparisons.
57 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
59 NM/01 Bugzilla 1197 - Spec typo
60 Bugzilla 1196 - Spec examples corrections
62 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
64 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
65 gnutls_enable_pkcs11, but renamed to more accurately indicate its
68 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
69 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
71 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
72 "acl {{name}{arg}...}", and optional args on acl condition
75 JH/05 Permit multiple router/transport headers_add/remove lines.
77 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
79 JH/07 Avoid using a waiting database for a single-message-only transport.
80 Performance patch from Paul Fisher. Bugzilla 1262.
82 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
85 JH/09 Add $headers_added variable, with content from use of ACL modifier
86 add_header (but not yet added to the message). Bugzilla 199.
88 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
89 Pulled from Bugzilla 817 by Wolfgang Breyha.
91 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
93 (nb: this is the same fix as in Exim 4.80.1)
95 JH/11 Add A= logging on delivery lines, and a client_set_id option on
98 JH/12 Add optional authenticated_sender logging to A= and a log_selector
101 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
103 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
104 advertise SMTP AUTH mechanism to us, instead of a generic
105 protocol violation error. Also, make Exim more robust to bad
106 data from the Dovecot auth socket.
108 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
110 When a queue runner is handling a message, Exim first routes the
111 recipient addresses, during which it prunes them based on the retry
112 hints database. After that it attempts to deliver the message to
113 any remaining recipients. It then updates the hints database using
116 So if a recipient address works intermittently, it can get repeatedly
117 deferred at routing time. The retry hints record remains fresh so the
118 address never reaches the final cutoff time.
120 This is a fairly common occurrence when a user is bumping up against
121 their storage quota. Exim had some logic in its local delivery code
122 to deal with this. However it did not apply to per-recipient defers
123 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
125 This change adds a proper retry rule check during routing so that the
126 final cutoff time is checked against the message's age. We only do
127 this check if there is an address retry record and there is not a
128 domain retry record; this implies that previous attempts to handle
129 the address had the retry_use_local_parts option turned on. We use
130 this as an approximation for the destination being like a local
131 delivery, as in LMTP.
133 I suspect this new check makes the old local delivery cutoff check
134 redundant, but I have not verified this so I left the code in place.
136 TF/02 Correct gecos expansion when From: is a prefix of the username.
138 Test 0254 submits a message to Exim with the header
142 When I ran the test suite under the user fanf2, Exim expanded
143 the header to contain my full name, whereas it should have added
144 a Resent-Sender: header. It erroneously treats any prefix of the
145 username as equal to the username.
147 This change corrects that bug.
149 GF/01 DCC debug and logging tidyup
150 Error conditions log to paniclog rather than rejectlog.
151 Debug lines prefixed by "DCC: " to remove any ambiguity.
153 TF/03 Avoid unnecessary rebuilds of lookup-related code.
155 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
156 Bug spotted by Jeremy Harris; was flawed since initial commit.
157 Would have resulted in OCSP responses post-SNI triggering an Exim
158 NULL dereference and crash.
160 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
162 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
163 Bug detection, analysis and fix by Samuel Thibault.
164 Bugzilla 1331, Debian bug #698092.
166 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
168 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
169 Server implementation by Todd Lyons, client by JH.
170 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
171 config variable "prdr_enable" controls whether the server
172 advertises the facility. If the client requests PRDR a new
173 acl_data_smtp_prdr ACL is called once for each recipient, after
174 the body content is received and before the acl_smtp_data ACL.
175 The client is controlled by bolth of: a hosts_try_prdr option
176 on the smtp transport, and the server advertisement.
177 Default client logging of deliveries and rejections involving
178 PRDR are flagged with the string "PRDR".
180 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
181 fclose(). Diagnosis by Todd Lyons.
183 PP/17 Update configure.default to handle IPv6 localhost better.
184 Patch by Alain Williams (plus minor tweaks).
187 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
188 This is now consistent with GnuTLS, and is now documented: the
189 previous undocumented portable approach to treating the option as
190 unset was to force an expansion failure. That still works, and
191 an empty string is now equivalent.
193 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
194 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
195 not performing validation itself.
197 PP/20 Added force_command boolean option to pipe transport.
198 Patch from Nick Koston, of cPanel Inc.
200 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
203 TF/04 Added udpsend ACL modifer and hexquote expansion operator
205 PP/21 Fix eximon continuous updating with timestamped log-files.
206 Broken in a format-string cleanup in 4.80, missed when I repaired the
207 other false fix of the same issue.
208 Report and fix from Heiko Schlichting.
211 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
212 Report from Prashanth Katuri.
214 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
215 It's SecureTransport, so affects any MacOS clients which use the
216 system-integrated TLS libraries, including email clients.
218 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
219 using a MIME ACL for non-SMTP local injection.
220 Report and assistance in diagnosis by Warren Baker.
222 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
224 JH/16 Fix comparisons for 64b. Bugzilla 1385.
226 TL/09 Add expansion variable $authenticated_fail_id to keep track of
227 last id that failed so it may be referenced in subsequent ACL's.
229 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
232 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
233 ldap library initialization, allowing self-signed CA's to be
234 used. Also properly sets require_cert option later in code by
235 using NULL (global ldap config) instead of ldap handle (per
236 session). Bug diagnosis and testing by alxgomz.
238 TL/12 Enhanced documentation in the ratelimit.pl script provided in
239 the src/util/ subdirectory.
241 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
242 renamed to Transport Post Delivery Action by Jeremy Harris, as
245 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
246 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
247 redis_servers = needs to be configured which will be used by the redis
248 lookup. Patch from Warren Baker, of The Packet Hub.
250 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
252 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
253 hostname or reverse DNS when processing a host list. Used suggestions
254 from multiple comments on this bug.
256 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
258 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
259 Missed a few lines, added it to make the runtest require no keyboard
262 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
263 contains upper case chars. Make router use caseful_local_part.
265 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
266 support when GnuTLS has been built with p11-kit.
272 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
274 This, or similar/improved, will also be change PP/11 of 4.82.
280 PP/01 Handle short writes when writing local log-files.
281 In practice, only affects FreeBSD (8 onwards).
282 Bugzilla 1053, with thanks to Dmitry Isaikin.
284 NM/01 Bugzilla 949 - Documentation tweak
286 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
289 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
291 PP/02 Implemented gsasl authenticator.
293 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
295 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
296 `pkg-config foo` for cflags/libs.
298 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
299 with rest of GSASL and with heimdal_gssapi.
301 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
302 `pkg-config foo` for cflags/libs for the TLS implementation.
304 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
305 properties get this fed in as external SSF. A number of robustness
306 and debugging improvements to the cyrus_sasl authenticator.
308 PP/08 cyrus_sasl server now expands the server_realm option.
310 PP/09 Bugzilla 1214 - Log authentication information in reject log.
311 Patch by Jeremy Harris.
313 PP/10 Added dbmjz lookup type.
315 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
317 PP/12 MAIL args handles TAB as well as SP, for better interop with
318 non-compliant senders.
319 Analysis and variant patch by Todd Lyons.
321 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
322 Bug report from Lars Müller <lars@samba.org> (via SUSE),
323 Patch from Dirk Mueller <dmueller@suse.com>
325 PP/13 tls_peerdn now print-escaped for spool files.
326 Observed some $tls_peerdn in wild which contained \n, which resulted
327 in spool file corruption.
329 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
330 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
331 or write after TLS renegotiation, which otherwise led to messages
334 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
335 as a tracking header (ie: a signed header comes before the signature).
336 Patch from Wolfgang Breyha.
338 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
339 comma-sep list; embedded commas doubled.
341 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
343 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
345 Report and patch from Dmitry Banschikov.
347 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
348 Removed SSL_clear() after SSL_new() which led to protocol negotiation
349 failures. We appear to now support TLS1.1+ with Exim.
351 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
352 lets Exim select keys and certificates based upon TLS SNI from client.
353 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
354 before an outbound SMTP session. New log_selector, +tls_sni.
356 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
357 NULL dereference. Report and patch from Alun Jones.
359 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
360 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
361 Not seeing resolver debug output on NetBSD, but suspect this is a
362 resolver implementation change.
364 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
365 Left warnings. Added "eximon gdb" invocation mode.
367 PP/21 Defaulting "accept_8bitmime" to true, not false.
369 PP/22 Added -bw for inetd wait mode support.
371 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
372 locate the relevant includes and libraries. Made this the default.
374 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
375 Bugzilla 1246, report and most of solution from Tomasz Kusy.
377 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
378 This may cause build issues on older platforms.
380 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
381 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
382 gnutls_require_mac & gnutls_require_protocols (no longer supported).
383 Added SNI support via GnuTLS too.
384 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
386 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
388 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
389 Applied second patch from Janne, implementing suggestion to default
390 multiple-strings-in-record handling to match SPF spec.
392 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
394 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
395 read-only, out of scope).
396 Patch from Wolfgang Breyha, report from Stuart Northfield.
398 PP/29 Fix three issues highlighted by clang analyser static analysis.
399 Only crash-plausible issue would require the Cambridge-specific
400 iplookup router and a misconfiguration.
401 Report from Marcin Mirosław.
403 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
405 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
406 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
407 As part of this, removing so much warning spew let me fix some minor
408 real issues in debug logging.
410 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
411 assignment on my part. Fixed.
413 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
414 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
415 Janne Snabb (who went above and beyond: thank you).
417 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
418 string otherwise requires a connection and a bunch more work and it's
419 relatively easy to get wrong. Should also expose TLS library linkage
422 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
423 64-bit ${eval} (JH/03).
425 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
426 GNU libc to support some of the 64-bit stuff, should not lead to
427 conflicts. Defined before os.h is pulled in, so if a given platform
428 needs to override this, it can.
430 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
431 protection layer was required, which is not implemented.
432 Bugzilla 1254, patch from Wolfgang Breyha.
434 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
435 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
436 tls_dhparam take prime identifiers. Also unbreak combination of
437 OpenSSL+DH_params+TLSSNI.
439 PP/39 Disable SSLv2 by default in OpenSSL support.
445 PP/01 Solaris build fix for Oracle's LDAP libraries.
446 Bugzilla 1109, patch from Stephen Usher.
448 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
450 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
453 TF/02 Fix a couple more cases where we did not log the error message
454 when unlink() failed. See also change 4.74-TF/03.
456 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
457 lock up or crash if it happened to be inside a call to libc when it
458 got a SIGUSR1 from exiwhat.
460 The SIGUSR1 handler appends the current process status to the process
461 log which is later printed by exiwhat. It used to use the general
462 purpose logging code to do this, but several functions it calls are
463 not safe for signals.
465 The new output code in the SIGUSR1 handler is specific to the process
466 log, and simple enough that it's easy to inspect for signal safety.
467 Removing some special cases also simplifies the general logging code.
468 Removing the spurious timestamps from the process log simplifies
471 TF/04 Improved ratelimit ACL condition.
473 The /noupdate option has been deprecated in favour of /readonly which
474 has clearer semantics. The /leaky, /strict, and /readonly update modes
475 are mutually exclusive. The update mode is no longer included in the
476 database key; it just determines when the database is updated. (This
477 means that when you upgrde Exim will forget old rate measurements.)
479 Exim now checks that the per_* options are used with an update mode that
480 makes sense for the current ACL. For example, when Exim is processing a
481 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
482 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
483 must specify per_mail/readonly. If you omit the update mode it defaults to
484 /leaky where that makes sense (as before) or /readonly where required.
486 The /noupdate option is now undocumented but still supported for
487 backwards compatibility. It is equivalent to /readonly except that in
488 ACLs where /readonly is required you may specify /leaky/noupdate or
489 /strict/noupdate which are treated the same as /readonly.
491 A useful new feature is the /count= option. This is a generalization
492 of the per_byte option, so that you can measure the throughput of other
493 aggregate values. For example, the per_byte option is now equivalent
494 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
496 The per_rcpt option has been generalized using the /count= mechanism
497 (though it's more complicated than the per_byte equivalence). When it is
498 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
499 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
500 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
501 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
502 Note that using per_rcpt with a non-readonly update mode in more than
503 one ACL will cause the recipients to be double-counted. (The per_mail
504 and per_byte options don't have this problem.)
506 The handling of very low rates has changed slightly. If the computed rate
507 is less than the event's count (usually one) then this event is the first
508 after a long gap. In this case the rate is set to the same as this event's
509 count, so that the first message of a spam run is counted properly.
511 The major new feature is a mechanism for counting the rate of unique
512 events. The new per_addr option counts the number of different
513 recipients that someone has sent messages to in the last time period. It
514 behaves like per_rcpt if all the recipient addresses are different, but
515 duplicate recipient addresses do not increase the measured rate. Like
516 the /count= option this is a general mechanism, so the per_addr option
517 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
518 example, measure the rate that a client uses different sender addresses
519 with the options per_mail/unique=$sender_address. There are further
520 details in the main documentation.
522 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
524 TF/06 Removed a few PCRE remnants.
526 TF/07 Automatically extract Exim's version number from tags in the git
527 repository when doing development or release builds.
529 PP/02 Raise smtp_cmd_buffer_size to 16kB.
530 Bugzilla 879. Patch from Paul Fisher.
532 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
533 Heavily based on revision 40f9a89a from Simon Arlott's tree.
536 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
538 PP/05 Variable $av_failed, true if the AV scanner deferred.
539 Bugzilla 1078. Patch from John Horne.
541 PP/06 Stop make process more reliably on build failure.
542 Bugzilla 1087. Patch from Heiko Schlittermann.
544 PP/07 Make maildir_use_size_file an _expandable_ boolean.
545 Bugzilla 1089. Patch from Heiko Schlittermann.
547 PP/08 Handle ${run} returning more data than OS pipe buffer size.
548 Bugzilla 1131. Patch from Holger Weiß.
550 PP/09 Handle IPv6 addresses with SPF.
551 Bugzilla 860. Patch from Wolfgang Breyha.
553 PP/10 GnuTLS: support TLS 1.2 & 1.1.
555 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
558 PP/11 match_* no longer expand right-hand-side by default.
559 New compile-time build option, EXPAND_LISTMATCH_RHS.
560 New expansion conditions, "inlist", "inlisti".
562 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
564 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
566 PP/14 fix log_write() format string regression from TF/03.
567 Bugzilla 1152. Patch from Dmitry Isaikin.
573 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
575 PP/02 Harmonised TLS library version reporting; only show if debugging.
576 Layout now matches that introduced for other libraries in 4.74 PP/03.
578 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
580 PP/04 New "dns_use_edns0" global option.
582 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
585 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
586 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
588 TK/01 Updated PolarSSL code to 0.14.2.
589 Bugzilla 1097. Patch from Andreas Metzler.
591 PP/07 Catch divide-by-zero in ${eval:...}.
594 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
597 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
598 format-string attack -- SECURITY: remote arbitrary code execution.
600 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
601 time unintentionally subject to list matching rules, letting the header
602 cause arbitrary Exim lookups (of items which can occur in lists, *not*
603 arbitrary string expansion). This allowed for information disclosure.
605 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
606 INT_MIN/-1 -- value coerced to INT_MAX.
612 NM/01 Workround for PCRE version dependency in version reporting
615 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
616 This fixes portability to compilers other than gcc, notably
617 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
619 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
620 makefiles for portability to HP-UX and POSIX correctness.
622 PP/01 Permit LOOKUP_foo enabling on the make command-line.
623 Also via indented variable definition in the Makefile.
624 (Debugging by Oliver Heesakkers).
626 PP/02 Restore caching of spamd results with expanded spamd_address.
627 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
629 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
630 Improves build reliability. Fix from: Frank Elsner
632 NM/02 Fix wide character breakage in the rfc2047 coding
633 Fixes bug 1064. Patch from Andrey N. Oktyabrski
635 NM/03 Allow underscore in dnslist lookups
636 Fixes bug 1026. Patch from Graeme Fowler
638 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
639 Code patches from Adam Ciarcinski of NetBSD.
641 NM/04 Fixed exiqgrep to cope with mailq missing size issue
644 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
645 is logged, to avoid truncation. Patch from John Horne.
647 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
648 Patch from Jakob Hirsch.
650 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
651 SQL string expansion failure details.
652 Patch from Andrey Oktyabrski.
654 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
655 Patch from Simon Arlott.
657 PP/09 New lookups functionality failed to compile on old gcc which rejects
658 extern declarations in function scope.
659 Patch from Oliver Fleischmann
661 PP/10 Use sig_atomic_t for flags set from signal handlers.
662 Check getgroups() return and improve debugging.
663 Fixed developed for diagnosis in bug 927 (which turned out to be
666 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
667 Patch from Mark Zealey.
669 PP/12 Bugzilla 1056: Improved spamd server selection.
670 Patch from Mark Zealey.
672 PP/13 Bugzilla 1086: Deal with maildir quota file races.
673 Based on patch from Heiko Schlittermann.
675 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
676 Patch from Uwe Doering, sign-off by Michael Haardt.
678 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
679 variable declaration deep within a block. Bug and patch from
682 PP/15 lookups-Makefile IRIX compatibilty coercion.
684 PP/16 Make DISABLE_DKIM build knob functional.
686 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
687 Patch by Simon Arlott
689 TF/03 Fix valgrind.h portability to C89 compilers that do not support
690 variable argument macros. Our copy now differs from upstream.
696 TF/01 Failure to get a lock on a hints database can have serious
697 consequences so log it to the panic log.
699 TF/02 Log LMTP confirmation messages in the same way as SMTP,
700 controlled using the smtp_confirmation log selector.
702 TF/03 Include the error message when we fail to unlink a spool file.
704 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
705 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
706 for maintaining out-of-tree patches for some time.
708 PP/01 Bugzilla 139: Documentation and portability issues.
709 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
710 Handle per-OS dynamic-module compilation flags.
712 PP/02 Let /dev/null have normal permissions.
713 The 4.73 fixes were a little too stringent and complained about the
714 permissions on /dev/null. Exempt it from some checks.
715 Reported by Andreas M. Kirchwitz.
717 PP/03 Report version information for many libraries, including
718 Exim version information for dynamically loaded libraries. Created
719 version.h, now support a version extension string for distributors
720 who patch heavily. Dynamic module ABI change.
722 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
723 privilege escalation vulnerability whereby the Exim run-time user
724 can cause root to append content of the attacker's choosing to
727 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
730 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
731 If dropping privileges for untrusted macros, we disabled normal logging
732 on the basis that it would fail; for the Exim run-time user, this is not
733 the case, and it resulted in successful deliveries going unlogged.
734 Fixed. Reported by Andreas Metzler.
740 PP/01 Date: & Message-Id: revert to normally being appended to a message,
741 only prepend for the Resent-* case. Fixes regression introduced in
742 Exim 4.70 by NM/22 for Bugzilla 607.
744 PP/02 Include check_rfc2047_length in configure.default because we're seeing
745 increasing numbers of administrators be bitten by this.
747 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
749 PP/03 Bugzilla 994: added openssl_options main configuration option.
751 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
753 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
755 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
757 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
758 without a peer certificate, leading to a segfault because of an
759 assumption that peers always have certificates. Be a little more
760 paranoid. Problem reported by Martin Tscholak.
762 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
763 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
764 NB: ClamAV planning to remove STREAM in "middle of 2010".
765 CL also introduces -bmalware, various -d+acl logging additions and
766 more caution in buffer sizes.
768 PP/09 Implemented reverse_ip expansion operator.
770 PP/10 Bugzilla 937: provide a "debug" ACL control.
772 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
774 PP/12 Bugzilla 973: Implement --version.
776 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
778 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
780 PP/15 Bugzilla 816: support multiple condition rules on Routers.
782 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
783 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
784 ignore trailing whitespace.
786 JJ/02 prevent non-panic DKIM error from being sent to paniclog
788 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
791 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
792 Notification from Dr Andrew Aitchison.
794 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
795 ExtendedDetectionInfo response format.
796 Notification from John Horne.
798 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
801 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
802 XSL and documented dependency on system catalogs, with examples of how
805 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
808 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
809 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
810 configuration file which is writeable by the Exim user or group.
812 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
813 of configuration files to cover files specified with the -C option if
814 they are going to be used with root privileges, not just the default
817 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
818 option (effectively making it always true).
820 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
821 files to be used while preserving root privileges.
823 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
824 that rogue child processes cannot use them.
826 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
827 run-time user, instead of root.
829 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
830 Exim run-time user without dropping privileges.
832 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
833 result string, instead of calling string_vformat() twice with the same
836 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
837 for other users. Others should always drop root privileges if they use
838 -C on the command line, even for a whitelisted configure file.
840 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
842 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
848 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
849 $data_path, and $header_path variables; fixed documentation bugs and
852 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
853 exipick to access non-standard spools, including the "frozen" queue
856 NM/01 Bugzilla 965: Support mysql stored procedures.
857 Patch from Alain Williams
859 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
861 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
862 Patch from Andreas Metzler
864 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
865 Patch from Kirill Miazine
867 NM/05 Bugzilla 671: Added umask to procmail example.
869 JJ/03 installed exipick 20100323.0, fixing doc bug
871 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
872 directory. Notification and patch from Dan Rosenberg.
874 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
876 TK/02 Improve log output when DKIM signing operation fails.
878 MH/01 Treat the transport option dkim_domain as a colon separated
879 list, not as a single string, and sign the message with each element,
880 omitting multiple occurences of the same signer.
882 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
883 Bugzilla 985, 986. Patch by Simon Arlott
885 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
886 Patch by Simon Arlott
888 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
889 MBX locking. Notification from Dan Rosenberg.
895 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
897 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
899 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
901 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
903 NM/04 Bugzilla 915: Flag broken perl installation during build.
909 TK/01 Added patch by Johannes Berg that expands the main option
910 "spamd_address" if it starts with a dollar sign.
912 TK/02 Write list of recipients to X-Envelope-Sender header when building
913 the mbox-format spool file for content scanning (suggested by Jakob
916 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
917 (http://www.dcc-servers.net/) support via dccifd. Activated by
918 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
920 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
921 by Mark Daniel Reidel <mr@df.eu>.
923 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
924 When building exim an external PCRE library is now needed -
925 PCRE is a system library on the majority of modern systems.
926 See entry on PCRE_LIBS in EDITME file.
928 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
929 conversation. Added nologin parameter to request.
930 Patch contributed by Kirill Miazine.
932 TF/01 Do not log submission mode rewrites if they do not change the address.
934 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
936 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
937 log files in place. Contributed by Roberto Lima.
939 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
941 TF/03 Bugzilla 615: When checking the local_parts router precondition
942 after a local_part_suffix or local_part_prefix option, Exim now
943 does not use the address's named list lookup cache, since this
944 contains cached lookups for the whole local part.
946 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
947 Robert Millan. Documentation is in experimental-spec.txt.
949 TF/04 Bugzilla 668: Fix parallel build (make -j).
951 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
953 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
954 Patch provided by Jan Srzednicki.
956 TF/05 Leading white space used to be stripped from $spam_report which
957 wrecked the formatting. Now it is preserved.
959 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
960 that they are available at delivery time.
962 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
964 TF/08 TLS error reporting now respects the incoming_interface and
965 incoming_port log selectors.
967 TF/09 Produce a more useful error message if an SMTP transport's hosts
968 setting expands to an empty string.
970 NM/06 Bugzilla 744: EXPN did not work under TLS.
971 Patch provided by Phil Pennock.
973 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
974 Patch provided by Richard Godbee.
976 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
977 acl_smtp_notquit, added index entry.
979 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
980 Patch provided by Eugene Bujak.
982 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
983 accept(). Patch provided by Maxim Dounin.
985 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
986 Patch provided by Phil Pennock.
988 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
990 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
991 Patch provided by Brad "anomie" Jorsch.
993 NM/14 Bugzilla 622: Added timeout setting to transport filter.
994 Patch provided by Dean Brooks.
996 TK/05 Add native DKIM support (does not depend on external libraries).
998 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
999 Patch provided by Graeme Fowler.
1001 NM/16 Bugzilla 851: Documentation example syntax fix.
1003 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1005 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1008 NM/19 Bugzilla 745: TLS version reporting.
1009 Patch provided by Phil Pennock.
1011 NM/20 Bugzilla 167: bool: condition support.
1012 Patch provided by Phil Pennock.
1014 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1015 clients. Patch provided by Phil Pennock.
1017 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1018 Patch provided by Brad "anomie" Jorsch.
1020 NM/23 Bugzilla 687: Fix misparses in eximstats.
1021 Patch provided by Heiko Schlittermann.
1023 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1024 Patch provided by Heiko Schlittermann.
1026 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1027 plus update to original patch.
1029 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1031 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1032 Patch provided by David Brownlee.
1034 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1036 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1038 NM/30 Bugzilla 888: TLS documentation bugfixes.
1040 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1042 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1043 Unlike the original bugzilla I have changed all shell scripts in src tree.
1045 NM/33 Bugzilla 898: Transport filter timeout fix.
1046 Patch by Todd Rinaldo.
1048 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1049 Patch by Serge Demonchaux.
1051 NM/35 Bugzilla 39: Base64 decode bug fixes.
1052 Patch by Jakob Hirsch.
1054 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1056 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1058 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1060 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1066 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1067 ALT-N's libdkim that I have put here:
1068 http://duncanthrax.net/exim-experimental/
1070 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1071 were called 'true' and it seems that C99 defines that as a reserved
1072 keyword to be used with 'bool' variable types. That means you could
1073 not include C99-style headers which use bools without triggering
1074 build errors in sieve.c.
1076 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1077 as mailq or other aliases. Changed the --help handling significantly
1078 to do whats expected. exim_usage() emits usage/help information.
1080 SC/01 Added the -bylocaldomain option to eximstats.
1082 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1084 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1086 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1092 PH/01 Another patch from the Sieve maintainer.
1094 PH/02 When an IPv6 address is converted to a string for single-key lookup
1095 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1096 dots are used instead of colons so that keys in lsearch files need not
1097 contain colons. This was done some time before quoting was made available
1098 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1099 (notated using the quote facility) so as to distinguish them from IPv4
1100 keys. This meant that lookups for IP addresses in host lists did not work
1101 for iplsearch lookups.
1103 This has been fixed by arranging for IPv6 addresses to be expressed with
1104 colons if the lookup type is iplsearch. This is not incompatible, because
1105 previously such lookups could never work.
1107 The situation is now rather anomolous, since one *can* have colons in
1108 ordinary lsearch keys. However, making the change in all cases is
1109 incompatible and would probably break a number of configurations.
1111 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1114 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1115 conversion specification without a maximum field width, thereby enabling
1116 a rogue spamd server to cause a buffer overflow. While nobody in their
1117 right mind would setup Exim to query an untrusted spamd server, an
1118 attacker that gains access to a server running spamd could potentially
1119 exploit this vulnerability to run arbitrary code as the Exim user.
1121 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1122 $primary_hostname instead of what libspf2 thinks the hosts name is.
1124 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1125 a directory entry by the name of the lookup key. Previously, if a
1126 symlink pointed to a non-existing file or a file in a directory that
1127 Exim lacked permissions to read, a lookup for a key matching that
1128 symlink would fail. Now it is enough that a matching directory entry
1129 exists, symlink or not. (Bugzilla 503.)
1131 PH/03 The body_linecount and body_zerocount variables are now exported in the
1134 PH/04 Added the $dnslist_matched variable.
1136 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1137 This means they are set thereafter only if the connection becomes
1140 PH/06 Added the client_condition to authenticators so that some can be skipped
1141 by clients under certain conditions.
1143 PH/07 The error message for a badly-placed control=no_multiline_responses left
1144 "_responses" off the end of the name.
1146 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1148 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1149 (without spaces) instead of just copying the configuration text.
1151 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1153 PH/11 Added $max_received_linelength.
1155 PH/12 Added +ignore_defer and +include_defer to host lists.
1157 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1158 way in which PCRE > 7.0 is built.
1160 PH/14 Implemented queue_only_load_latch.
1162 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1163 MAIL command. The effect was to mangle the value on 64-bit systems.
1165 PH/16 Another patch from the Sieve maintainer.
1167 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1169 PH/18 If a system quota error occurred while trying to create the file for
1170 a maildir delivery, the message "Mailbox is full" was not appended to the
1171 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1172 only to a quota excession during the actual writing of the file.
1174 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1175 characters?) which causes problems in log lines. The DN values are now
1176 passed through string_printing() before being added to log lines.
1178 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1179 and InterBase are left for another time.)
1181 PH/21 Added message_body_newlines option.
1183 PH/22 Guard against possible overflow in moan_check_errorcopy().
1185 PH/23 POSIX allows open() to be a macro; guard against that.
1187 PH/24 If the recipient of an error message contained an @ in the local part
1188 (suitably quoted, of course), incorrect values were put in $domain and
1189 $local_part during the evaluation of errors_copy.
1195 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1196 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1199 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1200 issue a MAIL command.
1202 PH/02 In an ACL statement such as
1204 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1206 if a client was not listed at all, or was listed with a value other than
1207 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1208 the condition was not true (as it should be), so access was not denied.
1209 The bug was that the ! inversion was incorrectly passed on to the second
1210 item. This has been fixed.
1212 PH/03 Added additional dnslists conditions == and =& which are different from
1213 = and & when the dns lookup returns more than one IP address.
1215 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1216 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1218 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1219 FSYNC, which compiles an option called disable_fsync that allows for
1220 bypassing fsync(). The documentation is heavily laced with warnings.
1222 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1224 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1225 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1226 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1227 including adding "make clean"; (3) Added -fPIC when compiling the test
1228 dynamically loaded module, to get rid of a warning.
1230 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1231 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1232 = 0s. The bug is otherwise harmless.
1234 PH/07 There was a bug in the dovecot authenticator such that the value of
1235 $auth1 could be overwritten, and so not correctly preserved, after a
1236 successful authentication. This usually meant that the value preserved by
1237 the server_setid option was incorrect.
1239 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1241 PH/09 Installed PCRE release 7.0.
1243 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1244 run for batched SMTP input. It is now run at the start of every message
1245 in the batch. While fixing this I discovered that the process information
1246 (output by running exiwhat) was not always getting set for -bs and -bS
1247 input. This is fixed, and it now also says "batched" for BSMTP.
1249 PH/11 Added control=no_pipelining.
1251 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1252 patch, slightly modified), and move the expansion of helo_data till after
1253 the connection is made in the smtp transport (so it can use these
1256 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1258 PH/14 Added log_selector = +pid.
1260 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1262 PH/16 Add ${if forany and ${if forall.
1264 PH/17 Added dsn_from option to vary the From: line in DSNs.
1266 PH/18 Flush SMTP output before performing a callout, unless control =
1267 no_callout_flush is set.
1269 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1270 was true (the default) a successful delivery failed to delete the retry
1271 item, thus causing premature timeout of the address. The bug is now
1274 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1276 PH/21 Long custom messages for fakedefer and fakereject are now split up
1277 into multiline reponses in the same way that messages for "deny" and
1278 other ACL rejections are.
1280 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1281 with slight modification.
1283 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1284 draft, changing the syntax and factoring some duplicate code".
1286 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1287 for deliveries of the second and subsequent messages over the same SMTP
1290 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1291 ${reduce, with only minor "tidies".
1293 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1295 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1296 expansion side effects.
1298 PH/27 When a message times out after an over-quota error from an Exim-imposed
1299 quota, the bounce message says "mailbox is full". This message was not
1300 being given when it was a system quota that was exceeded. It now should
1303 MH/03 Made $recipients available in local_scan(). local_scan() already has
1304 better access to the recipient list through recipients_list[], but
1305 $recipients can be useful in postmaster-provided expansion strings.
1307 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1308 in the case of a MAIL command with additional options following the
1309 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1310 were accidentally chopped off.
1312 PH/29 SMTP synchronization checks are implemented when a command is read -
1313 there is a check that no more input is waiting when there shouldn't be
1314 any. However, for some commands, a delay in an ACL can mean that it is
1315 some time before the response is written. In this time, more input might
1316 arrive, invalidly. So now there are extra checks after an ACL has run for
1317 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1318 pipelining has not been advertised.
1320 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1322 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1323 returned in the SMTP response when smtp_return_error_details was false.
1324 This has been fixed.
1326 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1327 instead of the C I/O that was originally supplied, because problems were
1328 reported on Solaris.
1330 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1331 Exim which did not show up earlier: it was assuming that a call to
1332 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1333 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1334 was a macro that became an assignment, so it seemed to work. This has
1335 changed to a proper function call with a void return, hence the compile
1336 error. Exim's code has been fixed.
1338 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1341 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1343 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1345 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1346 because it turns out to be too big or there is a timeout) there is no
1347 "Completed" line in the log. When some messages of this type were
1348 selected by exigrep, they were listed as "not completed". Others were
1349 picked up by some special patterns. I have improved the selection
1350 criteria to be more general.
1352 PH/38 The host_find_failed option in the manualroute router can now be set
1353 to "ignore", to completely ignore a host whose IP address cannot be
1354 found. If all hosts are ignored, the behaviour is controlled by the new
1355 host_all_ignored option.
1357 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1358 homing or because of multiple MX records with /mx) generated more than
1359 one IP address, and the following item turned out to be the local host,
1360 all the secondary addresses of the first item were incorrectly removed
1361 from the list, along with the local host and any following hosts (which
1362 is what is supposed to happen).
1364 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1365 whoever called Exim into the -H file. In the case of the daemon it was
1366 behaving confusingly. When first started, it used values for whoever
1367 started the daemon, but after a SIGHUP it used the Exim user (because it
1368 calls itself on a restart). I have changed the code so that it now always
1371 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1372 message are rejected with the same error (e.g. no authentication or bad
1373 sender address), and a DATA command is nevertheless sent (as can happen
1374 with PIPELINING or a stupid MUA), the error message that was given to the
1375 RCPT commands is included in the rejection of the DATA command. This is
1376 intended to be helpful for MUAs that show only the final error to their
1379 PH/42 Another patch from the Sieve maintainer.
1381 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1382 Eximstats - Fixed some broken HTML links and added missing column headers
1384 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1385 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1387 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1389 PH/43 Yet another patch from the Sieve maintainer.
1391 PH/44 I found a way to check for a TCP/IP connection going away before sending
1392 the response to the final '.' that terminates a message, but only in the
1393 case where the client has not sent further data following the '.'
1394 (unfortunately, this is allowed). However, in many cases there won't be
1395 any further data because there won't be any more messages to send. A call
1396 to select() can be used: if it shows that the input is "ready", there is
1397 either input waiting, or the socket has been closed. An attempt to read
1398 the next input character can distinguish the two cases. Previously, Exim
1399 would have sent an OK response which the client would never have see.
1400 This could lead to message repetition. This fix should cure that, at
1401 least in a lot of common cases.
1403 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1404 advertised in response to EHLO.
1410 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1411 fixed by 4.65/MH/01 (is this a record?) are fixed:
1413 (i) An empty string was always treated as zero by the numeric comparison
1414 operators. This behaviour has been restored.
1416 (ii) It is documented that the numeric comparison operators always treat
1417 their arguments as decimal numbers. This was broken in that numbers
1418 starting with 0 were being interpreted as octal.
1420 While fixing these problems I realized that there was another issue that
1421 hadn't been noticed. Values of message_size_limit (both the global option
1422 and the transport option) were treated as octal if they started with 0.
1423 The documentation was vague. These values are now always treated as
1424 decimal, and I will make that clear in the documentation.
1430 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1431 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1434 MH/01 Don't check that the operands of numeric comparison operators are
1435 integers when their expansion is in "skipping" mode (fixes bug
1436 introduced by 4.64-PH/07).
1438 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1439 child addresses, Exim now panics and dies. Previously, because the count
1440 is held in a short int, deliveries were likely to be lost. As such a
1441 large number of recipients for a single message is ridiculous
1442 (performance will be very, very poor), I have chosen to impose a limit
1443 rather than extend the field.
1449 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1450 leftover -K file (the existence of which was triggered by #402).
1451 While we were at it, introduced process PID as part of the -K
1452 filename. This should rule out race conditions when creating
1455 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1456 processing considerably. Previous code took too long for large mails,
1457 triggering a timeout which in turn triggers #401.
1459 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1460 in the DK code in transports.c. sendfile() is not really portable,
1461 hence the _LINUX specificness.
1463 TF/01 In the add_headers option to the mail command in an Exim filter,
1464 there was a bug that Exim would claim a syntax error in any
1465 header after the first one which had an odd number of characters
1468 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1469 callout verification, Exim cached a "reject" for the entire domain. This
1470 is correct for most verifications, but it is not correct for a recipient
1471 verification with use_sender or use_postmaster set, because in that case
1472 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1473 case of MAIL FROM:<> rejection from other early rejections (e.g.
1474 rejection of HELO). When verifying a recipient using a non-null MAIL
1475 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1476 Whatever the result of the callout, the value of the domain cache is
1477 left unchanged (for any other kind of callout, getting as far as trying
1478 RCPT means that the domain itself is ok).
1480 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1483 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1484 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1485 session, the daemon ignores ECONNECT errors and logs others; it now
1486 ignores EPIPE as well.
1488 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1489 (quoted-printable decoding).
1491 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1492 later the small subsequent patch to fix an introduced bug.
1494 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1496 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1498 PH/08 An error is now given if message_size_limit is specified negative.
1500 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1501 to be given (somewhat) arbitrary names.
1503 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1506 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1507 miscellaneous code fixes
1509 PH/10 Added the log_reject_target ACL modifier to specify where to log
1512 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1513 hostname. This is wrong, because it relates to the incoming message (and
1514 probably the interface on which it is arriving) and not to the outgoing
1515 callout (which could be using a different interface). This has been
1516 changed to use the value of the helo_data option from the smtp transport
1517 instead - this is what is used when a message is actually being sent. If
1518 there is no remote transport (possible with a router that sets up host
1519 addresses), $smtp_active_hostname is used.
1521 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1522 tweaks were necessary in order to get it to work (see also 21 below):
1523 (a) The code assumed that strncpy() returns a negative number on buffer
1524 overflow, which isn't the case. Replaced with Exim's string_format()
1526 (b) There were several signed/unsigned issues. I just did the minimum
1527 hacking in of casts. There is scope for a larger refactoring.
1528 (c) The code used strcasecmp() which is not a standard C function.
1529 Replaced with Exim's strcmpic() function.
1530 (d) The code set only $1; it now sets $auth1 as well.
1531 (e) A simple test gave the error "authentication client didn't specify
1532 service in request". It would seem that Dovecot has changed its
1533 interface. Fortunately there's a specification; I followed it and
1534 changed what the client sends and it appears to be working now.
1536 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1539 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1540 address A is aliased to B and C, where B exists and C does not. Without
1541 -v the output is "A verified" because verification stops after a
1542 successful redirection if more than one address is generated. However,
1543 with -v the child addresses are also verified. Exim was outputting "A
1544 failed to verify" and then showing the successful verification for C,
1545 with its parentage. It now outputs "B failed to verify", showing B's
1546 parentage before showing the successful verification of C.
1548 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1549 look up a TXT record in a specific list after matching in a combined
1552 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1553 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1554 they consult the DNS. I had assumed they would set it the way they
1555 wanted; and indeed my experiments on Linux seem to show that in some
1556 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1557 To be on the safe side, however, I have now made the interface to
1558 host_find_byname() similar to host_find_bydns(), with an argument
1559 containing the DNS resolver options. The host_find_byname() function now
1560 sets these options at its start, just as host_find_bydns() does. The smtp
1561 transport options dns_qualify_single and dns_search_parents are passed to
1562 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1563 of host_find_byname() use the default settings of RES_DEFNAMES
1564 (qualify_single) but not RES_DNSRCH (search_parents).
1566 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1567 spool_read_header() do less string testing, by means of a preliminary
1568 switch on the second character of optional "-foo" lines. (This is
1569 overdue, caused by the large number of possibilities that now exist.
1570 Originally there were few.) While I was there, I also converted the
1571 str(n)cmp tests so they don't re-test the leading "-" and the first
1572 character, in the hope this might squeeze out yet more improvement.
1574 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1575 flag allowing group syntax was set by the header_syntax check but not
1576 turned off, possible causing trouble later; (2) The flag was not being
1577 set at all for the header_verify test, causing "group"-style headers to
1578 be rejected. I have now set it in this case, and also caused header_
1579 verify to ignore an empty address taken from a group. While doing this, I
1580 came across some other cases where the code for allowing group syntax
1581 while scanning a header line wasn't quite right (mostly, not resetting
1582 the flag correctly in the right place). These bugs could have caused
1583 trouble for malformed header lines. I hope it is now all correct.
1585 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1586 with the "reply" argument non-NULL. The code, however (which originally
1587 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1588 but it didn't always do it. This confused somebody who was copying the
1589 code for some other use. I have removed all the tests.
1591 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1592 feature that was used to support insecure browsers during the U.S. crypto
1593 embargo. It requires special client support, and Exim is probably the
1594 only MTA that supported it -- and would never use it because real RSA is
1595 always available. This code has been removed, because it had the bad
1596 effect of slowing Exim down by computing (never used) parameters for the
1597 RSA_EXPORT functionality.
1599 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1600 authenticator to fail if there's a tab character in the incoming data
1601 (there should never be unless someone is messing about, as it's supposed
1602 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1603 if the connection is using TLS or if the remote IP is the same as the
1604 local IP, and the "valid-client-cert option" if a client certificate has
1607 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1608 authenticators. This can be used for authorization after authentication
1609 succeeds. (In the case of plaintext, it servers for both authentication
1612 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1613 if any retry times were supplied.
1615 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1616 connection, where there is no client IP address to check. In this
1617 situation, the verify now always succeeds.
1619 PH/25 Applied John Jetmore's -Mset patch.
1621 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1623 PH/27 In a string expansion for a processed (not raw) header when multiple
1624 headers of the same name were present, leading whitespace was being
1625 removed from all of them, but trailing whitespace was being removed only
1626 from the last one. Now trailing whitespace is removed from each header
1627 before concatenation. Completely empty headers in a concatenation (as
1628 before) are ignored.
1630 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1631 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1633 PH/29 [Removed. This was a change that I later backed out, and forgot to
1634 correct the ChangeLog entry (that I had efficiently created) before
1635 committing the later change.]
1637 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1638 address errors (4xx response to RCPT) over the same connection as other
1639 messages routed to the same hosts. Such deliveries are always "forced",
1640 so retry times are not inspected. This resulted in far too many retries
1641 for the affected addresses. The effect occurred only when there were more
1642 hosts than the hosts_max_try setting in the smtp transport when it had
1643 the 4xx errors. Those hosts that it had tried were not added to the list
1644 of hosts for which the message was waiting, so if all were tried, there
1645 was no problem. Two fixes have been applied:
1647 (i) If there are any address or message errors in an SMTP delivery, none
1648 of the hosts (tried or untried) are now added to the list of hosts
1649 for which the message is waiting, so the message should not be a
1650 candidate for sending over the same connection that was used for a
1651 successful delivery of some other message. This seems entirely
1652 reasonable: after all the message is NOT "waiting for some host".
1653 This is so "obvious" that I'm not sure why it wasn't done
1654 previously. Hope I haven't missed anything, but it can't do any
1655 harm, as the worst effect is to miss an optimization.
1657 (ii) If, despite (i), such a delivery is accidentally attempted, the
1658 routing retry time is respected, so at least it doesn't keep
1659 hammering the server.
1661 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1662 in ${readsocket because some servers need this prod.
1664 PH/32 Added some extra debug output when updating a wait-xxx database.
1666 PH/33 The hint "could be header name not terminated by colon", which has been
1667 given for certain expansion errors for a long time, was not being given
1668 for the ${if def:h_colon_omitted{... case.
1670 PH/34 The spec says: "With one important exception, whenever a domain list is
1671 being scanned, $domain contains the subject domain." There was at least
1672 one case where this was not true.
1674 PH/35 The error "getsockname() failed: connection reset by peer" was being
1675 written to the panic log as well as the main log, but it isn't really
1676 panic-worthy as it just means the connection died rather early on. I have
1677 removed the panic log writing for the ECONNRESET error when getsockname()
1680 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1681 runs only) independently of the message's sender address. This meant
1682 that, if the 4xx error was in fact related to the sender, a different
1683 message to the same recipient with a different sender could confuse
1684 things. In particualar, this can happen when sending to a greylisting
1685 server, but other circumstances could also provoke similar problems.
1686 I have changed the default so that the retry time for these errors is now
1687 based a combination of the sender and recipient addresses. This change
1688 can be overridden by setting address_retry_include_sender=false in the
1691 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1692 remote server are returned as part of bounce messages. This was not
1693 happening for LMTP over a pipe (the lmtp transport), but now it is the
1694 same for both kinds of LMTP.
1696 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1697 in header lines that were in fact CNAMEs. This is no longer the case.
1699 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1700 and queue runs started by the daemon processed all messages. This has
1701 been fixed so that -R and -S can now usefully be given with -q<time>.
1703 PH/40 Import PCRE release 6.7 (fixes some bugs).
1705 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1707 PH/42 Give an error if -q is specified more than once.
1709 PH/43 Renamed the variables $interface_address and $interface_port as
1710 $received_ip_address and $received_port, to make it clear that these
1711 values apply to message reception, and not to the outgoing interface when
1712 a message is delivered. (The old names remain recognized, of course.)
1714 PH/44 There was no timeout on the connect() call when using a Unix domain
1715 socket in the ${readsocket expansion. There now is.
1717 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1718 be meaningful with "accept".
1720 SC/01 Eximstats V1.43
1721 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1723 SC/02 Eximstats V1.44
1724 Use a glob alias rather than an array ref in the generated
1725 parser. This improves both readability and performance.
1727 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1728 Collect SpamAssassin and rejection statistics.
1729 Don't display local sender or destination tables unless
1730 there is data to show.
1731 Added average volumes into the top table text output.
1733 SC/04 Eximstats V1.46
1734 Collect data on the number of addresses (recipients)
1735 as well as the number of messages.
1737 SC/05 Eximstats V1.47
1738 Added 'Message too big' to the list of mail rejection
1739 reasons (thanks to Marco Gaiarin).
1741 SC/06 Eximstats V1.48
1742 Mainlog lines which have GMT offsets and are too short to
1743 have a flag are now skipped.
1745 SC/07 Eximstats V1.49 (Alain Williams)
1746 Added the -emptyok flag.
1748 SC/08 Eximstats V1.50
1749 Fixes for obtaining the IP address from reject messages.
1751 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1752 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1753 whitesspace changes from 4.64-PH/27
1755 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1758 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1759 are found, allow negative numbers in numeric criteria)
1761 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1763 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1764 to match changes made in 4.64-PH/43
1766 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1768 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1769 "domain" is actually an IP address, and force a failure. This locks out
1770 those revolvers/nameservers that support "A-for-A" lookups, in
1771 contravention of the specifications.
1773 PH/48 When a host name was looked up from an IP address, and the subsequent
1774 forward lookup of the name timed out, the host name was left in
1775 $sender_host_name, contrary to the specification.
1777 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1778 restricted to single-key lookups, Exim was not diagnosing an error if
1779 * or *@ was used with a query-style lookup.
1781 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1783 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1784 long ago, but noone interested enough thought of it. Let's just say that
1785 the "1.1" means that there are some new functions that weren't there at
1786 some point in the past.
1788 PH/51 Error processing for expansion failure of helo_data from an smtp
1789 transport during callout processing was broken.
1791 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1792 tested/used via the -bh/-bhc/-bs options.
1794 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1795 bug, fixed in subsequent PCRE releases).
1797 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1798 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1800 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1806 SC/01 Use a glob alias rather than an array ref in eximstats generated
1807 parser. This improves both readability and performance.
1809 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1810 Don't display local sender or destination tables in eximstats unless
1811 there is data to show.
1812 Added average volumes into the eximstats top table text output.
1814 SC/03 Collect data on the number of addresses (recipients) as well
1815 as the number of messages in eximstats.
1817 TF/01 Correct an error in the documentation for the redirect router. Exim
1818 does (usually) call initgroups() when daemonizing.
1820 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1821 with consistent privilege compared to when running as a daemon.
1823 TF/03 Note in the spec that $authenticated_id is not set for local
1824 submissions from trusted users.
1826 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1827 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1829 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1830 by adding some example configuration directives to the default
1831 configuration file. A little bit of work is required to uncomment the
1832 directives and define how usernames and passwords are checked, but
1833 there is now a framework to start from.
1835 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1836 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1837 without this. I don't know how relevant this is to other LDAP libraries.
1839 PH/02 Add the verb name to the "unknown ACL verb" error.
1841 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1843 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1845 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1846 directory not expanded when it should be if an expanded home directory
1847 was set for the address (which is overridden by the transport).
1849 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1852 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1853 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1854 because it is too late at that time, and has no effect.
1856 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1857 security issue with \' (bugzilla #107). I could not use the
1858 PQescapeStringConn() function, because it needs a PGconn value as one of
1861 PH/08 When testing addresses using -bt, indicate those final addresses that
1862 are duplicates that would not cause an additional delivery. At least one
1863 person was confused, thinking that -bt output corresponded to deliveries.
1864 (Suppressing duplicates isn't a good idea as you lose the information
1865 about possibly different redirections that led to the duplicates.)
1867 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1868 systems where poll() doesn't work, in particular OS X.
1870 PH/10 Added more information to debugging output for retry time not reached.
1872 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1873 operations in malware.c.
1875 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1878 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1879 syslog_duplication set false, log lines that would normally be written
1880 both the the main log and to the reject log were not written to syslog at
1883 PH/14 In the default configuration, change the use of "message" in ACL warn
1884 statements to "add_header".
1886 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1887 not followed by a command (e.g. "seen endif").
1889 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1890 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1893 PH/17 Added extra conditions to the default value of delay_warning_condition
1897 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1898 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1899 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1902 The Auto-Submitted: and various List- headers are standardised, whereas I
1903 don't think Precedence: ever was.
1905 PH/18 Refactored debugging code in route_finduser() to show more information,
1906 in particular, the error code if getpwnam() issues one.
1908 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1909 This is apparently needed in addition to the PH/07 change above to avoid
1910 any possible encoding problems.
1912 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1913 but not after initializing Perl.
1915 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1916 output them only if debugging. By default they are written stderr,
1917 apparently, which is not desirable.
1919 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1922 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1925 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1927 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1928 authenticated or an ident call has been made. Suppress the default
1929 values for $authenticated_id and $authenticated_sender (but permit -oMai
1930 and -oMas) when testing with -bh.
1932 PH/24 Re-jigged the order of the tests in the default configuration so that the
1933 tests for valid domains and recipients precede the DNS black list and CSA
1934 tests, on the grounds that those ones are more expensive.
1936 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1937 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1938 This bug exists in every version of Exim that I still have, right back to
1941 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1942 However, an attempt to turn on case-sensitivity in a regex key by
1943 including (?-i) didn't work because the subject string was already
1944 lowercased, and the effects were non-intuitive. It turns out that a
1945 one-line patch can be used to allow (?-i) to work as expected.
1951 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1952 other effects) broke the use of negated acl sub-conditions.
1954 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1957 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1958 "Deny" causes Exim to reject the incoming connection with a 554 error.
1959 Unfortunately, if there is a major crisis, such as a disk failure,
1960 tcp-wrappers gives "deny", whereas what one would like would be some
1961 kind of temporary error. A kludge has been added to help with this.
1962 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1963 554 error is used if errno is still zero or contains ENOENT (which occurs
1964 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1967 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1969 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1970 errors. Otherwise a message that provokes a temporary error (when other
1971 messages do not) can cause a whole host to time out.
1973 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1974 addresses were routed directly to files or pipes from a redirect router.
1975 File deliveries just didn't batch; pipe deliveries might have suffered
1978 PH/06 A failure to get a lock for a hints database would erroneously always say
1979 "Failed to get write lock", even when it was really a read lock.
1981 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1982 of 0600. This has been changed to use the value of the lockfile_mode
1983 option (which defaults to 0600).
1985 PH/08 Applied small patch from the Sieve maintainer.
1987 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1988 folder from quota calculations, a direct delivery into this folder messed
1989 up the contents of the maildirsize file. This was because the regex was
1990 used only to exclude .Trash (or whatever) when the size of the mailbox
1991 was calculated. There was no check that a delivery was happening into an
1992 excluded directory. This bug has been fixed by ignoring all quota
1993 processing for deliveries into excluded directories.
1995 PH/10 Added the maildirfolder_create_regex option to appendfile.
2001 PH/01 The code for finding all the local interface addresses on a FreeBSD
2002 system running IPv6 was broken. This may well have applied to all BSD
2003 systems, as well as to others that have similar system calls. The broken
2004 code found IPv4 interfaces correctly, but gave incorrect values for the
2005 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2006 that it would not match correctly against @[] and not recognize the IPv6
2009 PH/02 The ipliteral router was not recognizing addresses of the form user@
2010 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2012 PH/03 Added disable_ipv6.
2014 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2015 decoded form, because it is most often used to construct To: headers
2016 lines in autoreplies, and the decoded form may well be syntactically
2017 invalid. However, $reply_address has leading white space removed, and all
2018 newlines turned into spaces so that the autoreply transport does not
2021 PH/05 If group was specified without a user on a router, and no group or user
2022 was specified on a transport, the group from the router was ignored.
2024 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2025 for visible compile-time settings that can be used to change these
2026 numbers, for those that want even more. Backwards compatibility with old
2027 spool files has been maintained. However, going back to a previous Exim
2028 release will lost any variables that are in spool files.
2030 PH/07 Two small changes when running in the test harness: increase delay when
2031 passing a TCP/IP connection to a new process, in case the original
2032 process has to generate a bounce, and remove special handling of
2033 127.0.0.2 (sic), which is no longer necessary.
2035 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2036 be the same on different OS.
2038 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2041 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2042 whether --show-vars was specified or not
2044 JJ/02 exipick: Added support for new ACL variable spool format introduced
2047 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2048 syntactically invalid From: or Reply-to: line, and a filter used this to
2049 generate an autoreply, and therefore failed to obtain an address for the
2050 autoreply, Exim could try to deliver to a non-existent relative file
2051 name, causing unrelated and misleading errors. What now happens is that
2052 it logs this as a hard delivery error, but does not attempt to create a
2055 PH/11 The exinext utility has a -C option for testing purposes, but although
2056 the given file was scanned by exinext itself; it wasn't being passed on
2057 when Exim was called.
2059 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2060 an end-of-file indication when reading a command response.
2062 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2063 compiled. In many other places in Exim, IPv6 addresses are always
2064 recognized, so I have changed this. It also means that IPv4 domain
2065 literals of the form [IPV4:n.n.n.n] are now always recognized.
2067 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2068 used if the router is not running as root, for example, when verifying at
2069 ACL time, or when using -bh. The debugging output from this situation was
2070 non-existent - all you got was a failure to exec. I have made two
2073 (a) Failures to set uid/gid, the current directory, or a process leader
2074 in a subprocess such as that created by queryprogram now generate
2075 suitable debugging ouput when -d is set.
2077 (b) The queryprogram router detects when it is not running as root,
2078 outputs suitable debugging information if -d is set, and then runs
2079 the subprocess without attempting to change uid/gid.
2081 PH/15 Minor change to Makefile for building test_host (undocumented testing
2084 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2085 additional section of a DNS packet that returns MX or SRV records.
2086 Instead, it always explicitly searches for A/AAAA records. This avoids
2087 major problems that occur when a DNS server includes only records of one
2088 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2089 fixed another bug: if SRV records were looked up and the corresponding
2090 address records were *not* found in the additional section, the port
2091 values from the SRV records were lost.
2093 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2094 using the correct key (the original address) when searching the retry
2095 rules in order to find which one to use for generating the retry hint.
2097 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2098 adding the default one. Similarly, if it contains a Reply-To: header, the
2099 errors_reply_to option, if set, is not used.
2101 PH/19 When calculating a retry time, Exim used to measure the "time since
2102 failure" by looking at the "first failed" field in the retry record. Now
2103 it does not use this if it is later than than the arrival time of the
2104 message. Instead it uses the arrival time. This makes for better
2105 behaviour in cases where some deliveries succeed, thus re-setting the
2106 "first failed" field. An example is a quota failure for a huge message
2107 when small messages continue to be delivered. Without this change, the
2108 "time since failure" will always be short, possible causing more frequent
2109 delivery attempts for the huge message than are intended.
2110 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2112 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2113 $1, $2, $3) because the numerical variables can be reset during some
2114 expansion items (e.g. "match"), thereby losing the authentication data.
2116 PH/21 Make -bV show the size of off_t variables so that the test suite can
2117 decide whether to run tests for quotas > 2G.
2119 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2120 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2121 filecount value is greater than 2G or if a quota value is greater than 2G
2122 on a system where the size of off_t is not greater than 4, a panic error
2125 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2126 never match. The debug and -bh output now contains an explicit error
2127 message indicating a malformed IPv4 address or mask.
2129 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2130 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2131 PH/23 above applies.
2133 PH/25 Do not write to syslog when running in the test harness. The only
2134 occasion when this arises is a failure to open the main or panic logs
2135 (for which there is an explicit test).
2137 PH/26 Added the /no_tell option to "control=freeze".
2139 PH/27 If a host name lookup failed very early in a connection, for example, if
2140 the IP address matched host_lookup and the reverse lookup yielded a name
2141 that did not have a forward lookup, an error message of the form "no IP
2142 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2143 could be logged. Now it outputs the IP address instead of "NULL".
2145 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2146 allows the sender and the authenticated sender to be set when
2147 submitting a message from within Exim. Since child_open_exim() is
2148 documented for local_scan(), the new function should be too.
2150 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2151 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2152 results in an empty string is now treated as unset.
2154 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2156 PH/31 Added sender_verify_fail logging option.
2158 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2159 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2160 tidied the source and removed it altogether.
2162 PH/33 When a queue run was abandoned because the load average was too high, a
2163 log line was always written; now it is written only if the queue_run log
2164 selector is set. In addition, the log line for abandonment now contains
2165 information about the queue run such as the pid. This is always present
2166 in "start" and "stop" lines but was omitted from the "abandon" line.
2168 PH/34 Omit spaces between a header name and the colon in the error message that
2169 is given when verify = headers_syntax fails (if there are lots of them,
2170 the message gets confusing).
2172 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2173 names, as there are now some PTR records that contain slashes. This check
2174 is only to protect against broken name servers that fall over on strange
2175 characters, so the fact that it applies to all lookups doesn't matter.
2177 PH/36 Now that the new test suite is complete, we can remove some of the
2178 special code in Exim that was needed for the old test suite. For example,
2179 sorting DNS records because real resolvers return them in an arbitrary
2180 order. The new test suite's fake resolver always returns records in the
2183 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2184 bounces) except when queue_only is set, to avoid logging races between
2185 the different processes.
2187 PH/38 Panic-die if .include specifies a non-absolute path.
2189 PH/39 A tweak to the "H" retry rule from its user.
2191 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2192 a label. They prevented compilation on older perls.
2194 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2195 a warning to be raised on newish perls.
2197 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2198 on queue. Changes to match documented behaviour of showing count of
2199 messages matching specified criteria.
2201 PH/40 Changed the default ident timeout from 30s to 5s.
2203 PH/41 Added support for the use of login_cap features, on those BSD systems
2204 that have them, for controlling the resources used by pipe deliveries.
2206 PH/42 The content-scanning code uses fopen() to create files in which to put
2207 message data. Previously it was not paying any attention to the mode of
2208 the files. Exim runs with umask(0) because the rest of the code creates
2209 files with open(), and sets the required mode explicitly. Thus, these
2210 files were ending up world-writeable. This was not a big issue, because,
2211 being within the spool directory, they were not world-accessible. I have
2212 created a function called modefopen, which takes an additional mode
2213 argument. It sets umask(777), creates the file, chmods it to the required
2214 mode, then resets the umask. All the relevant calls to fopen() in the
2215 content scanning code have been changed to use this function.
2217 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2218 to 24 hours. This avoids potential overflow problems when processing G
2219 and H retry rules. I suspect nobody ever tinkers with this value.
2221 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2223 PH/45 When the plaintext authenticator is running as a client, the server's
2224 challenges are checked to ensure they are valid base64 strings. By
2225 default, the authentication attempt is cancelled if an invalid string is
2226 received. Setting client_ignore_invalid_base64 true ignores these errors.
2227 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2228 they are received. Thus, the responses can be made to depend on the
2229 challenges. If an invalid string is ignored, an empty string is placed in
2232 PH/46 Messages that are created by the autoreply transport now contains a
2233 References: header, in accordance with RFCs 2822 and 3834.
2235 PH/47 Added authenticated_sender_force to the smtp transport.
2237 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2239 PH/49 Installed latest patch from the Sieve maintainer.
2241 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2242 was also set, the appendfile transport was unnecessarily scanning a
2243 directory of message files (e.g. for maildir delivery) to find the count
2244 of files (along with the size), even though it did not need this
2245 information. It now does the scan only if it needs to find either the
2246 size of the count of files.
2248 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2250 PH/52 Two bugs concerned with error handling when the smtp transport is
2253 (i) Exim was not creating retry information for temporary errors given
2254 for individual recipients after the DATA command when the smtp transport
2255 was used in LMTP mode. This meant that they could be retried too
2256 frequently, and not timed out correctly.
2258 (ii) Exim was setting the flag that allows error details to be returned
2259 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2260 recipients that were returned after the DATA command.
2262 PH/53 This is related to PH/52, but is more general: for any failing address,
2263 when detailed error information was permitted to be returned to the
2264 sender, but the error was temporary, then after the final timeout, only
2265 "retry timeout exceeded" was returned. Now it returns the full error as
2266 well as "retry timeout exceeded".
2268 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2269 do this, and (what is worse) MTAs that accept it.
2271 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2272 will now be deprecated.
2274 PH/56 New os.c-cygwin from the Cygwin maintainer.
2276 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2277 formats (previously only available in exim formats via -bpr, -bpru,
2278 and -bpra. Now also available in native and exiqgrep formats)
2280 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2281 with very large, slow to parse queues
2283 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2285 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2287 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2288 responses to authentication challenges, though it was showing the
2289 challenges; (ii) I've removed the CR characters from the debug output for
2292 PH/58 Allow for the insertion of a newline as well as a space when a string
2293 is turned into more than one encoded-word during RFC 2047 encoding. The
2294 Sieve code now uses this.
2296 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2297 data_4xx, lost_connection, tls_required.
2299 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2300 message was being sent as an SMTP response.
2302 PH/61 Add -l and -k options to exicyclog.
2304 PH/62 When verifying, if an address was redirected to one new address, so that
2305 verification continued, and the new address failed or deferred after
2306 having set something in $address_data, the value of $address_data was not
2307 passed back to the ACL. This was different to the case when no
2308 redirection occurred. The value is now passed back in both cases.
2310 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2311 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2312 use login_cap.h, so on its own it isn't the distinguishing feature. The
2313 new name refers directly to the setclassresources() function.
2315 PH/65 Added configuration files for NetBSD3.
2317 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2319 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2320 is preferred over IPv4.
2322 PH/68 The bounce_return_message and bounce_return_body options were not being
2323 honoured for bounces generated during the reception of non-SMTP messages.
2324 In particular, this applied to messages rejected by the ACL. This bug has
2325 been fixed. However, if bounce_return_message is true and bounce_return_
2326 body is false, the headers that are returned for a non-SMTP message
2327 include only those that have been read before the error was detected.
2328 (In the case of an ACL rejection, they have all been read.)
2330 PH/69 The HTML version of the specification is now built in a directory called
2331 spec_html instead of spec.html, because the latter looks like a path with
2332 a MIME-type, and this confuses some software.
2334 PH/70 Catch two compiler warnings in sieve.c.
2336 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2337 function verify_get_ident() calls ip_connect() to connect a socket, but
2338 if the "connect()" function timed out, ip_connect() used to close the
2339 socket. However, verify_get_ident() also closes the socket later, and in
2340 between Exim writes to the log, which may get opened at this point. When
2341 the socket was closed in ip_connect(), the log could get the same file
2342 descriptor number as the socket. This naturally causes chaos. The fix is
2343 not to close the socket in ip_connect(); the socket should be closed by
2344 the function that creates it. There was only one place in the code where
2345 this was missing, in the iplookup router, which I don't think anybody now
2346 uses, but I've fixed it anyway.
2348 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2349 well as to direct DNS lookups. Otherwise the handling of names in host
2350 lists is inconsistent and therefore confusing.
2356 PH/01 Two changes to the default runtime configuration:
2358 (1) Move the checks for relay_from_hosts and authenticated clients from
2359 after to before the (commented out) DNS black list checks.
2361 (2) Add control=submission to the relay_from_hosts and authenticated
2362 clients checks, on the grounds that messages accepted by these
2363 statements are most likely to be submissions.
2365 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2367 (1) Generate an error if the third argument for the ${prvs expansion is
2370 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2373 (3) Reset the variables that are obtained from the first argument of
2374 ${prvscheck and used in the second argument before leaving the code,
2375 because their memory is reclaimed, so using them afterwards may do
2378 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2379 one (it's much easier than Tom thought :-).
2381 (5) Because of (4), we can now allow for the use of $prvscheck_result
2382 inside the third argument.
2384 PH/03 For some reason, the default setting of PATH when running a command from
2385 a pipe transport was just "/usr/bin". I have changed it to
2388 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2389 anything to be listed in the output from -bV.
2391 PH/05 When a filter generated an autoreply, the entire To: header line was
2392 quoted in the delivery log line, like this:
2394 => >A.N.Other <ano@some.domain> <original@ddress> ...
2396 This has been changed so that it extracts the operative address. There
2397 may be more than one such address. If so, they are comma-separated, like
2400 => >ano@some.domain,ona@other.domain <original@ddress> ...
2402 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2403 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2404 not being looked up in the rDNS to get a host name, Exim was showing the
2405 IP address twice in Received: lines, even though the IP addresses were
2406 identical. For example:
2408 Received: from [1.2.3.4] (helo=[1.2.3.4])
2410 However, if the real host name was known, it was omitting the HELO data
2411 if it matched the actual IP address. This has been tidied up so that it
2412 doesn't show the same IP address twice.
2414 PH/07 When both +timestamp and +memory debugging was on, the value given by
2415 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2416 called by the debug printing, thereby overwriting the timestamp buffer.
2417 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2419 PH/08 When the original message was included in an autoreply transport, it
2420 always said "this is a copy of the message, including all the headers",
2421 even if body_only or headers_only was set. It now gives an appropriate
2424 PH/09 Applied a patch from the Sieve maintainer which:
2426 o fixes some comments
2427 o adds the (disabled) notify extension core
2428 o adds some debug output for the result of if/elsif tests
2429 o points to the current vacation draft in the documentation
2430 and documents the missing references header update
2434 o fixes a bug in processing the envelope test (when testing
2435 multiple envelope elements, the last element determinted the
2438 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2439 Electronic Mail") by including:
2441 Auto-submitted: auto-generated
2443 in the messages that it generates (bounce messages and others, such as
2444 warnings). In the case of bounce messages for non-SMTP mesages, there was
2445 also a typo: it was using "Auto_submitted" (underscore instead of
2446 hyphen). Since every message generated by Exim is necessarily in response
2447 to another message, thes have all been changed to:
2449 Auto-Submitted: auto-replied
2451 in accordance with these statements in the RFC:
2453 The auto-replied keyword:
2455 - SHOULD be used on messages sent in direct response to another
2456 message by an automatic process,
2458 - MUST NOT be used on manually-generated messages,
2460 - MAY be used on Delivery Status Notifications (DSNs) and Message
2461 Disposition Notifications (MDNs),
2463 - MUST NOT be used on messages generated by automatic or periodic
2464 processes, except for messages which are automatic responses to
2467 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2468 to the default Received: header definition.
2470 PH/12 Added log selector acl_warn_skipped (default on).
2472 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2473 variables because (a) they are in the wrong storage pool and (b) even if
2474 they were copied, it wouldn't work properly because of the caching.
2476 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2477 checking when decoding. Apparently there are clients that generate
2478 overlong encoded strings. Why am I not surprised?
2480 PH/15 If the first argument of "${if match_address" was not empty, but did not
2481 contain an "@" character, Exim crashed. Now it writes a panic log message
2482 and treats the condition as false.
2484 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2486 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2487 extension "envelope-auth". The code is finished and in agreement with
2488 other implementations, but there is no documentation so far and in fact,
2489 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2490 not changing the active code.
2492 Print executed "if" and "elsif" statements when debugging is used. This
2493 helps a great deal to understand what a filter does.
2495 Document more things not specified clearly in RFC3028. I had all this
2496 sorted out, when out of a sudden new issues came to my mind. Oops."
2498 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2501 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2502 canonical form (as documented). However, after a host name lookup from
2503 the IP address, check_host() was doing a simple string comparison with
2504 addresses acquired from the DNS when checking that the found name did
2505 have the original IP as one of its addresses. Since any found IPv6
2506 addresses are likely to be in abbreviated form, the comparison could
2507 fail. Luckily, there already exists a function for doing the comparison
2508 by converting both addresses to binary, so now that is used instead of
2509 the text comparison.
2511 PH/20 There was another similar case to PH/19, when a complete host name was
2512 given in a host list; looking up its IP address could give an abbreviated
2513 form, whereas the current host's name might or might not be abbreviated.
2514 The same fix has been applied.
2520 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2521 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2524 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2525 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2527 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2529 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2530 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2531 Also, for "Auto-Submitted", treat anything other than "no" as
2532 non-personal, in accordance with RFC 3834. (Previously it treated
2533 anything starting "auto-" as non-personal.)
2535 TF/01 The control=submission/name=... option had a problem with syntax
2536 errors if the name included a slash character. The /name= option
2537 now slurps the rest of the string, so it can include any characters
2538 but it must come last in the list of options (after /sender_retain
2541 PH/05 Some modifications to the interface to the fake nameserver for the new
2549 TK/01 Added the "success_on_redirect" address verification option. See
2550 NewStuff for rationale and an example.
2552 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2554 PH/02 Patch to exigrep to allow it to work on syslog lines.
2556 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2557 fread() to skip over the body file's header line, because in Cygwin the
2558 header line is locked and is inaccessible.
2560 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2561 co-exist for some time) to make it clear that it is the Exim ID that is
2562 referenced, not the Message-ID: header line.
2564 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2565 string_format() function, because snprintf() does not exist on all
2568 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2569 ${stat: expansion item.
2571 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2572 protocol synchronization error", to keep the pedants happy.
2574 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2575 well as for IRIX systems, when gcc is being used. See the host.c source
2578 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2580 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2583 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2584 transport and to the smtp transport in LMTP mode.
2586 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2588 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2589 run a filter in a subprocess. This could lead to confusion in subsequent
2590 lookups in the parent process. There should also be a search_tidyup() at
2591 the end of the subprocess.
2593 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2594 only if the host matched helo_try_verify_hosts, which caused the
2595 verification to occur when the EHLO/HELO command was issued. The ACL just
2596 tested the remembered result. Now, if a previous verification attempt has
2597 not happened, "verify = helo" does it there and then.
2599 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2601 TK/03 Fix log output including CR from clamd.
2603 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2604 exist provoked a memory error which could cause a segfault.
2606 PH/15 Installed PCRE 6.2
2608 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2610 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2611 of the problem. Specifically, suggested +O2 rather than +O1 for the
2614 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2616 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2617 record, the port was not correctly logged when the outgoing_port log
2618 selector was set (it logged the transort's default port).
2620 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2621 fallback_hosts, and "hosts" in the smtp transport.
2623 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2624 host errors such as "Connection refused".
2626 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2627 authentication with radiusclient 0.4.9:
2629 - Error returned from rc_read_config was caught wrongly
2630 - Username/password not passed on to radius server due to wrong length.
2632 The presumption is that some radiusclient API changes for 4.51/PH/17
2633 were not taken care of correctly. The code is still untested by me (my
2634 Linux distribution still has 0.3.2 of radiusclient), but it was
2635 contributed by a Radius user.
2637 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2638 expanding the "port" option of the smtp transport.
2640 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2641 while reading a MIME header. Thanks to Tom Hughes for a patch.
2643 PH/24 Include config.h inside local_scan.h so that configuration settings are
2646 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2647 that in an ACL for RCPT (for example), you can examine exactly what was
2650 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2651 commands, but it was not correctly comparing the address with the actual
2652 client host address. Thus, it would show the EHLO address in Received:
2653 header lines when this was not necessary.
2655 PH/27 Added the % operator to ${eval:}.
2657 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2658 it should be ignoring failures (because with -C, for example, it has lost
2659 privilege). It wasn't ignoring creation failures other than "already
2662 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2665 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2666 because an input line was too long, either on its own, or by virtue of
2667 too many continuations, the temporary file was not being removed, and the
2668 return code was incorrect.
2670 PH/31 Missing "BOOL" in function definition in filtertest.c.
2672 PH/32 Applied Sieve patches from the maintainer.
2674 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2676 PH/33 Added "verify = not_blind".
2678 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2679 Local/Makefile (with some defaults set). These are used in built scripts
2680 such as exicyclog, but they have never been used in the exim_install
2681 script (though there are many overriding facilities there). I have
2682 arranged that the exim_install script now takes note of these two
2685 PH/35 Installed configuration files for Dragonfly.
2687 PH/36 When a locally submitted message by a trusted user did not contain a
2688 From: header, and the sender address was obtained from -f or from an SMTP
2689 MAIL command, and the trusted user did not use -F to supply a sender
2690 name, $originator_name was incorrectly used when constructing a From:
2691 header. Furthermore, $originator_name was used for submission mode
2692 messages from external hosts without From: headers in a similar way,
2693 which is clearly wrong.
2695 PH/37 Added control=suppress_local_fixups.
2697 PH/38 When log_selector = +received_sender was set, and the addition of the
2698 sender made the log line's construction buffer exactly full, or one byte
2699 less than full, an overflow happened when the terminating "\n" was
2702 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2703 when the result of a list match is failure because a DNS lookup failed.
2705 PH/40 RM_COMMAND is now used in the building process.
2707 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2708 the "build-* directories that it finds.
2710 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2711 address in a domain literal was a prefix of an interface address.
2713 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2714 when verifying a sender address, unless rewrite_headers is false.
2716 PH/44 Wrote a long comment about why errors_to addresses are verified as
2717 recipients, not senders.
2719 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2720 the ratelimit ACL was added.
2722 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2724 PH/46 Added extra information about PostgreSQL errors to the error string.
2726 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2727 avoiding the need to install special zones in a real server. This is
2728 backwards compatible; if it can't find the fake resolver, it drops back.
2729 Thus, both old and new test suites can be run.
2731 TF/02 Added util/ratelimit.pl
2733 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2734 clock is set back in time.
2736 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2737 Candler <B.Candler@pobox.com>.
2739 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2740 OK for addresses that are the result of redirections.
2742 PH/48 A number of further additions for the benefit of the new test suite,
2743 including a fake gethostbyname() that interfaces to the fake DNS resolver
2746 TF/06 The fix for widen_domains has also been applied to qualify_single and
2747 search_parents which are the other dnslookup options that can cause
2750 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2753 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2755 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2756 service" if it encountered an MX record with an empty target hostname.
2757 The message is now "an MX or SRV record indicated no SMTP service".
2759 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2760 if the DNS of the sending site is misconfigured. This is quite a
2761 common situation. This change restores the behaviour of treating a
2762 helo verification defer as a failure.
2764 PH/51 If self=fail was set on a router, the bounce message did not include the
2765 actual error message.
2771 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2773 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2774 command itself ended in error, the underlying message about the transport
2775 filter timeout was being overwritten with the pipe command error. Now the
2776 underlying error message should be appended to the second error message.
2778 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2780 PH/02 Reduce the amount of output that "make" produces by default. Full output
2781 can still be requested.
2783 PH/03 The warning log line about a condition test deferring for a "warn" verb
2784 was being output only once per connection, rather than after each
2785 occurrence (because it was using the same function as for successful
2786 "warn" verbs). This seems wrong, so I have changed it.
2788 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2789 it should not have, which might have caused a crash in the right
2790 circumstances, but probably never did.
2792 PH/04 Installed a modified version of Tony Finch's patch to make submission
2793 mode fix the return path as well as the Sender: header line, and to
2794 add a /name= option so that you can make the user's friendly name appear
2797 TF/03 Added the control = fakedefer ACL modifier.
2799 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2800 Mark Lowes for thorough testing.
2802 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2804 TK/03 Merged latest SRS patch from Miles Wilton.
2806 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2807 with the definition in sysexits.h (which is #included earlier).
2808 Fortunately, Exim does not actually use EX_OK. The code used to try to
2809 preserve the sysexits.h value, by assumimg that macro definitions were
2810 scanned for macro replacements. I have been disabused of this notion,
2811 so now the code just undefines EX_OK before #including unistd.h.
2813 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2814 in the smtp transport. When a block could not be written in a single
2815 write() function, the timeout was being re-applied to each part-write.
2816 This seems wrong - if the receiver was accepting one byte at a time it
2817 would take for ever. The timeout is now adjusted when this happens. It
2818 doesn't have to be particularly precise.
2820 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2821 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2823 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2824 without a domain if the check to <postmaster@domain> fails.
2826 SC/01 Eximstats: added -xls and the ability to specify output files
2827 (patch written by Frank Heydlauf).
2829 SC/02 Eximstats: use FileHandles for outputing results.
2831 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2833 SC/04 Eximstats: fixed display of large numbers with -nvr option
2835 SC/05 Eximstats: fixed merging of reports with empty tables.
2837 SC/06 Eximstats: added the -include_original_destination flag
2839 SC/07 Eximstats: removed tabs and trailing whitespace.
2841 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2843 TK/06 MBOX spool code: Add real "From " MBOX separator line
2844 so the .eml file is really in mbox format (even though
2845 most programs do not really care). Patch from Alex Miller.
2847 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2848 The latter is generated from $received_to and is only set if the
2849 message has one envelope recipient. SA can use these headers,
2850 obviously out-of-the-box. Patch from Alex Miller.
2852 PH/08 The ${def test on a variable was returning false if the variable's
2853 value was "0", contrary to what the specification has always said!
2854 The result should be true unless the variable is empty.
2856 PH/09 The syntax error of a character other than { following "${if
2857 def:variable_name" (after optional whitespace) was not being diagnosed.
2858 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2859 accidental colon was present, for example, could give incorrect results.
2861 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2862 result is used (not including appendfile, where other changes are about
2865 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2866 This involved changing a lot of size variables from int to off_t. It
2867 should work with maildirs and everything.
2869 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2870 spamd dying while we are connected to it.
2872 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2875 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2876 function for BDB 4.3.
2878 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2880 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2881 the sources for PCRE have been re-arranged and more files are now
2884 PH/15 The code I had for printing potentially long long variables in PH/11
2885 above was not the best (it lost precision). The length of off_t variables
2886 is now inspected at build time, and an appropriate printing format (%ld
2887 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2888 to be "long long int" or "long int". This is needed for the internal
2889 formatting function string_vformat().
2891 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2892 the configuration file to be ":syslog", then the script "guesses" where
2893 the logs files are, rather than using the compiled in default. In our
2894 case the guess is not the same as the compiled default, so the script
2895 suddenly stopped working when I started to use syslog. The patch checks
2896 to see if log_file_path is "". If so, it attempts to read it from exim
2897 with no configuration file to get the compiled in version, before it
2898 falls back to the previous guessing code."
2900 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2901 implementing BATV in an Exim configuration. See NewStuff for the gory
2904 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2905 Makefile that are specific to HP-UX.
2907 PH/18 If the "use_postmaster" option was set for a recipient callout together
2908 with the "random" option, the postmaster address was used as the MAIL
2909 FROM address for the random test, but not for the subsequent recipient
2910 test. It is now used for both.
2912 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2913 patch removes a few documentation additions to RFC 3028, because the
2914 latest draft now contains them. It adds the new en;ascii-case comparator
2915 and a new error check for 8bit text in MIME parts. Comparator and
2916 require names are now matched exactly. I enabled the subaddress
2917 extension, but it is not well tested yet (read: it works for me)."
2919 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2920 rework some of the code of TK/09 above to avoid the hardwired use of
2921 "%lld" and "long long". Replaced the call to snprintf() with a call to
2924 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2925 records point to non-existent hosts", "retry timeout exceeded", and
2926 "retry time not reached for any host after a long failure period".
2928 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2929 experimental DomainKeys support:
2931 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2932 (2) On an error such as an illegally used "control", the wrong name for
2933 the control was given.
2935 These problems did NOT occur unless DomainKeys support was compiled.
2937 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2939 PH/24 Added ${if match_ip condition.
2941 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2942 databases so that it will be absolutely obvious if a crash occurs in the
2943 DB library. This is a regular occurrence (often caused by mis-matched
2946 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2947 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2948 picked up on a user's system that detects such things. There doesn't seem
2949 to be a gcc warning option for this - only an attribute that has to be
2950 put on the function's prototype. It seems that in Fedora Core 4 they have
2951 set this on a number of new functions. No doubt there will be more in due
2954 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2955 specify a transport. However, if an address that was verified by such a
2956 router was the subject of a callout, Exim crashed because it tried to
2957 read the rcpt_include_affixes from the non-existent transport. Now it
2958 just assumes that the setting of that option is false. This bug was
2959 introduced by 4.51/PH/31.
2961 PH/28 Changed -d+all to exclude +memory, because that information is very
2962 rarely of interest, but it makes the output a lot bigger. People tend to
2963 do -d+all out of habit.
2965 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2966 code in os-type was giving problems when libc.so lives in lib64, like on
2969 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2970 aren't the modern standard, and it seems that some systems' include files
2971 don't always have them. Exim was already checking for some of the newer
2972 ones like T_AAAA, and defining it itself. I've added checks for all the
2973 record types that Exim uses.
2975 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2976 not automatically generating a new one, as it is supposed to. This
2977 prevented TLS from working. If the file did exist, but contained invalid
2978 data, a new version was generated, as expected. It was only the case of a
2979 non-existent file that was broken.
2981 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2982 with a change in libdomainkeys > 0.64.
2984 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2985 from DNS. If the selector record carries the flag, it now has
2986 precedence over the domain-wide flag.
2988 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2990 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2991 the use of a port name that isn't defined in /etc/services) occurred, the
2992 message was deferred as in a normal delivery, and thus remained on the
2993 spool, instead of being failed because of the mua_wrapper setting. This
2994 is now fixed, and I tidied up some of the mua_wrapper messages at the
2997 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2998 the messages in a hash of arrays rather than using individual hashes.
2999 This is a bit cleaner and results in dramatic memory savings, albeit
3000 at a slight CPU cost.
3002 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3003 as requested by Marc Sherman.
3005 SC/10 Eximstats: added histograms for user specified patterns as requested
3008 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3010 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3011 fopen() in the content-scanning modules that did not already have it.
3017 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3018 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3020 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3022 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3024 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3025 to test Sieve filters that use "vacation".
3027 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3028 that changes the way the GnuTLS parameters are stored in the cache file.
3029 The new format can be generated externally. For backward compatibility,
3030 if the data in the cache doesn't make sense, Exim assumes it has read an
3031 old-format file, and it generates new data and writes a new file. This
3032 means that you can't go back to an older release without removing the
3035 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3036 work if there are any delivery delays because "one_time" forces the
3037 parent to be marked "delivered", so its unseen clone is never tried
3038 again. For this reason, Exim now forbids the simultaneous setting of
3041 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3042 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3043 entry if you want to know the details. The fix, however, neglected to
3044 consider the case when local delivery batching is involved. The test for
3045 "previously delivered" was not happening when checking to see if an
3046 address could be batched with a previous (undelivered) one; under
3047 certain circumstances this could lead to multiple deliveries to the same
3050 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3051 in its include files, and this causes problems building Exim.
3053 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3054 header_syntax) but Exim was just ignoring anything given after a slash.
3055 In particular, this caused confusion with an attempt to use "verify =
3056 reverse_host_lookup/defer_ok". An error is now given when options are
3057 supplied for verify items that do not have them. (Maybe reverse_host_
3058 lookup should have a defer_ok option, but that's a different point.)
3060 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3061 defined by RFC 821) to 2048, because there were problems with some AUTH
3062 commands, and RFC 1869 says the size should be increased for extended
3063 SMTP commands that take arguments.
3065 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3068 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3069 "unknown" error; now it says that the functionality isn't in the binary.
3071 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3072 an address' error message when a string expansion fails (syntax or
3073 whatever). Otherwise the password may appear in the log. Following change
3074 PH/42 below, there is no longer a chance of it appearing in a bounce
3077 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3079 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3080 of its addresses was ever tried. (Bugzilla bug #2.)
3082 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3083 the result incorrectly in the debug output. (It correctly added a newline
3084 to what was transported.)
3086 TF/01 Added $received_time.
3088 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3089 commented out examples of how to interface to a virus scanner and to
3090 SpamAssassin. Also added commented examples of av_scanner and
3091 spamd_address settings.
3093 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3094 and controls are allowed in which ACLs. There were a couple of minor
3095 errors. Some of the entries in the conditions table (which is a table of
3096 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3097 negation of where the condition IS allowed.
3099 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3101 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3102 header file does not have a version number, so I've had to invent a new
3103 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3104 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3105 radiusclient), but it was contributed by a Radius user.
3107 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3108 files or directories, for OpenSSL.
3110 PH/19 When an Exim process that is running as root has to create an Exim log
3111 file, it does so in a subprocess that runs as exim:exim so as to get the
3112 ownership right at creation (otherwise, other Exim processes might see
3113 the file with the wrong ownership). There was no test for failure of this
3114 fork() call, which would lead to the process getting stuck as it waited
3115 for a non-existent subprocess. Forks do occasionally fail when resources
3116 run out. I reviewed all the other calls to fork(); they all seem to check
3119 PH/20 When checking for unexpected SMTP input at connect time (before writing
3120 the banner), Exim was not dealing correctly with a non-positive return
3121 from the read() function. If the client had disconnected by this time,
3122 the result was a log entry for a synchronization error with an empty
3123 string after "input=" when read() returned zero. If read() returned -1
3124 (an event I could not check), uninitialized data bytes were printed.
3125 There were reports of junk text (parts of files, etc) appearing after
3128 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3130 PH/22 Added support for macro redefinition, and (re)definition in between
3131 driver and ACL definitions.
3133 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3134 forgetting to use the resulting value; it was using the unexpanded value.
3136 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3137 hadn't been configured. The fix is from Juergen Kreileder, who
3138 understands it better than I do:
3140 "Here's what I see happening with three configured cyrus_sasl
3141 authenticators configured (plain, login, cram-md5):
3143 On startup auth_cyrus_sasl_init() gets called for each of these.
3144 This means three calls to sasl_listmech() without a specified mech_list.
3145 => SASL tests which mechs of all available mechs actually work
3146 => three warnings about OTP not working
3147 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3149 With the patch, sasl_listmech() also gets called three times. But now
3150 SASL's mech_list option is set to the server_mech specified in the the
3151 authenticator. Or in other words, the answer from sasl_listmech()
3152 gets limited to just the mech you're testing for (which is different
3154 => the return list contains just 'plain' or 'login', 'cram-md5' or
3155 nothing depending on the value of ob->server_mech.
3157 I've just tested the patch: Authentication still works fine,
3158 unavailable mechs specified in the exim configuration are still
3159 caught, and the auth.log warnings about OTP are gone."
3161 PH/25 When debugging is enabled, the contents of the command line are added
3162 to the debugging output, even when log_selector=+arguments is not
3165 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3166 answer is "GNU", and only if the return is "GNU/something" is the answer
3169 PH/27 $acl_verify_message is now set immediately after the failure of a
3170 verification in an ACL, and so is available in subsequent modifiers. In
3171 particular, the message can be preserved by coding like this:
3173 warn !verify = sender
3174 set acl_m0 = $acl_verify_message
3176 Previously, $acl_verify_message was set only while expanding "message"
3177 and "log_message" when a very denied access.
3179 PH/28 Modified OS/os.c-Linux with
3181 -#ifndef OS_LOAD_AVERAGE
3182 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3184 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3185 nomenclature these days.)
3187 PH/29 Installed patch from the Sieve maintainer that adds the options
3188 sieve_useraddress and sieve_subaddress to the redirect router.
3190 PH/30 In these circumstances:
3191 . Two addresses routed to the same list of hosts;
3192 . First host does not offer TLS;
3193 . First host accepts first address;
3194 . First host gives temporary error to second address;
3195 . Second host offers TLS and a TLS session is established;
3196 . Second host accepts second address.
3197 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3198 and peerdn, if requested) that were in fact used only for the second
3201 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3202 attention to any local part prefix or suffix that was matched by the
3203 router that accepted the address. It now behaves in the same way as it
3204 does for delivery: the affixes are removed from the local part unless
3205 rcpt_include_affixes is set on the transport.
3207 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3208 timeout during the DATA phase of an incoming message.
3210 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3211 applied a patch sanctioned by the Sieve maintainer.
3213 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3214 the uid or gid is negative. A case of a negative gid caused this to be
3215 noticed. The fix allows for either to be negative.
3217 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3218 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3221 PH/36 The change PH/12 above was broken. Fixed it.
3223 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3224 the grounds that routing the same address twice would always produce the
3225 same answer. This might have been true once, but it is certainly no
3226 longer true now. Routing a child address may depend on the previous
3227 routing that produced that child. Some complicated redirection strategies
3228 went wrong when messages had multiple recipients, and made Exim's
3229 behaviour dependent on the order in which the addresses were given.
3231 I have moved the duplicate checking until after the routing is complete.
3232 Exim scans the addresses that are assigned to local and remote
3233 transports, and removes any duplicates. This means that more work will be
3234 done, as duplicates will always all be routed, but duplicates are
3235 presumably rare, so I don't expect this is of any significance.
3237 For deliveries to pipes, files, and autoreplies, the duplicate checking
3238 still happens during the routing process, since they are not going to be
3241 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3242 It corrects a timeout issue with spamd. This is Ian's comment: "The
3243 background is that sometimes spamd either never reads data from a
3244 connection it has accepted, or it never writes response data. The exiscan
3245 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3246 blindly assumes that writes won't block so it may never time out."
3248 PH/39 Allow G after quota size as well as K and M.
3250 PH/40 The value set for $authenticated_id in an authenticator may not contain
3251 binary zeroes or newlines because the value is written to log lines and
3252 to spool files. There was no check on this. Now the value is run through
3253 the string_printing() function so that such characters are converted to
3254 printable escape sequences.
3256 PH/41 $message_linecount is a new variable that contains the total number of
3257 lines in the message. Compare $body_linecount, which is the count for the
3260 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3261 bounce and delay warning messages, except in certain special cases, which
3264 (a) An SMTP error message from a remote host;
3265 (b) A message specified in a :fail: redirection;
3266 (c) A message specified in a "fail" command in a system filter;
3267 (d) A message specified in a FAIL return from the queryprogram router;
3268 (e) A message specified by the cannot_route_message router option.
3270 In these cases only, Exim does include the error details in bounce and
3271 warning messages. There are also a few cases where bland messages such
3272 as "unrouteable address" or "local delivery error" are given.
3274 PH/43 $value is now also set for the "else" part of a ${run expansion.
3276 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3277 being worked on, but at least Exim now implements the latest version to
3280 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3281 process to complete was treated as a delivery failure, a timeout while
3282 writing the message to the pipe was logged, but erroneously treated as a
3283 successful delivery. Such timeouts include transport filter timeouts. For
3284 consistency with the overall process timeout, these timeouts are now
3285 treated as errors, giving rise to delivery failures by default. However,
3286 there is now a new Boolean option for the pipe transport called
3287 timeout_defer, which, if set TRUE, converts the failures into defers for
3288 both kinds of timeout. A transport filter timeout is now identified in
3291 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3292 systems where "make" and "gmake" are different, calling "gmake" at top
3293 level broke things. I've arranged for the value of $(MAKE) to be passed
3294 from the Makefile to this script so that it can call the same version of
3298 A note about Exim versions 4.44 and 4.50
3299 ----------------------------------------
3301 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3302 changes of various kinds. As a consequence, a big documentation update was
3303 needed. This delayed the release for rather longer than seemed good, especially
3304 in the light of a couple of (minor) security issues. Therefore, the changes
3305 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3306 release. So 4.44 and 4.50 are in effect two different branches that both start
3309 I have left the 4.50 change log unchanged; it contains all the changes since
3310 4.43. The change log for 4.44 is below; many of its items are identical to
3311 those for 4.50. This seems to be the most sensible way to preserve the
3312 historical information.
3318 1. Minor wording change to the doc/README.SIEVE file.
3320 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3321 computation of the current number of files was incorrect.
3323 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3324 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3327 4. Give more explanation in the error message when the command for a transport
3328 filter fails to execute.
3330 5. There are several places where Exim runs a non-Exim command in a
3331 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3332 was being done only for the command run by the queryprogram router. It is
3333 now done for all such subprocesses. The other cases are: ${run, transport
3334 filters, and the commands run by the lmtp and pipe transports.
3336 6. Added CONFIGURE_GROUP build-time option.
3338 7. Some older OS have a limit of 256 on the maximum number of file
3339 descriptors. Exim was using setrlimit() to set 1000 as a large value
3340 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3341 systems. I've change it so that if it can't get 1000, it tries for 256.
3343 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3344 was an oversight, and furthermore, ever since the addition of extra
3345 controls (e.g. 4.43/32), the checks on when to allow different forms of
3346 "control" were broken. There should now be diagnostics for all cases when a
3347 control that does not make sense is encountered.
3349 9. Added the /retain_sender option to "control=submission".
3351 10. $recipients is now available in the predata ACL (oversight).
3353 11. Tidy the search cache before the fork to do a delivery from a message
3354 received from the command line. Otherwise the child will trigger a lookup
3355 failure and thereby defer the delivery if it tries to use (for example) a
3356 cached ldap connection that the parent has called unbind on.
3358 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3359 of $address_data from the recipient verification was clobbered by the
3360 sender verification.
3362 13. The value of address_data from a sender verification is now available in
3363 $sender_address_data in subsequent conditions in the ACL statement.
3365 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3367 15. Added a new option "connect=<time>" to callout options, to set a different
3370 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3371 was its contents. (It was OK if the option was not defined at all.)
3373 17. A "Completed" log line is now written for messages that are removed from
3374 the spool by the -Mrm option.
3376 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3377 information about exactly what failed.
3379 19. Added -dd to debug only the daemon process.
3381 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3382 handles timeouts, both on the server side and network timeouts. Renamed the
3383 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3385 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3386 It is now set to "smtps".
3388 22. $host_address is now set to the target address during the checking of
3389 ignore_target_hosts.
3391 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3392 being passed; this would have caused $sender_host_name to have been used if
3393 matching the list had actually called for a host name (not very likely,
3394 since this list is usually IP addresses). A host name is now passed as
3397 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3398 code that specifies a non-restarting handler (typically sigaction() in
3399 modern systems) in an attempt to fix a rare and obscure crash bug.
3401 25. Narrowed the window for a race in the daemon that could cause it to ignore
3402 SIGCHLD signals. This is not a major problem, because they are used only to
3403 wake it up if nothing else does.
3405 26. A malformed maildirsize file could cause Exim to calculate negative values
3406 for the mailbox size or file count. Odd effects could occur as a result.
3407 The maildirsize information is now recalculated if the size or filecount
3410 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3411 support for a long time. Removed HAVE_SYS_VFS_H.
3413 28. Installed the latest version of exipick from John Jetmore.
3415 29. In an address list, if the pattern was not a regular expression, an empty
3416 subject address (from a bounce message) matched only if the pattern was an
3417 empty string. Non-empty patterns were not even tested. This was the wrong
3418 because it is perfectly reasonable to use an empty address as part of a
3419 database query. An empty address is now tested by patterns that are
3420 lookups. However, all the other forms of pattern expect the subject to
3421 contain a local part and a domain, and therefore, for them, an empty
3422 address still always fails if the pattern is not itself empty.
3424 30. Exim went into a mad DNS loop when attempting to do a callout where the
3425 host was specified on an smtp transport, and looking it up yielded more
3426 than one IP address.
3428 31. Re-factored the code for checking spool and log partition space into a
3429 function that finds that data and another that does the check. The former
3430 is then used to implement four new variables: $spool_space, $log_space,
3431 $spool_inodes, and $log_inodes.
3433 32. The RFC2047 encoding function was originally intended for short strings
3434 such as real names; it was not keeping to the 75-character limit for
3435 encoded words that the RFC imposes. It now respects the limit, and
3436 generates multiple encoded words if necessary. To be on the safe side, I
3437 have increased the buffer size for the ${rfc2047: expansion operator from
3440 33. It is now permitted to omit both strings after an "if" condition; if the
3441 condition is true, the result is "true". As before, when the second string
3442 is omitted, a false condition yields an empty string. This makes it less
3443 cumbersome to write custom ACL and router conditions.
3445 34. Failure to deliver a bounce message always caused it to be frozen, even if
3446 there was an errors_to setting on the router. The errors_to setting is now
3449 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3450 canonical form (fully expanded) before being placed in
3451 $sender_host_address.
3453 36. The table in the code that translates DNS record types into text (T_A to
3454 "A" for instance) was missing entries for NS and CNAME. It is just possible
3455 that this could have caused confusion if both these types were looked up
3456 for the same domain, because the text type is used as part of Exim's
3457 per-process caching. But the chance of anyone hitting this buglet seems
3460 37. The dnsdb lookup has been extended in a number of ways.
3462 (1) There is a new type, "zns", which walks up the domain tree until it
3463 finds some nameserver records. It should be used with care.
3465 (2) There is a new type, "mxh", which is like "mx" except that it returns
3466 just the host names, not the priorities.
3468 (3) It is now possible to give a list of domains (or IP addresses) to be
3469 looked up. The behaviour when one of the lookups defers can be
3470 controlled by a keyword.
3472 (4) It is now possible to specify the separator character for use when
3473 multiple records are returned.
3475 38. The dnslists ACL condition has been extended: it is now possible to supply
3476 a list of IP addresses and/or domains to be looked up in a particular DNS
3479 39. Added log_selector=+queue_time_overall.
3481 40. When running the queue in the test harness, wait just a tad after forking a
3482 delivery process, to get repeatability of debugging output.
3484 41. Include certificate and key file names in error message when GnuTLS fails
3485 to set them up, because the GnuTLS error message doesn't include the name
3486 of the failing file when there is a problem reading it.
3488 42. Allow both -bf and -bF in the same test run.
3490 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3492 44. The "Exiscan patch" is now merged into the mainline Exim source.
3494 45. Sometimes the final signoff response after QUIT could fail to get
3495 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3496 < 0 before doing a fflush(). This bug looks as though it goes back to the
3497 introduction of TLS in release 3.20, but "sometimes" must have been rare
3498 because the tests only now provoked it.
3500 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3501 (this can affect the format of dates).
3503 47. exim_tidydb, when checking for the continued existence of a message for
3504 which it has found a message-specific retry record, was not finding
3505 messages that were in split spool directories. Consequently, it was
3506 deleting retry records that should have stayed in existence.
3508 48. Steve fixed some bugs in eximstats.
3510 49. The SPA authentication driver was not abandoning authentication and moving
3511 on to the next authenticator when an expansion was forced to fail,
3512 contradicting the general specification for all authenticators. Instead it
3513 was generating a temporary error. It now behaves as specified.
3515 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3516 (the order specifies the preference for clients). The order is now AES256,
3517 AES128, 3DES, ARCFOUR128.
3519 51. Small patch to Sieve code - explicitly set From: when generating an
3522 52. Exim crashed if a remote delivery caused a very long error message to be
3523 recorded - for instance if somebody sent an entire SpamAssassin report back
3524 as a large number of 550 error lines. This bug was coincidentally fixed by
3525 increasing the size of one of Exim's internal buffers (big_buffer) that
3526 happened as part of the Exiscan merge. However, to be on the safe side, I
3527 have made the code more robust (and fixed the comments that describe what
3530 53. Now that there can be additional text after "Completed" in log lines (if
3531 the queue_time_overall log selector is set), a one-byte patch to exigrep
3532 was needed to allow it to recognize "Completed" as not the last thing in
3535 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3536 patch that reportedly fixes this has been added. I am not expert enough to
3537 create a test for it. This is what the patch creator wrote:
3539 "I found a little strange behaviour of ldap code when working with
3540 Windows 2003 AD Domain, where users was placed in more than one
3541 Organization Units. When I tried to give exim partial DN, the exit code
3542 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3543 But simultaneously result of request was absolutely normal ldap result,
3544 so I produce this patch..."
3546 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3547 REFERENCE, so I have modified the code to exclude the patch when that macro
3550 55. Some experimental protocols are using DNS PTR records for new purposes. The
3551 keys for these records are domain names, not reversed IP addresses. The
3552 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3553 leaves it alone. Component reversal etc. now happens only for IP addresses.
3556 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3558 57. Double the size of the debug message buffer (to 2048) so that more of very
3559 long debug lines gets shown.
3561 58. The exicyclog utility now does better if the number of log files to keep
3562 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3564 59. Two changes related to the smtp_active_hostname option:
3566 (1) $smtp_active_hostname is now available as a variable.
3567 (2) The default for smtp_banner uses $smtp_active_hostname instead
3568 of $primary_hostname.
3570 60. The host_aton() function is supposed to be passed a string that is known
3571 to be a valid IP address. However, in the case of IPv6 addresses, it was
3572 not checking this. This is a hostage to fortune. Exim now panics and dies
3573 if the condition is not met. A case was found where this could be provoked
3574 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3575 components; fortuitously, this particular loophole had already been fixed
3576 by change 4.50/55 above.
3578 If there are any other similar loopholes, the new check in host_aton()
3579 itself should stop them being exploited. The report I received stated that
3580 data on the command line could provoke the exploit when Exim was running as
3581 exim, but did not say which command line option was involved. All I could
3582 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3583 running as the user.
3586 61. There was a buffer overflow vulnerability in the SPA authentication code
3587 (which came originally from the Samba project). I have added a test to the
3588 spa_base64_to_bits() function which I hope fixes it.
3591 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3592 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3594 63. The daemon start-up calls getloadavg() while still root for those OS that
3595 need the first call to be done as root, but it missed one case: when
3596 deliver_queue_load_max is set with deliver_drop_privilege. This is
3597 necessary for the benefit of the queue runner, because there is no re-exec
3598 when deliver_drop_privilege is set.
3600 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3601 This has been fixed.
3603 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3604 was in use, was not putting the data itself into the right store pool;
3605 consequently, it could be overwritten for a subsequent message in the same
3606 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3609 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3611 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3612 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3613 the function were treating the return as a boolean value, which happened to
3614 work because 0=false and not-0=true, but is not correct code.
3616 68. The host_aton() function was not handling scoped IPv6 addresses (those
3617 with, for example, "%eth0" on the end) correctly.
3619 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3620 negated items (that is, ~something) in unsigned ints. Some compilers
3621 apparently mutter when there is no cast.
3623 70. If an address verification called from an ACL failed, and did not produce a
3624 user-specific message (i.e. there was only a "system" message), nothing was
3625 put in $acl_verify_message. In this situation, it now puts the system
3628 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3629 session; change 4.31/43 added the unwanted input to the log line - except
3630 that it did not do this in the start of session case. It now does.
3632 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3633 This is wrong and can cause the other end to generate a synchronization
3634 error if it is another Exim or anything else that does the synchronization
3635 check. A QUIT command is no longer sent after a timeout.
3637 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3638 during host lookups.
3640 74. The defer_ok option of callout verification was not working if it was used
3641 when verifying addresses in header lines, that is, for this case:
3643 verify = header_sender/callout=defer_ok
3645 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3646 those file descriptors could be used for SMTP connections. If anything
3647 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3648 could be sent to the SMTP client, causing chaos. The daemon now opens
3649 stdin, stdout, and stderr to /dev/null when it puts itself into the
3652 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3653 log by default. The user can override this with suitable Perl magic.
3655 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3656 the log message when discard triggers, was not working for the DATA ACL or
3657 for the non-SMTP ACL.
3659 78. Error message wording change in sieve.c.
3661 79. If smtp_accept_max_per_host was set, the number of connections could be
3662 restricted to fewer than expected, because the daemon was trying to set up
3663 a new connection before checking whether the processes handling previous
3664 connections had finished. The check for completed processes is now done
3665 earlier. On busy systems, this bug wouldn't be noticed because something
3666 else would have woken the daemon, and it would have reaped the completed
3669 80. If a message was submitted locally by a user whose login name contained one
3670 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3671 It caused a spool format error. I have fixed the spool reading code. A
3672 related problem was that the "from" clause in the Received: line became
3673 illegal because of the space(s). It is now covered by ${quote_local_part.
3675 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3678 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3680 83. Patch from Sieve maintainer to fix unterminated string problem in
3681 "vacation" handling.
3683 84. Some minor changes to the Linux configuration files to help with other
3684 OS variants using glibc.
3686 85. One more patch for Sieve to update vacation handling to latest spec.
3689 ----------------------------------------------------
3690 See the note above about the 4.44 and 4.50 releases.
3691 ----------------------------------------------------
3697 1. Change 4.43/35 introduced a bug that caused file counts to be
3698 incorrectly computed when quota_filecount was set in an appendfile
3701 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3702 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3705 3. Give more explanation in the error message when the command for a transport
3706 filter fails to execute.
3708 4. There are several places where Exim runs a non-Exim command in a
3709 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3710 was being done only for the command run by the queryprogram router. It is
3711 now done for all such subprocesses. The other cases are: ${run, transport
3712 filters, and the commands run by the lmtp and pipe transports.
3714 5. Some older OS have a limit of 256 on the maximum number of file
3715 descriptors. Exim was using setrlimit() to set 1000 as a large value
3716 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3717 systems. I've change it so that if it can't get 1000, it tries for 256.
3719 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3720 was an oversight, and furthermore, ever since the addition of extra
3721 controls (e.g. 4.43/32), the checks on when to allow different forms of
3722 "control" were broken. There should now be diagnostics for all cases when a
3723 control that does not make sense is encountered.
3725 7. $recipients is now available in the predata ACL (oversight).
3727 8. Tidy the search cache before the fork to do a delivery from a message
3728 received from the command line. Otherwise the child will trigger a lookup
3729 failure and thereby defer the delivery if it tries to use (for example) a
3730 cached ldap connection that the parent has called unbind on.
3732 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3733 of $address_data from the recipient verification was clobbered by the
3734 sender verification.
3736 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3737 was its contents. (It was OK if the option was not defined at all.)
3739 11. A "Completed" log line is now written for messages that are removed from
3740 the spool by the -Mrm option.
3742 12. $host_address is now set to the target address during the checking of
3743 ignore_target_hosts.
3745 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3746 being passed; this would have caused $sender_host_name to have been used if
3747 matching the list had actually called for a host name (not very likely,
3748 since this list is usually IP addresses). A host name is now passed as
3751 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3752 code that specifies a non-restarting handler (typically sigaction() in
3753 modern systems) in an attempt to fix a rare and obscure crash bug.
3755 15. Narrowed the window for a race in the daemon that could cause it to ignore
3756 SIGCHLD signals. This is not a major problem, because they are used only to
3757 wake it up if nothing else does.
3759 16. A malformed maildirsize file could cause Exim to calculate negative values
3760 for the mailbox size or file count. Odd effects could occur as a result.
3761 The maildirsize information is now recalculated if the size or filecount
3764 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3765 support for a long time. Removed HAVE_SYS_VFS_H.
3767 18. Updated exipick to current release from John Jetmore.
3769 19. Allow an empty sender to be matched against a lookup in an address list.
3770 Previously the only cases considered were a regular expression, or an
3773 20. Exim went into a mad DNS lookup loop when doing a callout where the
3774 host was specified on the transport, if the DNS lookup yielded more than
3777 21. The RFC2047 encoding function was originally intended for short strings
3778 such as real names; it was not keeping to the 75-character limit for
3779 encoded words that the RFC imposes. It now respects the limit, and
3780 generates multiple encoded words if necessary. To be on the safe side, I
3781 have increased the buffer size for the ${rfc2047: expansion operator from
3784 22. Failure to deliver a bounce message always caused it to be frozen, even if
3785 there was an errors_to setting on the router. The errors_to setting is now
3788 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3789 canonical form (fully expanded) before being placed in
3790 $sender_host_address.
3792 24. Updated eximstats to version 1.33
3794 25. Include certificate and key file names in error message when GnuTLS fails
3795 to set them up, because the GnuTLS error message doesn't include the name
3796 of the failing file when there is a problem reading it.
3798 26. Expand error message when OpenSSL has problems setting up cert/key files.
3801 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3802 (this can affect the format of dates).
3804 28. exim_tidydb, when checking for the continued existence of a message for
3805 which it has found a message-specific retry record, was not finding
3806 messages that were in split spool directories. Consequently, it was
3807 deleting retry records that should have stayed in existence.
3809 29. eximstats updated to version 1.35
3810 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3811 1.35 - bugfix such that pie charts by volume are generated correctly
3813 30. The SPA authentication driver was not abandoning authentication and moving
3814 on to the next authenticator when an expansion was forced to fail,
3815 contradicting the general specification for all authenticators. Instead it
3816 was generating a temporary error. It now behaves as specified.
3818 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3819 (the order specifies the preference for clients). The order is now AES256,
3820 AES128, 3DES, ARCFOUR128.
3822 31. Small patch to Sieve code - explicitly set From: when generating an
3825 32. Exim crashed if a remote delivery caused a very long error message to be
3826 recorded - for instance if somebody sent an entire SpamAssassin report back
3827 as a large number of 550 error lines. This bug was coincidentally fixed by
3828 increasing the size of one of Exim's internal buffers (big_buffer) that
3829 happened as part of the Exiscan merge. However, to be on the safe side, I
3830 have made the code more robust (and fixed the comments that describe what
3833 33. Some experimental protocols are using DNS PTR records for new purposes. The
3834 keys for these records are domain names, not reversed IP addresses. The
3835 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3836 leaves it alone. Component reversal etc. now happens only for IP addresses.
3839 34. The host_aton() function is supposed to be passed a string that is known
3840 to be a valid IP address. However, in the case of IPv6 addresses, it was
3841 not checking this. This is a hostage to fortune. Exim now panics and dies
3842 if the condition is not met. A case was found where this could be provoked
3843 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3844 components; fortuitously, this particular loophole had already been fixed
3845 by change 4.50/55 or 4.44/33 above.
3847 If there are any other similar loopholes, the new check in host_aton()
3848 itself should stop them being exploited. The report I received stated that
3849 data on the command line could provoke the exploit when Exim was running as
3850 exim, but did not say which command line option was involved. All I could
3851 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3852 running as the user.
3855 35. There was a buffer overflow vulnerability in the SPA authentication code
3856 (which came originally from the Samba project). I have added a test to the
3857 spa_base64_to_bits() function which I hope fixes it.
3860 36. The daemon start-up calls getloadavg() while still root for those OS that
3861 need the first call to be done as root, but it missed one case: when
3862 deliver_queue_load_max is set with deliver_drop_privilege. This is
3863 necessary for the benefit of the queue runner, because there is no re-exec
3864 when deliver_drop_privilege is set.
3866 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3867 was in use, was not putting the data itself into the right store pool;
3868 consequently, it could be overwritten for a subsequent message in the same
3869 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3872 38. Sometimes the final signoff response after QUIT could fail to get
3873 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3874 < 0 before doing a fflush(). This bug looks as though it goes back to the
3875 introduction of TLS in release 3.20, but "sometimes" must have been rare
3876 because the tests only now provoked it.
3882 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3883 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3884 Now it returns an integer. A number of calls were still expecting a T/F
3885 return. Fortuitously, in all cases, the tests worked in OK situations,
3886 which is the norm. However, things would have gone wrong on any write
3887 failures on the smtp file descriptor. This function is used when sending
3888 messages over SMTP and also when doing verify callouts.
3890 2. When Exim is called to do synchronous delivery of a locally submitted
3891 message (the -odf or -odi options), it no longer closes stderr before doing
3894 3. Implemented the mua_wrapper option.
3896 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3898 5. Implemented the functions header_remove(), header_testname(),
3899 header_add_at_position(), and receive_remove_recipient(), and exported them
3902 6. If an ACL "warn" statement specified the addition of headers, Exim already
3903 inserted X-ACL-Warn: at the start if there was no header name. However, it
3904 was not making this test for the second and subsequent header lines if
3905 there were newlines in the string. This meant that an invalid header could
3906 be inserted if Exim was badly configured.
3908 7. Allow an ACL "warn" statement to add header lines at the start or after all
3909 the Received: headers, as well as at the end.
3911 8. Added the rcpt_4xx retry error code.
3913 9. Added postmaster_mailfrom=xxx to callout verification option.
3915 10. Added mailfrom=xxxx to the callout verification option, for verify=
3918 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3919 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3921 12. Inserted some casts to stop certain compilers warning when using pointer
3922 differences as field lengths or precisions in printf-type calls (mostly
3923 affecting debugging statements).
3925 13. Added optional readline() support for -be (dynamically loaded).
3927 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3928 same clock tick as a message's arrival, so that its received time was the
3929 same as the "first fail" time on the retry record, and that message
3930 remained on the queue past the ultimate address timeout, every queue runner
3931 would try a delivery (because it was past the ultimate address timeout) but
3932 after another failure, the ultimate address timeout, which should have then
3933 bounced the address, did not kick in. This was a "< instead of <=" error;
3934 in most cases the first failure would have been in the next clock tick
3935 after the received time, and all would be well.
3937 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3938 being recognized when the domain list was tested by the match_domain
3939 condition in an expansion string.
3941 16. Added the ${str2b64: operator.
3943 17. Exim was always calling setrlimit() to set a large limit for the number of
3944 processes, without checking whether the existing limit was already
3945 adequate. (It did check for the limit on file descriptors.) Furthermore,
3946 errors from getrlimit() and setrlimit() were being ignored. Now they are
3947 logged to the main and panic logs, but Exim does carry on, to try to do its
3948 job under whatever limits there are.
3950 18. Imported PCRE 5.0.
3952 19. Trivial typo in log message " temporarily refused connection" (the leading
3955 20. If the log selector return_path_on_delivery was set and an address was
3956 redirected to /dev/null, the delivery process crashed because it assumed
3957 that a return path would always be set for a "successful" delivery. In this
3958 case, the whole delivery is bypassed as an optimization, and therefore no
3961 21. Internal re-arrangement: the function for sending a challenge and reading
3962 a response while authentication was assuming a zero-terminated challenge
3963 string. It's now changed to take a pointer and a length, to allow for
3964 binary data in such strings.
3966 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3968 23. Exim was not respecting finduser_retries when seeking the login of the
3969 uid under which it was called; it was always trying 10 times. (The default
3970 setting of finduser_retries is zero.) Also, it was sleeping after the final
3971 failure, which is pointless.
3973 24. Implemented tls_on_connect_ports.
3975 25. Implemented acl_smtp_predata.
3977 26. If the domain in control=submission is set empty, Exim assumes that the
3978 authenticated id is a complete email address when it generates From: or
3979 Sender: header lines.
3981 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3982 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3983 chown and chgrp in /bin and hostname in /usr/bin.
3985 28. Exim was keeping the "process log" file open after each use, just as it
3986 does for the main log. This opens the possibility of it remaining open for
3987 long periods when the USR1 signal hits a daemon. Occasional processlog
3988 errors were reported, that could have been caused by this. Anyway, it seems
3989 much more sensible not to leave this file open at all, so that is what now
3992 29. The long-running daemon process does not normally write to the log once it
3993 has entered its main loop, and it closes the log before doing so. This is
3994 so that log files can straightforwardly be renamed and moved. However,
3995 there are a couple of unusual error situations where the daemon does write
3996 log entries, and I had neglected to close the log afterwards.
3998 30. The text of an SMTP error response that was received during a remote
3999 delivery was being truncated at 512 bytes. This is too short for some of
4000 the long messages that one sometimes sees. I've increased the limit to
4003 31. It is now possible to make retry rules that apply only when a message has a
4004 specific sender, in particular, an empty sender.
4006 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4007 it possible to be selective about when SMTP synchronization is enforced.
4009 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4011 32. Implemented hosts_connection_nolog.
4013 33. Added an ACL for QUIT.
4015 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4018 35. Added mailbox_size and mailbox_filecount to appendfile.
4020 36. Added control = no_multiline_responses to ACLs.
4022 37. There was a bug in the logic of the code that waits for the clock to tick
4023 in the case where the clock went backwards by a substantial amount such
4024 that the microsecond fraction of "now" was more than the microsecond
4025 fraction of "then" (but the whole seconds number was less).
4027 38. Added support for the libradius Radius client library this is found on
4028 FreeBSD (previously only the radiusclient library was supported).
4034 1. When certain lookups returned multiple values in the form name=value, the
4035 quoting of the values was not always being done properly. Specifically:
4036 (a) If the value started with a double quote, but contained no whitespace,
4038 (b) If the value contained whitespace other than a space character (i.e.
4039 tabs or newlines or carriage returns) it was not quoted.
4040 This fix has been applied to the mysql and pgsql lookups by writing a
4041 separate quoting function and calling it from the lookup code. The fix
4042 should probably also be applied to nisplus, ibase and oracle lookups, but
4043 since I cannot test any of those, I have not disturbed their existing code.
4045 2. A hit in the callout cache for a specific address caused a log line with no
4046 reason for rejecting RCPT. Now it says "Previous (cached) callout
4047 verification failure".
4049 3. There was an off-by-one bug in the queryprogram router. An over-long
4050 return line was truncated at 256 instead of 255 characters, thereby
4051 overflowing its buffer with the terminating zero. As well as fixing this, I
4052 have increased the buffer size to 1024 (and made a note to document this).
4054 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4055 when Exim is waiting for an SMTP response from a remote server, Exim
4056 restarts its select() call on the socket, thereby resetting its timeout.
4057 This is not a problem when such interrupts are rare. Somebody set up a cron
4058 job to run exiwhat every 2 minutes, which is less than the normal select()
4059 timeout (5 or 10 minutes). This meant that the select() timeout never
4060 kicked in because it was always reset. I have fixed this by comparing the
4061 time when an interrupt arrives with the time at the start of the first call
4062 to select(). If more time than the timeout has elapsed, the interrupt is
4063 treated as a timeout.
4065 5. Some internal re-factoring in preparation for the addition of Sieve
4066 extensions (by MH). In particular, the "personal" test is moved to a
4067 separate function, and given an option for scanning Cc: and Bcc: (which is
4068 not set for Exim filters).
4070 6. When Exim created an email address using the login of the caller as the
4071 local part (e.g. when creating a From: or Sender: header line), it was not
4072 quoting the local part when it contained special characters such as @.
4074 7. Installed new OpenBSD configuration files.
4076 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4077 try to make them clearer.
4079 9. Callout options, other than the timeout value, were being ignored when
4080 verifying sender addresses in header lines. For example, when using
4082 verify = header_sender/callout=no_cache
4084 the cache was (incorrectly) being used.
4086 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4087 only the Cygwin environment.
4089 11. When return_path_on_delivery was set as a log selector, if different remote
4090 addresses in the same message used different return paths and parallel
4091 remote delivery occurred, the wrong values would sometimes be logged.
4092 (Whenever a remote delivery process finished, the return path value from
4093 the most recently started remote delivery process was logged.)
4095 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4096 lines when AUTH and/or TLS are in use. This is the "received protocol"
4097 field. Exim used to use "asmtp" for authenticated SMTP, without any
4098 indication (in the protocol name) for TLS use. Now it follows the RFC and
4099 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4100 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4101 names appear in log lines as well as in Received: header lines.
4103 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4104 extensions, and comparison tests, and to fix some bugs.
4106 14. Changes to the "personal" filter test:
4108 (1) The test was buggy in that it was just doing the equivalent of
4109 "contains" tests on header lines. For example, if a user's address was
4110 anne@some.where, the "personal" test would incorrectly be true for
4112 To: susanne@some.where
4114 This test is now done by extracting each address from the header in turn,
4115 and checking the entire address. Other tests that are part of "personal"
4116 are now done using regular expressions (for example, to check local parts
4117 of addresses in From: header lines).
4119 (2) The list of non-personal local parts in From: addresses has been
4120 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4121 taken from the Sieve specification recommendations.
4123 (3) If the message contains any header line starting with "List-" it is
4124 treated as non-personal.
4126 (4) The test for "circular" in the Subject: header line has been removed
4127 because it now seems ill-conceived.
4129 15. Minor typos in src/EDITME comments corrected.
4131 16. Installed latest exipick from John Jetmore.
4133 17. If headers_add on a router specified a text string that was too long for
4134 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4135 of string_sprintf() is now avoided.
4137 18. $message_body_size was not set (it was always zero) when running the DATA
4138 ACL and the local_scan() function.
4140 19. For the "mail" command in an Exim filter, no default was being set for
4141 the once_repeat time, causing a random time value to be used if "once" was
4142 specified. (If the value happened to be <= 0, no repeat happened.) The
4143 default is now 0s, meaning "never repeat". The "vacation" command was OK
4144 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4145 (I found it when inspecting the code).
4147 20. There is now an overall timeout for performing a callout verification. It
4148 defaults to 4 times the callout timeout, which applies to individual SMTP
4149 commands during the callout. The overall timeout applies when there is more
4150 than one host that can be tried. The timeout is checked before trying the
4151 next host. This prevents very long delays if there are a large number of
4152 hosts and all are timing out (e.g. when the network connections are timing
4153 out). The value of the overall timeout can be changed by specifying an
4154 additional sub-option for "callout", called "maxwait". For example:
4156 verify = sender/callout=5s,maxwait=20s
4158 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4159 to the end before writing, but this should make it even safer).
4161 22. Exim was forgetting that it had advertised PIPELINING for the second and
4162 subsequent messages on an SMTP connection. It was also not resetting its
4163 memory on STARTTLS and an internal HELO.
4165 23. When Exim logs an SMTP synchronization error within a session, it now
4166 records whether PIPELINING has been advertised or not.
4168 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4169 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4170 rather than long int.
4172 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4174 26. Added the never_mail option to autoreply.
4180 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4181 crash if the getsockname() call failed; this can happen if a connection is
4182 closed very soon after it is established. The problem was simply in the
4183 order in which certain operations were done, causing Exim to try to write
4184 to the SMTP stream before it had set up the file descriptor. The bug has
4185 been fixed by making things happen in the correct order.
4191 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4192 before the connection was closed, thus losing the rejection response.
4194 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4195 some early Solaris releases, but causes trouble in current releases where
4196 socklen_t is defined.
4198 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4201 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4204 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4205 that releasing the top of it at the end releases what was used for sub-
4206 expansions (unless the block got too big). However, discard this block if
4207 the first thing is a variable or header, so that we can use its block when
4208 it is dynamic (useful for very large $message_headers, for example).
4210 6. Lookups now cache *every* query, not just the most recent. A new, separate
4211 store pool is used for this. It can be recovered when all lookup caches are
4212 flushed. Lookups now release memory at the end of their result strings.
4213 This has involved some general refactoring of the lookup sources.
4215 7. Some code has been added to the store_xxx() functions to reduce the amount
4216 of flapping under certain conditions.
4218 8. log_incoming_interface used to affect only the <= reception log lines. Now
4219 it causes the local interface and port to be added to several more SMTP log
4220 lines, for example "SMTP connection from", and rejection lines.
4222 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4224 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4226 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4227 could be overwritten at the end of the current message (or the start of a
4228 new message if it was set in a HELO ACL). The value is now preserved for
4229 the duration of the SMTP connection.
4231 12. If a transport had a headers_rewrite setting, and a matching header line
4232 contained an unqualified address, that address was qualified, even if it
4233 did not match any rewriting rules. The underlying bug was that the values
4234 of the flags that permit the existence of unqualified sender and recipient
4235 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4236 non-local messages, and by -bnq for local messages) were not being
4237 preserved with the message after it was received.
4239 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4240 "next input=" as part of the text comprising the host identity instead of
4241 the correct text. The code was using the same buffer for two different
4242 strings. However, depending on which order the printing function evaluated
4243 its arguments, the bug did not always show up. Under Linux, for example, my
4244 test suite worked just fine.
4246 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4247 doesn't work with some older versions of Perl. It has been changed to "my",
4248 which in any case is probably the better facility to use.
4250 15. A really picky compiler found some instances of statements for creating
4251 error messages that either had too many or two few arguments for the format
4254 16. The size of the buffer for calls to the DNS resolver has been increased
4255 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4256 for addresses that have a lot of PTR records. This alleviates a problem; it
4257 does not fully solve it.
4259 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4260 buffer now truncates the list and logs the incident, which is the same
4261 action as happens when Exim is looking up a host name and its aliases.
4262 Previously in this situation something unpredictable would happen;
4263 sometimes it was "internal error: store_reset failed".
4265 18. If a server dropped the connection unexpectedly when an Exim client was
4266 using GnuTLS and trying to read a response, the client delivery process
4267 crashed while trying to generate an error log message.
4269 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4270 string, for example:
4272 warn message = H1: something\nH2: something
4274 the text was added as a single header line from Exim's point of view
4275 though it ended up OK in the delivered message. However, searching for the
4276 second and subsequent header lines using $h_h2: did not work. This has been
4277 fixed. Similarly, if a system filter added multiple headers in this way,
4278 the routers could not see them.
4280 20. Expanded the error message when iplsearch is called with an invalid key to
4281 suggest using net-iplsearch in a host list.
4283 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4284 ACLs are no longer actually imposed (and a message to that effect is
4287 22. If a "gecos" field in a passwd entry contained escaped characters, in
4288 particular, if it contained a \" sequence, Exim got it wrong when building
4289 a From: or a Sender: header from that name. A second bug also caused
4290 incorrect handling when an unquoted " was present following a character
4291 that needed quoting.
4293 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4294 was not being matched caselessly.
4296 24. Arranged for all hyphens in the exim.8 source to be escaped with
4299 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4300 properly. Recipient callout cache records were still being keyed to include
4301 the sender, even when use_sender was set false. This led to far more
4302 callouts that were necessary. The sender is no longer included in the key
4303 when use_sender is false.
4305 26. Added "control = submission" modifier to ACLs.
4307 27. Added the ${base62d: operator to decode base 62 numbers.
4309 28. dnsdb lookups can now access SRV records.
4311 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4312 the configuration file.
4314 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4315 (-v) mode. This makes the output for a verbose queue run more intelligible.
4317 31. Added a use_postmaster feature to recipient callouts.
4319 32. Added the $body_zerocount variable, containing the number of binary zero
4320 bytes in the message body.
4322 33. The time of last modification of the "new" subdirectory is now used as the
4323 "mailbox time last read" when there is a quota error for a maildir
4326 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4328 35. Added +ignore_unknown as a special item in host lists.
4330 36. Code for decoding IPv6 addresses in host lists is now included, even if
4331 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4332 address was recognized as an IP address, but was then not correctly decoded
4333 into binary, causing unexpected and incorrect effects when compared with
4340 1. Very minor rewording of debugging text in manualroute to say "list of
4341 hosts" instead of "hostlist".
4343 2. If verify=header_syntax was set, and a header line with an unqualified
4344 address (no domain) and a large number of spaces between the end of the
4345 name and the colon was received, the reception process suffered a buffer
4346 overflow, and (when I tested it) crashed. This was caused by some obsolete
4347 code that should have been removed. The fix is to remove it!
4349 3. When running in the test harness, delay a bit after writing a bounce
4350 message to get a bit more predictability in the log output.
4352 4. Added a call to search_tidyup() just before forking a reception process. In
4353 theory, someone could use a lookup in the expansion of smtp_accept_max_
4354 per_host which, without the tidyup, could leave open a database connection.
4356 5. Added the variables $recipient_data and $sender_data which get set from a
4357 lookup success in an ACL "recipients" or "senders" condition, or a router
4358 "senders" option, similar to $domain_data and $local_part_data.
4360 6. Moved the writing of debug_print from before to after the "senders" test
4363 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4364 problems for message scanning, either using a data ACL, or using
4365 local_scan() because the Received: header was not generated till after they
4366 were called (in order to set the time as the time of reception completion).
4367 I have revised the way this works. The header is now generated after the
4368 body is received, but before the ACL or local_scan() are called. After they
4369 are run, the timestamp in the header is updated.
4375 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4376 before starting a queue runner without re-exec. This happened only when
4377 deliver_drop_privilege was set or when the Exim user was set to root. The
4378 effect of the bug was that timeouts during subsequent deliveries caused
4379 crashes instead of being properly handled. The handler is now left at its
4380 default (and expected) setting.
4382 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4383 message, again when deliver_drop_privilege is set or Exim is run as root.
4384 The bug described in (1) was not present in this case, but the tidying up
4385 of the other signals was missing. I have made the two cases consistent.
4387 3. The ignore_target_hosts setting on a manualroute router was being ignored
4388 for hosts that were looked up using the /MX notation.
4390 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4393 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4394 operated on the sender address. After changing the $sender_address to <>
4395 for the sender address verify, Exim was re-instated it as the original
4396 (before rewriting) address, but remembering that it had rewritten it, so it
4397 wasn't rewriting it again. This bug also had the effect of breaking the
4398 sender address verification caching when the sender address was rewritten.
4400 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4401 This has been changed so that if the ip literal address matches
4402 ignore_target_hosts, the router declines.
4404 7. Added expansion conditions match_domain, match_address, and match_local_
4405 part (NOT match_host).
4407 8. The placeholder for the Received: header didn't have a length field set.
4409 9. Added code to Exim itself and to exim_lock to test for a specific race
4410 condition that could lead to file corruption when using MBX delivery. The
4411 issue is with the lockfile that is created in /tmp. If this file is removed
4412 after a process has opened it but before that process has acquired a lock,
4413 there is the potential for a second process to recreate the file and also
4414 acquire a lock. This could lead to two Exim processes writing to the file
4415 at the same time. The added code performs the same test as UW imapd; it
4416 checks after acquiring the lock that its file descriptor still refers to
4417 the same named file.
4419 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4420 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4423 11. Added the smtp_active_hostname option. If used, this will typically be made
4424 to depend on the incoming interface address. Because $interface_address is
4425 not set up until the daemon has forked a reception process, error responses
4426 that can happen earlier (such as "too many connections") no longer contain
4429 12. If an expansion in a condition on a "warn" statement fails because a lookup
4430 defers, the "warn" statement is abandoned, and the next ACL statement is
4431 processed. Previously this caused the whole ACL to be aborted.
4433 13. Added the iplsearch lookup type.
4435 14. Added ident_timeout as a log selector.
4437 15. Added tls_certificate_verified as a log selector.
4439 16. Added a global option tls_require_ciphers (compare the smtp transport
4440 option of the same name). This controls incoming TLS connections.
4442 17. I finally figured out how to make tls_require_ciphers do a similar thing
4443 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4444 before starting the TLS session.
4446 18. Tabs are now shown as \t in -bP output.
4448 19. If the log selector return_path_on_delivery was set, Exim crashed when
4449 bouncing a message because it had too many Received: header lines.
4451 20. If two routers both had headers_remove settings, and the first one included
4452 a superfluous trailing colon, the final name in the first list and the
4453 first name in the second list were incorrectly joined into one item (with a
4454 colon in the middle).
4460 1. Added -C and -D options to the exinext utility, mainly to make it easier
4461 to include in the automated testing, but these could be helpful when
4462 multiple configurations are in use.
4464 2. The exinext utility was not formatting the output nicely when there was
4465 an alternate port involved in the retry record key, nor when there was a
4466 message id as well (for retries that were specific to a specific message
4467 and a specific host). It was also confused by IPv6 addresses, because of
4468 the additional colons they contain. I have fixed the IPv4 problem, and
4469 patched it up to do a reasonable job for IPv6.
4471 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4472 delivery, the log line now contains "pipelined" if PIPELINING was used.
4474 4. An SMTP transport process used to panic and die if the bind() call to set
4475 an explicit outgoing interface failed. This has been changed; it is now
4476 treated in the same way as a connect() failure.
4478 5. A reference to $sender_host_name in the part of a conditional expansion
4479 that was being skipped was still causing a DNS lookup. This no longer
4482 6. The def: expansion condition was not recognizing references to header lines
4483 that used bh_ and bheader_.
4485 7. Added the _cache feature to named lists.
4487 8. The code for checking quota_filecount in the appendfile transport was
4488 allowing one more file than it should have been.
4490 9. For compatibility with Sendmail, the command line option
4498 and sets the incoming protocol and host name (for trusted callers). The
4499 host name and its colon can be omitted when only the protocol is to be set.
4500 Note the Exim already has two private options, -pd and -ps, that refer to
4501 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4502 "s", but I don't think that's a major issue.
4504 10. A number of refactoring changes to the code, none of which should affect
4507 (a) The number of logging options was getting close to filling up the
4508 32-bit word that was used as a bit map. I have split them into two classes:
4509 those that are passed in the argument to log_write(), and those that are
4510 only ever tested independently outside of that function. These are now in
4511 separate 32-bit words, so there is plenty of room for expansion again.
4512 There is no change in the user interface or the logging behaviour.
4514 (b) When building, for example, log lines, the code previously used a
4515 macro that called string_cat() twice, in order to add two strings. This is
4516 not really sufficiently general. Furthermore, there was one instance where
4517 it was actually wrong because one of the argument was used twice, and in
4518 one call a function was used. (As it happened, calling the function twice
4519 did not affect the overall behaviour.) The macro has been replaced by a
4520 function that can join an arbitrary number of extra strings onto a growing
4523 (c) The code for expansion conditions now uses a table and a binary chop
4524 instead of a serial search (which was left over from when there were very
4525 few conditions). Also, it now recognizes conditions like "pam" even when
4526 the relevant support is not compiled in: a suitably worded error message is
4527 given if an attempt is made to use such a condition.
4529 11. Added ${time_interval:xxxxx}.
4531 12. A bug was causing one of the ddress fields not to be passed back correctly
4532 from remote delivery subprocesses. The field in question was not being
4533 subsequently used, so this caused to problems in practice.
4535 13. Added new log selectors queue_time and deliver_time.
4537 14. Might have fixed a bug in maildirsizefile handling that threw up
4538 "unexpected character" debug warnings, and recalculated the data
4539 unnecessarily. In any case, I expanded the warning message to give more
4542 15. Added the message "Restricted characters in address" to the statements in
4543 the default ACL that block characters like @ and % in local parts.
4545 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4546 Three changes have been made:
4548 (a) There was a serious bug; a negative response to MAIL caused the whole
4549 recipient domain to be cached as invalid, thereby blocking all messages
4550 to all local parts at the same domain, from all senders. This bug has
4551 been fixed. The domain is no longer cached after a negative response to
4552 MAIL if the sender used is not empty.
4554 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4557 (c) A new callout option, "use_sender" has been added for people who want
4558 the modified behaviour.
4564 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4567 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4568 indeed breaks things for older releases.
4570 3. Added additional logging to the case where there is a problem reading data
4571 from a filter that is running in a subprocess using a pipe, in order to
4572 try to track down a specific problem.
4574 4. Testing facility fudge: when running in the test harness and attempting
4575 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4576 getting "No route to host". Convert this to a timeout.
4578 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4581 6. Some OS don't have socklen_t but use size_t instead. This affects the
4582 fifth argument of getsockopt() amongst other things. This is now
4583 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4584 can be set for individual OS. I have set it for SunOS5, OSF1, and
4585 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4586 some earlier ones do not.
4588 7. Change 4.30/15 was not doing the test caselessly.
4590 8. The standard form for an IPv6 address literal was being rejected by address
4591 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4592 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4593 this, as well as the form without the "IPv6" on the front (but only when
4594 address literals are enabled, of course).
4596 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4598 10. Exim crashed if a message with an empty sender address specified by -f
4599 encountered a router with an errors_to setting. This could be provoked only
4600 by a command such as
4604 where an empty string was supplied; "<>" did not hit this bug.
4606 11. Installed PCRE release 4.5.
4608 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4609 remained set. It is now erased.
4611 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4612 times from message ids (which are base 36 rather than the normal 62).
4614 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4615 were being counted as actual protocol errors, and logged if the log
4616 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4617 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4618 and DATA following a set of rejected RCPTs do not count as protocol errors.
4619 In other words, Exim assumes they were pipelined, though this may not
4620 actually be the case. Of course, in all cases the client gets an
4621 appropriate error code.
4623 15. If a lookup fails in an ACL condition, a message about the failure may
4624 be available; it is used if testing the ACL cannot continue, because most
4625 such messages specify what the cause of the deferral is. However, some
4626 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4627 that caused an old message to be retained and used if a later statement
4628 caused a defer, replacing the real cause of the deferral.
4630 16. If an IP address had so many PTR records that the DNS lookup buffer
4631 was not large enough to hold them, Exim could crash while trying to process
4632 the truncated data. It now detects and logs this case.
4634 17. Further to 4.21/58, another change has been made: if (and only if) the
4635 first line of a message (the first header line) ends with CRLF, a bare LF
4636 in a subsequent header line has a space inserted after it, so as not to
4637 terminate the header.
4639 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4640 unnecessarily, used atoi() instead of strtol(), and didn't check the
4641 termination when getting file sizes from file names by regex.
4643 19. Completely re-implemented the support for maildirsize files, in the light
4644 of a number of problems with the previous contributed implementation
4645 (4.30/29). In particular:
4647 . If the quota is zero, the maildirsize file is maintained, but no quota is
4650 . If the maildir directory does not exist, it is created before any attempt
4651 to write a maildirsize file.
4653 . The quota value in the file is just a cache; if the quota is changed in
4654 the transport, the new value overrides.
4656 . A regular expression is available for excluding directories from the
4659 20. The autoreply transport checks the characters in options that define the
4660 message's headers; it allows continued headers, but it was checking with
4661 isspace() after an embedded newline instead of explicitly looking for a
4664 21. If all the "regular" hosts to which an address was routed had passed their
4665 expiry times, and had not reached their retry times, the address was
4666 bounced, even if fallback hosts were defined. Now Exim should go on to try
4669 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4670 equivalent code in the SMTP transport. Some hosts send humungous responses
4671 to HELO/EHLO, more than 1024 it seems.
4673 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4674 gives compiler warnings in some environments. I've now done it "properly",
4677 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4678 (because of problems with the built-in one) was declared to return uschar *
4679 instead of char *, causing compiler failure.
4681 25. Fixed a file descriptor leak when processing alias/forward files.
4683 26. Fixed a minor format string issue in dbfn.c.
4685 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4687 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4688 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4691 29. When a sender address is verified, it is cached, to save repeating the test
4692 when there is more than one recipient in a message. However, when the
4693 verification involves a callout, it is possible for different callout
4694 options to be set for different recipients. It is too complicated to keep
4695 track of this in the cache, so now Exim always runs a verification when a
4696 callout is required, relying on the callout cache for the optimization.
4697 The overhead is duplication of the address routing, but this should not be
4700 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4701 to be verified with callout=postmaster, and the main callout worked but the
4702 postmaster check failed, the verification correctly failed. However, if a
4703 subsequent RCPT command asked for sender verification *without* the
4704 postmaster check, incorrect caching caused this verification also to fail,
4707 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4708 it was not caching the DNS options (qualify_single, search_parents) that
4709 were used when the lookup failed. A subsequent lookup with different
4710 options therefore always gave the same answer, though there were cases
4711 where it should not have. (Example: a "domains = !$mx_any" option on a
4712 dnslookup router: the "domains" option is always processed without any
4713 widening, but the router might have qualify_single set.) Now Exim uses the
4714 cached value only when the same options are set.
4716 32. Added John Jetmore's "exipick" utility to the distribution.
4718 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4719 than a timeout (e.g. a certificate is required, and is not provided), an
4720 Exim server now closes the connection immediately. Previously it waited for
4721 the client to close - but if the client is SSL, it seems that they each
4722 wait for each other, leading to a delay before one of them times out.
4724 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4725 maintained 0.8.x compatibility because I don't think many are using it, and
4726 it is clearly obsolete.
4728 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4731 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4732 client certificate was expired. A simple patch fixes this, though I don't
4733 understand the full logic of why the verify callback is called multiple
4736 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4737 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4738 which causes problems with some clients (such as the Certicom SSL Plus
4739 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4740 disables the coutermeasure allowing Eudora to connect."
4742 38. Exim was not checking that a write() to a log file succeeded. This could
4743 lead to Bad Things if a log got too big, in particular if it hit a file
4744 size limit. Exim now panics and dies if it cannot write to a log file, just
4745 as it does if it cannot open a log file.
4747 39. Modified OS/Makefile-Linux so that it now contains
4749 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4751 The two -D definitions ensure that Exim is compiled with large file
4752 support, which makes it possible to handle log files that are bigger than
4755 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4756 instance) a domain was checked against a named list that involved a lookup,
4757 causing $domain_data to be set, then another domain was checked against the
4758 same list, then the first domain was re-checked, the value of $domain_data
4759 after the final check could be wrong. In particular, if the second check
4760 failed, it could be set empty. This bug probably also applied to
4763 41. The strip_trailing_dot option was not being applied to the address given
4764 with the -f command-line option.
4766 42. The code for reading a message's header from the spool was incrementing
4767 $received_count, but never initializing it. This meant that the value was
4768 incorrect (doubled) while delivering a message in the same process in which
4769 it was received. In the most common configuration of Exim, this never
4770 happens - a fresh exec is done - but it can happen when
4771 deliver_drop_privilege is set.
4773 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4774 it now includes up to 150 characters of the unexpected data in the log
4777 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4778 and building data strings. The size of both of these buffers was 10 000
4779 bytes - far larger than anybody would *ever* want, thought I. Needless to
4780 say, somebody hit the limit. I have increased the maximum line length to
4781 20 000 and the maximum data length of concatenated lines to 100 000. I have
4782 also fixed two bugs, because there was no checking on these buffers. Tsk,
4783 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4784 buffer is too small.
4786 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4787 lsearch lookups. Now it does.
4789 46. When parsing a route_list item in a manualroute router, a fixed-length
4790 buffer was used for the list of hosts. I made this 1024 bytes long,
4791 thinking that nobody would ever have a list of hosts that long. Wrong.
4792 Somebody had a whole pile of complicated expansion conditions, and the
4793 string was silently truncated, leading to an expansion error. It turns out
4794 that it is easier to change to an unlimited length (owing to other changes
4795 that have happened since this code was originally written) than to build
4796 structure for giving a limitation error. The length of the item that
4797 expands into the list of hosts is now unlimited.
4799 47. The lsearch lookup could not handle data where the length of text line was
4800 more than 4095 characters. Such lines were truncated, leading to shortened
4801 data being returned. It should now handle lines of any length.
4803 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4804 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4807 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4808 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4809 no longer happens in comment lines. A list of replacements is now placed
4810 at the head of all of the source files, except those whose only change is
4811 to replace PERL_COMMAND in the very first #! line.
4813 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4814 messages on the queue, with a bottom-up merge sort, using code contributed
4815 by Michael Haardt. This should make operations like -bp somewhat faster on
4816 large queues. It won't affect queue runners, except when queue_run_in_order
4819 51. Installed eximstats 1.31 in the distribution.
4821 52. Added support for SRV lookups to the dnslookup router.
4823 53. If an ACL referred to $message_body or $message_body_end, the value was not
4824 reset for any messages that followed in the same SMTP session.
4826 54. The store-handling optimization for building very long strings was not
4827 differentiating between the different store pools. I don't think this
4828 actually made any difference in practice, but I've tidied it.
4830 55. While running the routers to verify a sender address, $sender_address
4831 was still set to the sender address. This is wrong, because when routing to
4832 send a bounce to the sender, it would be empty. Therefore, I have changed
4833 it so that, while verifying a sender address, $sender_address is set to <>.
4834 (There is no change to what happens when verifying a recipient address.)
4836 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4837 target A or AAAA records (if not already returned) without resetting the
4838 qualify_single or search_parents options of the DNS resolver. These are
4839 inappropriate in this case because the targets of MX and SRV records must
4840 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4841 target that, when qualified, matched something in the local domain. These
4842 two options are now turned off when doing these lookups.
4844 57. It seems that at least some releases of Reiserfs (which does not have the
4845 concept of a fixed number of inodes) returns zero and not -1 for the
4846 number of available inodes. This interacted badly with check_spool_inodes,
4847 which assumed that -1 was the "no such thing" setting. What I have done is
4848 to check that the total number of inodes is greater than zero before doing
4849 the test of how many are available.
4851 58. When a "warn" ACL statement has a log_message modifier, the message is
4852 remembered, and not repeated. This is to avoid a lot of repetition when a
4853 message has many recipients that cause the same warning to be written.
4854 Howewer, Exim was preserving the list of already written lines for an
4855 entire SMTP session, which doesn't seem right. The memory is now reset if a
4856 new message is started.
4858 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4859 debugging output unless log_rewrite was also set.
4861 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4862 of (int)(handle) when we know that handle contains (void *)(-1).
4864 61. The Exim daemon panic-logs an error return when it closes the incoming
4865 connection. However "connection reset by peer" seems to be common, and
4866 isn't really an error worthy of noting specially, so that particular error
4869 62. When Exim is trying to find all the local interfaces, it used to panic and
4870 die if the ioctl to get the interface flags failed. However, it seems that
4871 on at least one OS (Solaris 9) it is possible to have an interface that is
4872 included in the list of interfaces, but for which you get a failure error
4873 for this call. This happens when the interface is not "plumbed" into a
4874 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4875 failure of the "get flags" call assumes that the interface is down.
4877 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4878 makes life easier for people who are doing arithmetic on fields extracted
4879 from dates, where you often get leading zeros that should not be
4880 interpreted as octal.
4882 64. Added qualify_domain to the redirect router, to override the global
4885 65. If a pathologically long header line contained very many addresses (the
4886 report of this problem mentioned 10 000) and each of them was rewritten,
4887 Exim could use up a very large amount of memory. (It kept on making new
4888 copies of the header line as it rewrote, and never released the old ones.)
4889 At the expense of a bit more processing, the header rewriting function has
4890 been changed so that it no longer eats memory in this way.
4892 66. The generation of the Received: header has been moved from the time that a
4893 message starts to be received, to the time that it finishes. The timestamp
4894 in the Received: header should now be very close to that of the <= log
4895 line. There are two side-effects of this change:
4897 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4898 logged header lines no longer include the local Received: line, because
4899 it has not yet been created. The same applies to a copy of the message
4900 that is returned to a non-SMTP sender when a message is rejected.
4902 (b) When a filter file is tested using -bf, no additional Received: header
4903 is added to the test message. After some thought, I decided that this
4906 This change does not affect the value of $received_for. It is still set
4907 after address rewriting, but before local_scan() is called.
4909 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4911 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4912 gave an unhelpful panic error message, and a defer error. I have managed to
4913 change this behaviour so that it now rejects any supplied certificate,
4914 which seems right, as the list of acceptable certificates is empty.
4916 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4917 gave an unhelpful defer error. I have not managed to make this reject any
4918 supplied certificates, but the error message it gives is "no certificate
4919 supplied", which is not helpful.
4921 70. exigrep's output now also includes lines that are not associated with any
4922 message, but which match the given pattern. Implemented by a patch from
4923 Martin Sluka, which also tidied up the Perl a bit.
4925 71. Recipient callout verification, like sender verification, was using <> in
4926 the MAIL FROM command. This isn't really the right thing, since the actual
4927 sender may affect whether the remote host accepts the recipient or not. I
4928 have changed it to use the actual sender in the callout; this means that
4929 the cache record is now keyed on a recipient/sender pair, not just the
4930 recipient address. There doesn't seem to be a real danger of callout loops,
4931 since a callout by the remote host to check the sender would use <>.
4932 [SEE ABOVE: changed after hitting problems.]
4934 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4935 temporary errors. However, in the case of such a code being given after
4936 the end of a data transmission (i.e. after ".") Exim was failing to write
4937 a retry record for the message. (Yes, there was some broken host that was
4938 actually sending 8xx at this point.)
4940 73. An unknown lookup type in a host list could cause Exim to panic-die when
4941 the list was checked. (An example that provoked this was putting <; in the
4942 middle of a list instead of at the start.) If this happened during a DATA
4943 ACL check, a -D file could be left lying around. This kind of configuration
4944 error no longer causes Exim to die; instead it causes a defer errror. The
4945 incident is still logged to the main and panic logs.
4947 74. Buglet left over from Exim 3 conversion. The message "too many messages
4948 in one connection" was written to the rejectlog but not the mainlog, except
4949 when address rewriting (yes!) was being logged.
4951 75. Added write_rejectlog option.
4953 76. When a system filter was run not as root (that is, when system_filter_user
4954 was set), the values of the $n variables were not being returned to the
4955 main process; thus, they were not subsequently available in the $sn
4958 77. Added +return_path_on_delivery log selector.
4960 78. A connection timeout was being treated differently from recipients deferred
4961 when testing hosts_max_try with a message that was older than the host's
4962 retry timeout. (The host should not be counted, thus allowing all hosts to
4963 be tried at least once before bouncing.) This may have been the cause of an
4964 occasionally reported bug whereby a message would remain on the queue
4965 longer than the retry timeout, but would be bounced if a delivery was
4966 forced. I say "may" because I never totally pinned down the problem;
4967 setting up timeout/retry tests is difficult. See also the next item.
4969 79. The ultimate address timeout was not being applied to errors that involved
4970 a combination of host plus message (for example, a timeout on a MAIL
4971 command). When an address resolved to a number of possible hosts, and they
4972 were not all tried for each delivery (e.g. because of hosts_max_try), a
4973 message could remain on the queue longer than the retry timeout.
4975 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4978 81. Fixed an obscure SMTP outgoing bug which required at least the following
4979 conditions: (a) there was another message waiting for the same server;
4980 (b) the server returned 5xx to all RCPT commands in the first message so
4981 that the message was not completed; (c) the server dropped the connection
4982 or gave a negative response to the RSET that Exim sends to abort the
4983 transaction. The observed case was a dropped connection after DATA that had
4984 been sent in pipelining mode. That is, the server had advertised PIPELINING
4985 but was not implementing it correctly. The effect of the bug was incorrect
4986 behaviour, such as trying another host, and this could lead to a crash.
4992 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4993 and daemon.c were passed as pointers to ints; they should have been
4994 pointers to socklen_t variables (which are typically unsigned ints).
4996 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4999 3. Fixed a really odd bug that affected only the testing scheme; patching a
5000 certain fixed string in the binary changed the value of another string that
5001 happened to be identical to the end of the original first string.
5003 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5004 name", it returns that address as the IP address. On some operating
5005 systems (e.g. Solaris), it also passes back the IP address string as the
5006 "host name". However, on others (e.g. Linux), it passes back an empty
5007 string. Exim wasn't checking for this, and was changing the host name to an
5008 empty string, assuming it had been canonicized.
5010 5. Although rare, it is permitted to have more than one PTR record for a given
5011 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5012 all the names associated with an address, because they do in Solaris.
5013 However, it seems that they do not in Linux for data that comes from the
5014 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5015 I found this out when I moved to a new Linux workstation and tried to run
5016 the Exim test suite.
5018 To get round this problem I have changed the code so that it now does its
5019 own call to the DNS to look up PTR records when searching for a host name.
5020 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5021 addresses that are only in /etc/hosts are still found.
5023 This behaviour is, however, controlled by an option called host_lookup_
5024 order, which defaults to "bydns:byaddr". If people want to use the other
5025 order, or indeed, just use one or the other means of lookup, they can
5026 specify it in this variable.
5028 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5029 some operating systems, this comes back from gethostbyaddr() as an empty
5030 string, and this is what Exim used to test for. However, it seems that in
5031 other systems, "." is yielded. Exim now tests for this case too.
5033 7. The values of check_spool_space and check_log_space are now held internally
5034 as a number of kilobytes instead of an absolute number of bytes. If a
5035 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5036 kilobyte. This means that much larger values can be stored.
5038 8. Exim monitor: an attempt to get the action menu when not actually pointing
5039 at a message produces an empty menu entitled "No message selected". This
5040 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5041 no entries in it ("Shell widget menu has zero width and/or height"). So I
5042 have added a single, blank menu entry in this case.
5044 9. Added ${quote_local_part.
5046 10. MIME decoding is now applied to the contents of Subject: header lines when
5049 11. Now that a reference to $sender_host_address automatically causes a reverse
5050 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5051 host lookup before query-style lookups in lists that might use this
5052 variable. This has therefore been abolished, and the "net-" prefix is no
5053 longer necessary for query-style lookups.
5055 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5056 have been a typo for LFLAGS, so it has been changed.
5058 13. The install script calls Exim with "-C /dev/null" in order to find the
5059 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5060 to be output. Howeve, since Exim outputs its version number before the
5061 error, it didn't break the script. It just looked ugly. I fixed this by
5062 always allowing "-C /dev/null" if the caller is root.
5064 14. Ignore overlarge ACL variable number when reading spool file - insurance
5065 against a later release with more variables having written the file.
5067 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5068 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5069 this, as well as the form without the "IPv6" on the front.
5071 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5072 OS/Makefile-Darwin file.
5074 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5075 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5077 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5078 variable. However, some OS use ints for inodes. I've added cast to long int
5079 to get rid of the compiler warning.
5081 19. I had forgotten to lock out "/../" in configuration file names when
5082 ALT_CONFIG_PREFIX was set.
5084 20. Routers used for verification do not need to specify transports. However,
5085 if such a router generated a host list, and callout was configured, Exim
5086 crashed, because it could not find a port number from the (non-existent)
5087 transport. It now assumes port 25 in this circumstance.
5089 21. Added the -t option to exigrep.
5091 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5092 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5093 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5094 them is set, LOOKUP_LSEARCH is forced.
5096 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5098 24. Added sender and host information to the "rejected by local_scan()" log
5099 line; previously there was no indication of these.
5101 25. Added .include_if_exists.
5103 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5104 Linux. It turns out that not all file systems support this. Apparently some
5105 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5106 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5107 not supported on the file descriptor, is now ignored when Exim is trying to
5108 sync a directory. This applies only to Linux.
5110 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5112 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5113 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5116 29. Added support for maildirsize files from supplied patch (modified a bit).
5118 30. The use of :fail: followed by an empty string could lead Exim to respond to
5119 sender verification failures with (e.g.):
5121 550 Verification failed for <xxx>
5122 550 Sender verify failed
5124 where the first response line was missing the '-' that indicates it is not
5125 the final line of the response.
5127 31. The loop for finding the name of the user that called Exim had a hardwired
5128 limit of 10; it now uses the value of finduser_retries, which is used for
5129 all other user lookups.
5131 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5134 33. Exim was neglecting to zero errno before one call of strtol() when
5135 expanding a string and expecting an integer value. On some systems this
5136 resulted in spurious "integer overflow" errors. Also, it was casting the
5137 result into an int without checking.
5139 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5140 did not work. The code looks as if it has *never* worked, though it appears
5141 to have been documented since at least releast 1.62. I have made it work.
5143 35. The "timeout_DNS" error in retry rules, also documented since at least
5144 1.62, also never worked. As it isn't clear exactly what this means, and
5145 clearly it isn't a major issue, I have abolished the feature by treating it
5146 as "timeout", and writing a warning to the main and panic logs.
5148 36. The display of retry rules for -brt wasn't always showing the error code
5151 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5152 timeout_connect_A, timeout_connect_MX.
5154 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5155 to the empty sender.
5157 39. The daemon was not analysing the content of -oX till after it had closed
5158 stderr and disconnected from the controlling terminal. This meant that any
5159 syntax errors were only noted on the panic log, and the return code from
5160 the command was 0. By re-arranging the code a little, I've made the
5161 decoding happen first, so such errors now appear on stderr, and the return
5162 code is 1. However, the actual setting up of the sockets still happens in
5163 the disconnected process, so errors there are still only recorded on the
5166 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5167 connections (as happens on some IP stacks) was logged at start up time as
5168 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5169 it from "IPv6 and IPv4", which means that two separate sockets are being
5172 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5173 says whether AF_INET or AF_INET6 was passed as an argument.
5175 42. Exiwhat output was messed up when time zones were included in log
5178 43. Exiwhat now gives more information about the daemon's listening ports,
5179 and whether -tls-on-connect was used.
5181 44. The "port" option of the smtp transport is now expanded.
5183 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5184 silently ignored. Now an error message is written to the main and panic
5187 46. There's a new ACL modifier called "logwrite" which writes to a log file
5188 as soon as it is encountered.
5190 47. Added $local_user_uid and $local_user_gid at routing time.
5192 48. Exim crashed when trying to verify a sender address that was being
5195 49. Exim was recognizing only a space character after ".include". It now also
5196 recognizes a tab character.
5198 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5199 extracting the relevant information from the specification. The man page no
5200 longer contains scrambled data for the -d option, and I've added a section
5201 at the front about calling Exim under different names.
5203 51. Added "extra_headers" argument to the "mail" command in filter files.
5205 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5208 53. Installed eximstats 1.29.
5210 54. Added transport_filter_timeout as a generic transport option.
5212 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5213 Cc: header lines. This was required by RFC 822, but it not required by RFC
5216 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5217 incoming messages that did not have them. Now it does so only if the
5218 message originates locally, that is, if there is no associated remote host
5219 address. When Resent- header lines are present, this applies to the Resent-
5220 lines rather than the non-Resent- lines.
5222 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5223 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5225 58. Messages for configuration errors now include the name of the main
5226 configuration file - useful now that there may be more than one file in a
5227 list (.included file names were always shown).
5229 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5230 for those rare installations that do not start the daemon as root or run it
5231 setuid root. I've cut out the call to initgroups() if the daemon is not
5234 60. The Exim user and group can now be bound into the binary as text strings
5235 that are looked up at the start of Exim's processing.
5237 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5239 62. Added $mailstore_basename variable.
5241 63. Installed patch to sieve.c from Michael Haardt.
5243 64. When Exim failed to open the panic log after failing to open the main log,
5244 the original message it was trying to log was written to stderr and debug
5245 output, but if they were not available (the usual case in production), it
5246 was lost. Now it is written to syslog before the two lines that record the
5247 failures to open the logs.
5249 65. Users' Exim filters run in subprocesses under the user's uid. It is
5250 possible for a "deliver" command or an alias in a "personal" command to
5251 provoke an address rewrite. If logging of address rewriting is configured,
5252 this fails because the process is not running as root or exim. There may be
5253 a better way of dealing with this, but for the moment (because 4.30 needs
5254 to be released), I have disabled address rewrite logging when running a
5255 filter in a non-root, non-exim process.
5261 1. The buildconfig auxiliary program wasn't quoting the value set for
5262 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5263 not defined. This bug was masked in 4.22 by the effect that was fixed in
5266 2. Some messages that were rejected after a message id was allocated were
5267 shown as "incomplete" by exigrep. It no longer does this for messages that
5268 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5270 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5271 have allow_domain_literals set, the ID did not get logged in the <= line.
5272 Domain literals are now always recognized in Message-ID: header lines.
5274 4. The first argument for a ${extract expansion item is the key name or field
5275 number. Leading and trailing spaces in this item were not being ignored,
5276 causing some misleading effects.
5278 5. When deliver_drop_privilege was set, single queue runner processes started
5279 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5280 same command in the process it spins off) were not dropping privilege.
5282 6. When the daemon running as "exim" started a queue runner, it always
5283 re-executed Exim in the spun-off process. This is a waste of effort when
5284 deliver_drop_privilege is set. The new process now just calls the
5285 queue-runner function directly.
5291 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5294 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5295 ignored. Though the use of -oP was forcing the writing of a pid file, it
5296 was always written to the default place.
5298 3. If the message "no IP address found for host xxxx" is generated during
5299 incoming verification, it is now followed by identification of the incoming
5300 connection (so you can more easily find what provoked it).
5302 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5304 5. Added some features to "harden" Exim a bit more against certain attacks:
5306 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5307 be put in Local/Makefile. This is like the never_users runtime option,
5308 but it cannot be overridden. The default setting is "root".
5310 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5311 prefix string with which any file named in a -C command line option
5314 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5315 is retained for -C and -D only if the caller of Exim is root. Without
5316 it, the exim user may also use -C and -D and retain privilege.
5318 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5319 command line option is disabled.
5321 6. Macro names set by the -D option must start with an upper case letter, just
5322 like macro names defined in the configuration file.
5324 7. Added "dereference=" facility to LDAP.
5326 8. Two instances of the typo "uknown" in the source files are fixed.
5328 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5329 the Configure-Makefile script screwed up while processing it.
5331 10. Incorporated PCRE 4.4.
5333 11. The SMTP synchronization check was not operating right at the start of an
5334 SMTP session. For example, it could not catch a HELO sent before the client
5335 waited for the greeting. There is now a check for outstanding input at the
5336 point when the greeting is written. Because of the duplex, asynchronous
5337 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5338 way, but not yet received, when the check is performed.
5340 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5341 on TCP/IP sockets, because this apparently causes some broken clients to
5344 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5345 unchanged) from the Cygwin maintainer.
5347 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5348 was supported instead of testing for mbx. Effectively a typo.
5350 15. The spa authenticator server code was not checking that the input it
5351 received was valid base64.
5353 16. The debug output line for the "set" modifier in ACLs was not showing the
5354 name of the variable that was being set.
5356 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5358 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5359 lookup on its own. Something else was needed to trigger the lookup. For
5360 example, a match in host_lookup or the need for a host name in a host list.
5361 Now, if $sender_host_name is referenced and the host name has not yet been
5362 looked up, a lookup is performed. If the lookup fails, the variable remains
5363 empty, and $host_lookup_failed is set to "1".
5365 19. Added "eqi" as a case-independent comparison operator.
5367 20. The saslauthd authentication condition could segfault if neither service
5368 nor realm was specified.
5370 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5371 error message that was logged was misleading, and incoming SMTP
5372 connections were dropped. The message is now more accurate, and temporary
5373 errors are given to SMTP connections.
5375 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5376 (except RSET) with a 421 error, until QUIT is received. However, it was
5377 failing to send a response to QUIT.
5379 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5380 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5381 ACL is now run after the helo_try_verify_hosts code.
5383 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5384 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5385 case-independent, so other case variants are also recognized). Apparently
5386 some systems use these upper case variants.
5388 25. If more than two messages were waiting for the same host, and a transport
5389 filter was specified for the transport, Exim sent two messages over the
5390 same TCP/IP connection, and then failed with "socket operation on non-
5391 socket" when it tried to send the third.
5393 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5395 27. The extern definition of crypt16() in expand.c was not being excluded when
5396 the OS had its own crypt16() function.
5398 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5399 as a preferred synonym for return_size_limit, both as an option and as an
5402 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5404 30. Changed the default configuration ACL to relax the local part checking rule
5405 for addresses that are not in any local domains. For these addresses,
5406 slashes and pipe symbols are allowed within local parts, but the sequence
5407 /../ is explicitly forbidden.
5409 31. SPA server authentication was not clearing the challenge buffer before
5412 32. log_message in a "warn" ACL statement was writing to the reject log as
5413 well as to the main log, which contradicts the documentation and doesn't
5414 seem right (because no rejection is happening). So I have stopped it.
5416 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5417 However, I am unable to do any testing of this.
5419 34. Fixed an infelicity in the appendfile transport. When checking directories
5420 for a mailbox, to see if any needed to be created, it was accidentally
5421 using path names with one or more superfluous leading slashes; tracing
5422 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5424 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5425 contents are added to the log line that is written for every discarded
5426 recipient. (Previously a log_message setting was ignored.)
5428 36. The ${quote: operator now quotes the string if it is empty.
5430 37. The install script runs exim in order to find its version number. If for
5431 some reason other than non-existence or emptiness, which it checks, it
5432 could not run './exim', it was installing it with an empty version number,
5433 i.e. as "exim-". This error state is now caught, and the installation is
5436 38. An argument was missing from the function that creates an error message
5437 when Exim fails to connect to the socket for saslauthd authentication.
5438 This could cause Exim to crash, or give a corrupted message.
5440 39. Added isip, isip4, and isip6 to ${if conditions.
5442 40. The ACL variables $acl_xx are now saved with the message, and can be
5443 accessed later in routers, transports, and filters.
5445 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5446 strings in the file are not string-expanded.
5448 42. If a MAIL command specified a SIZE value that was too large to fit into an
5449 int variable, the check against message_size_limit failed. Such values are
5450 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5451 day this will have to be increased, but I don't think I want to be around
5452 when emails are that large.
5459 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5460 iconv() is not standard in FreeBSD.
5462 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5463 IPv6 enabled. The observed symptom was a segmentation fault on return from
5464 the function os_common_find_running_interfaces() in src/os.c.
5466 3. In the check_special_case() function in daemon.c I had used "errno" as an
5467 argument name, which causes warnings on some systems. This was basically a
5468 typo, since it was named "eno" in the comments!
5470 4. The code that waits for the clock to tick (at a resolution of some fraction
5471 of a second) so as to ensure message-id uniqueness was always waiting for
5472 at least one whole tick, when it could have waited for less. [This is
5473 almost certainly not relevant at current processor speeds, where it is
5474 unlikely to ever wait at all. But we try to future-proof.]
5476 5. The function that sleeps for a time interval that includes fractions of a
5477 second contained a race. It did not block SIGALRM between setting the
5478 timer, and suspending (a couple of lines later). If the interval was short
5479 and the sigsuspend() was delayed until after it had expired, the suspension
5480 never ended. On busy systems this could lead to processes getting stuck for
5483 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5484 process, before it forks any delivery processes. The open lookup caching
5485 mechanism meant that the open file or database connection was passed into
5486 the delivery process. The problem was that delivery processes always tidy
5487 up cached lookup data. This could cause a problem for the next delivery
5488 process started by the queue runner, because the external queue runner
5489 process does not know about the closure. So the next delivery process
5490 still has data in the lookup cache. In the case of a file lookup, there was
5491 no problem because closing a file descriptor in a subprocess doesn't affect
5492 the parent. However, if the lookup was caching a connection to a database,
5493 the connection was closed, and the second delivery process was likely to
5494 see errors such as "PGSQL: query failed: server closed the connection
5495 unexpectedly". The problem has been fixed by closing all cached lookups
5496 in a queue runner before running a delivery process.
5498 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5499 seem to have the "const" qualifier which it has on other OS. I've
5502 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5503 *of the same type* (client or server) with the same public name that an
5504 error should be diagnosed.
5506 9. When Exim looked up a host name for an IP address, but failed to find the
5507 original IP address when looking up the host name (a safety check), it
5508 output the message "<ip address> does not match any IP for NULL", which was
5509 confusing, to say the least. The bug was that the host name should have
5510 appeared instead of "NULL".
5512 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5513 user that is built into the binary, and the -C or -D options is used, root
5514 privilege is dropped before the configuration file is read. In addition,
5515 logging is switched to stderr instead of the normal log files. If the
5516 configuration then re-defines the Exim user, the unprivileged environment
5517 is probably not what is expected, so Exim logs a panic warning message (but
5520 However, if deliver_drop_privilege is set, the unprivileged state may well
5521 be exactly what is intended, so the warning has been cut out in that case,
5522 and Exim is allowed to try to write to its normal log files.
5528 1. smtp_return_error_details was not giving details for temporary sender
5529 or receiver verification errors.
5531 2. Diagnose a configuration error if two authenticators have the same public
5534 3. Exim used not to create the message log file for a message until the first
5535 delivery attempt. This could be confusing when incoming messages were held
5536 for policy or load reasons. The message log file is now created at the time
5537 the message is received, and an initial "Received" line is written to it.
5539 4. The automatically generated man page for command line options had a minor
5540 bug that caused no ill effects; however, a more serious problem was that
5541 the procedure for building the man page automatically didn't always
5542 operate. Consequently, release 4.20 contains an out-of-date version. This
5543 shouldn't happen again.
5545 5. When building Exim with embedded Perl support, the script that builds the
5546 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5547 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5549 6. The freeze_tell option was not being used for messages that were frozen on
5550 arrival, either by an ACL or by local_scan().
5552 7. Added the smtp_incomplete_transaction log selector.
5554 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5555 was accepting AUTH without a new EHLO.
5557 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5558 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5561 10. Logging of TCP/IP connections (when configured) now happens in the main
5562 daemon process instead of the child process, so that the TCP/IP connection
5563 count is more accurate (but it can never be perfect).
5565 11. The use of "drop" in a nested ACL was not being handled correctly in the
5566 outer ACL. Now, if condition failure induced by the nested "drop" causes
5567 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5568 or "require"), the connection is dropped.
5570 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5571 that yield "discard" can now be used with an "accept" or a "discard" verb,
5572 but an error is generated for any others (because I can't see a useful way
5573 to define what should happen).
5575 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5576 are now processed in the same way as lines in the Exim configuration file.
5577 In particular, continuation lines are supported.
5579 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5581 15. Added -ti meaning -t -i.
5583 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5584 domains, and warn by logging if others are found.
5586 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5587 returned from GIFCONF when Exim is trying to find the list of interfaces on
5588 a host. The code in os.c has been modified to copy each ifreq to an aligned
5589 structure in all cases.
5591 Also, in some cases, the returned ifreq's were being copied to a 'struct
5592 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5593 means the last couple of bytes of an IPv6 address could be chopped if the
5594 ifreq contained only a normal sockaddr (14 bytes storage).
5596 18. Named domain lists were not supported in the hosts_treat_as_local option.
5597 An entry such as +xxxx was not recognized, and was treated as a literal
5600 19. Ensure that header lines added by a DATA ACL are included in the reject log
5601 if the ACL subsequently rejects the message.
5603 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5604 MD5 (which is deprecated).
5606 21. When testing a filter file using -bf, Exim was writing a message when it
5607 took the sender from a "From " line in the message, but it was not doing so
5608 when it took $return_path from a Return-Path: header line. It now does.
5610 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5611 with a valid header line field name (a series of printing characters
5612 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5614 23. Changed "disc" in the source to "disk" to conform to the documentation and
5615 the book and for uniformity.
5617 24. Ignore Sendmail's -Ooption=value command line item.
5619 25. When execve() failed while trying to run a command in a pipe transport,
5620 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5621 could be confused with a return value of 69 from the command itself. This
5622 has been changed to 127, the value the shell returns if it is asked to run
5623 a non-existent command. The wording for the related log line suggests a
5624 non-existent command as the problem.
5626 26. If received_header_text expands to an empty string, do not add a Received:
5627 header line to the message. (Well, it adds a token one on the spool, but
5628 marks it "old" so that it doesn't get used or transmitted.)
5630 27. Installed eximstats 1.28 (addition of -nt option).
5632 28. There was no check for failure on the call to getsockname() in the daemon
5633 code. This can fail if there is a shortage of resources on the system, with
5634 ENOMEM, for example. A temporary error is now given on failure.
5636 29. Contrary to the C standard, it seems that in some environments, the
5637 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5638 program. Exim now does this explicitly; it affects the formatting of
5639 timestamps using strftime().
5641 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5642 complaints. I've now initialized all the variables, to avoid this.
5644 32. Header lines added by a system filter were not being "seen" during
5645 transport-time rewrites.
5647 33. The info_callback() function passed to OpenSSL is set up with type void
5648 (*)(SSL *, int, int), as described somewhere. However, when calling the
5649 function (actually a macro) that sets it up, the type void(*)() is
5650 expected. I've put in a cast to prevent warnings from picky compilers.
5652 34. If a DNS black list lookup found a CNAME record, but there were no A
5653 records associated with the domain it pointed at, Exim crashed.
5655 35. If a DNS black list lookup returned more than one A record, Exim ignored
5656 all but the first. It now scans all returned addresses if a particular IP
5657 value is being sought. In this situation, the contents of the
5658 $dnslist_value variable are a list of all the addresses, separated by a
5661 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5662 to accept a host name and all its aliases if the forward lookup for any of
5663 them yielded the IP address of the incoming connection. Now it accepts only
5664 those names whose forward lookup yields the correct IP address. Any other
5665 names are discarded. This closes a loophole whereby a rogue DNS
5666 administrator could create reverse DNS records to break through a
5667 wildcarded host restriction in an ACL.
5669 37. If a user filter or a system filter that ran in a subprocess used any of
5670 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5671 the wrong values were passed to the pipe command ($thisaddress had the
5672 value of $0, $0 had the value of $1, etc). This bug was introduced by
5673 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5675 38. Improved the line breaking for long SMTP error messages from ACLs.
5676 Previously, if there was no break point between 40 and 75 characters, Exim
5677 left the rest of the message alone. Two changes have been made: (a) I've
5678 reduced the minimum length to 35 characters; (b) if it can't find a break
5679 point between 35 and 75 characters, it looks ahead and uses the first one
5680 that it finds. This may give the occasional overlong line, but at least the
5681 remaining text gets split now.
5683 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5684 file descriptors might be low, and that setting 1000 would always raise it.
5685 It turns out that in some environments, the limit is already over 1000 and
5686 that lowering it causes trouble. So now Exim takes care not to decrease it.
5688 40. When delivering a message, the value of $return_path is set to $sender_
5689 address at the start of routing (routers may change the value). By an
5690 oversight, this default was not being set up when an address was tested by
5691 -bt or -bv, which affected the outcome if any router or filter referred to
5694 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5695 log only when "message" or "log_message" are set. However, if one of the
5696 conditions was an address verification, or a call to a nested ACL, the
5697 messages generated by the underlying test were being passed through. This
5698 no longer happens. The underlying message is available in $acl_verify_
5699 message for both "message" and "log_message" expansions, so it can be
5700 passed through if needed.
5702 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5703 new expansion $bh_ to give the encoded byte string without charset
5704 translation. Translation happens only if iconv() is available; HAVE_ICONV
5705 indicates this at build time. HEADERS_CHARSET gives the charset to
5706 translate to; headers_charset can change it in the configuration, and
5707 "headers charset" can change it in an individual filter file.
5709 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5710 that creates RFC 2047 encoded "words" labels them as that charset instead
5711 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5712 expansion operator; (ii) when Exim creates a From: line for a local
5713 message; (iii) when a header line is rewritten to include a "phrase" part.
5715 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5716 buggy, causing it to skip the first lines of messages whose message ID
5717 ended in 'D'. This would not have bitten before Exim release 4.14, because
5718 message IDs were unlikely to end in 'D' before then. The effect was to have
5719 incorrect size information for certain domains.
5721 45. #include "config.h" was missing at the start of the crypt16.c module. This
5722 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5725 46. If there was a timeout during a "random" callout check, Exim treated it as
5726 a failure of the random address, and carried on sending RSET and the real
5727 address. If the delay was just some slowness somewhere, the response to the
5728 original RCPT would be taken as a response to RSET and so on, causing
5729 mayhem of various kinds.
5731 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5732 when I implemented it. It didn't allow for the fact that some option values
5733 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5734 the right test for positive values.
5736 48. Domain names in DNS records are case-independent. Exim always looks them up
5737 in lower case. Some resolvers return domain names in exactly the case they
5738 appear in the zone file, that is, they may contain uppercase letters. Not
5739 all resolvers do this - some return always lower case. Exim was treating a
5740 change of case by a resolver as a change of domain, similar to a widening
5741 of a domain abbreviation. This triggered its re-routing code and so it was
5742 trying to route what was effectively the same domain again. This normally
5743 caused routing to fail (because the router wouldn't handle the domain
5744 twice). Now Exim checks for this case specially, and just changes the
5745 casing of the domain that it ultimately uses when it transmits the message
5748 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5751 50. If a filter generated a file delivery with a non-absolute name (possible if
5752 no home directory exists for the router), the forbid_file option was not
5755 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5756 the existing equality matching.
5758 52. Exim was using ints instead of ino_t variables in some places where it was
5759 dealing with inode numbers.
5761 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5762 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5763 called TMPDIR, and if it finds it is different, it changes its value.
5765 54. The smtp_printf() function is now made available to local_scan() so
5766 additional output lines can be written before returning. There is also an
5767 smtp_fflush() function to enable the detection of a dropped connection.
5768 The variables smtp_input and smtp_batched_input are exported to
5771 55. Changed the default runtime configuration: the message "Unknown user"
5772 has been removed from the ACL, and instead placed on the localuser router,
5773 using the cannot_route_message feature. This means that any verification
5774 failures that generate their own messages won't get overridden. Similarly,
5775 the "Unrouteable address" message that was in the ACL for unverifiable
5776 relay addresses has also been removed.
5778 56. Added hosts_avoid_esmtp to the smtp transport.
5780 57. The exicyclog script was not checking for the esoteric option
5781 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5782 will work only if exicyclog is run under the appropriate euid.
5784 58. Following a discussion on the list, the rules by which Exim recognises line
5785 endings on incoming messages have been changed. The -dropcr and drop_cr
5786 options are now no-ops, retained only for backwards compatibility. The
5787 following line terminators are recognized: LF CRLF CR. However, special
5788 processing applies to CR:
5790 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5791 nor a local message in the state where . is a terminator.
5793 (ii) If a bare CR is encountered in a header line, an extra space is added
5794 after the line terminator so as not to end the header. The reasoning
5795 behind this is that bare CRs in header lines are most likely either
5796 to be mistakes, or people trying to play silly games.
5798 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5799 was being incorrectly given as 18 bytes larger than it should have been.
5800 This is a VOB (very old bug).
5802 60. This may never have affected anything current, but just in case it has:
5803 When the local host is found other than at the start of a list of hosts,
5804 the local host, those with the same MX, and any that follow, are discarded.
5805 When the list in question was part of a longer list of hosts, the following
5806 hosts (not currently being processed) were also being discarded. This no
5807 longer happens. I'm not sure if this situation could ever has previously
5810 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5813 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5814 header. This is something that is recommended in a new Internet Draft, and
5815 is something that is documented as being done by Sendmail. There are two
5816 possible values. For messages generated by the autoreply transport, Exim
5819 Auto-Submitted: auto-replied
5821 whereas for all other generated messages (e.g. bounces) it adds
5823 Auto-Submitted: auto-generated
5825 63. The "personal" condition in filters now includes a test for the
5826 Auto-Submitted: header. If it contains the string "auto-" the message it
5827 not considered personal.
5829 64. Added rcpt_include_affixes as a generic transport option.
5831 65. Added queue_only_override (default true).
5833 66. Added the syslog_duplication option.
5835 67. If what should have been the first header line of a message consisted of
5836 a space followed by a colon, Exim was mis-interpreting it as a header line.
5837 It isn't of course - it is syntactically invalid and should therefore be
5838 treated as the start of the message body. The misbehaviour could have
5839 caused a number of strange effects, including loss of data in subsequent
5840 header lines, and spool format errors.
5842 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5843 client host had authenticated. This control can now be exercised by an ACL
5844 for more flexibility.
5846 69. By default, callouts do not happen when testing with -bh. There is now a
5847 variant, -bhc, which does actually run the callout code, including
5848 consulting and updating the callout cache.
5850 70. Added support for saslauthd authentication, courtesy of Alexander
5853 71. If statvfs() failed on the spool or log directories while checking their
5854 size for availability, Exim confusingly gave the error "space shortage".
5855 Furthermore, in debugging mode it crashed with a floating point exception.
5856 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5857 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5858 serious problem, Exim now writes to the main and panic logs when this
5859 happens, with details of the failure. It then refuses to accept the
5860 incoming message, giving the message "spool directory problem" or "log
5861 directory problem" with a 421 code for SMTP messages.
5863 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5864 0, 1, and 2 exist, because some OS complain for execs without them (see
5865 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5866 descriptors. However, the code omitted to check that the open succeeded,
5867 causing mysterious errors if for some reason the permissions on /dev/null
5868 got screwed. Now Exim writes a message to the main and panic logs, and
5869 bombs out if it can't open /dev/null.
5871 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5872 interact so that it is all more flexible. It is supposed to remain
5873 backwards compatible. Also added extra_local_interfaces.
5875 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5876 to bomb out with an assertion failure - to the client this appears as a
5877 connection drop. This problem occurs in the part of the code that was taken
5878 from the Samba project. Fortunately, the assertion is in a very simple
5879 function, so I have fixed this by reproducing the function inline in the
5880 one place where it is called, and arranging for authentication to fail
5881 instead of killing the process with assert().
5883 75. The SPA client code was not working when the server requested OEM rather
5884 than Unicode encoding.
5886 76. Added code to make require_files with a specific uid setting more usable in
5887 the case where statting the file as root fails - usually a non-root-mounted
5888 NFS file system. When this happens and the failure is EACCES, Exim now
5889 forks a subprocess and does the per-uid checking as the relevant uid.
5891 77. Added process_log_path.
5893 78. If log_file_path was not explicitly set, a setting of check_log_space or
5894 check_log_inodes was ignored.
5896 79. If a space check for the spool or log partitions fails, the incident is now
5897 logged. Of course, in the latter case the data may get lost...
5899 80. Added the %p formatting code to string_format() so that it can be used to
5900 print addresses in debug_print(). Adjusted all the address printing in the
5901 debugging in store.c to use %p rather than %d.
5903 81. There was a concern that a line of code in smtp_in.c could overflow a
5904 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5905 initially expressed, the concern was not well-founded, because trailing
5906 spaces are removed early. However, if the trailing spaces were followed by
5907 a NULL, they did not get removed, so the overflow was possible. Two fixes
5910 (a) I re-wrote the offending code in a cleaner fashion.
5911 (b) If an incoming SMTP command contains a NULL character, it is rejected
5914 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5915 runs initgroups(), so that if the Exim user is in any additional groups,
5916 they will be used during message reception.
5922 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob