1 Change log file for Exim from version 4.21
2 -------------------------------------------
8 TF/01 Correctly close the server side of TLS when forking for delivery.
10 When a message was received over SMTP with TLS, Exim failed to clear up
11 the incoming connection properly after forking off the child process to
12 deliver the message. In some situations the subsequent outgoing
13 delivery connection happened to have the same fd number as the incoming
14 connection previously had. Exim would try to use TLS and fail, logging
15 a "Bad file descriptor" error.
17 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
18 utilities have not been installed.
20 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
21 temporary space as the ACL may create new global variables.
23 TL/01 LDAP support uses per connection or global context settings, depending
24 upon the detected version of the libraries at build time.
26 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
27 to extract and use the src ip:port in logging and expansions as if it
28 were a direct connection from the outside internet.
34 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
36 PP/02 Make -n do something, by making it not do something.
37 When combined with -bP, the name of an option is not output.
39 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
42 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
43 $sender_host_name and config options to manage this, and basic check
46 PP/05 DSCP support for outbound connections and control modifier for inbound.
48 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
49 (Only plugin which currently uses this is kerberos4, which nobody should
50 be using, but we should make it available and other future plugins might
51 conceivably use it, even though it would break NAT; stuff *should* be
52 using channel bindings instead).
54 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
55 name; added for Sendmail compatibility; requires admin caller.
56 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
57 just ignore it); requires trusted caller.
58 Also parse but ignore: -Ac -Am -X<logfile>
61 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
63 TL/02 Add +smtp_confirmation as a default logging option.
65 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
66 Patch by Magnus Holmgren from 2007-02-20.
68 TL/04 Bugzilla 1281 - Spec typo.
69 Bugzilla 1283 - Spec typo.
70 Bugzilla 1290 - Spec grammar fixes.
72 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
74 TL/06 Add Experimental DMARC support using libopendmarc libraries.
76 TL/07 Fix an out of order global option causing a segfault. Reported to dev
77 mailing list by by Dmitry Isaikin.
79 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
81 JH/02 Support "G" suffix to numbers in ${if comparisons.
83 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
85 NM/01 Bugzilla 1197 - Spec typo
86 Bugzilla 1196 - Spec examples corrections
88 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
90 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
91 gnutls_enable_pkcs11, but renamed to more accurately indicate its
94 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
95 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
97 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
98 "acl {{name}{arg}...}", and optional args on acl condition
101 JH/05 Permit multiple router/transport headers_add/remove lines.
103 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
105 JH/07 Avoid using a waiting database for a single-message-only transport.
106 Performance patch from Paul Fisher. Bugzilla 1262.
108 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
111 JH/09 Add $headers_added variable, with content from use of ACL modifier
112 add_header (but not yet added to the message). Bugzilla 199.
114 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
115 Pulled from Bugzilla 817 by Wolfgang Breyha.
117 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
119 (nb: this is the same fix as in Exim 4.80.1)
121 JH/11 Add A= logging on delivery lines, and a client_set_id option on
124 JH/12 Add optional authenticated_sender logging to A= and a log_selector
127 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
129 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
130 advertise SMTP AUTH mechanism to us, instead of a generic
131 protocol violation error. Also, make Exim more robust to bad
132 data from the Dovecot auth socket.
134 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
136 When a queue runner is handling a message, Exim first routes the
137 recipient addresses, during which it prunes them based on the retry
138 hints database. After that it attempts to deliver the message to
139 any remaining recipients. It then updates the hints database using
142 So if a recipient address works intermittently, it can get repeatedly
143 deferred at routing time. The retry hints record remains fresh so the
144 address never reaches the final cutoff time.
146 This is a fairly common occurrence when a user is bumping up against
147 their storage quota. Exim had some logic in its local delivery code
148 to deal with this. However it did not apply to per-recipient defers
149 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
151 This change adds a proper retry rule check during routing so that the
152 final cutoff time is checked against the message's age. We only do
153 this check if there is an address retry record and there is not a
154 domain retry record; this implies that previous attempts to handle
155 the address had the retry_use_local_parts option turned on. We use
156 this as an approximation for the destination being like a local
157 delivery, as in LMTP.
159 I suspect this new check makes the old local delivery cutoff check
160 redundant, but I have not verified this so I left the code in place.
162 TF/02 Correct gecos expansion when From: is a prefix of the username.
164 Test 0254 submits a message to Exim with the header
168 When I ran the test suite under the user fanf2, Exim expanded
169 the header to contain my full name, whereas it should have added
170 a Resent-Sender: header. It erroneously treats any prefix of the
171 username as equal to the username.
173 This change corrects that bug.
175 GF/01 DCC debug and logging tidyup
176 Error conditions log to paniclog rather than rejectlog.
177 Debug lines prefixed by "DCC: " to remove any ambiguity.
179 TF/03 Avoid unnecessary rebuilds of lookup-related code.
181 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
182 Bug spotted by Jeremy Harris; was flawed since initial commit.
183 Would have resulted in OCSP responses post-SNI triggering an Exim
184 NULL dereference and crash.
186 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
188 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
189 Bug detection, analysis and fix by Samuel Thibault.
190 Bugzilla 1331, Debian bug #698092.
192 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
194 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
195 Server implementation by Todd Lyons, client by JH.
196 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
197 config variable "prdr_enable" controls whether the server
198 advertises the facility. If the client requests PRDR a new
199 acl_data_smtp_prdr ACL is called once for each recipient, after
200 the body content is received and before the acl_smtp_data ACL.
201 The client is controlled by bolth of: a hosts_try_prdr option
202 on the smtp transport, and the server advertisement.
203 Default client logging of deliveries and rejections involving
204 PRDR are flagged with the string "PRDR".
206 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
207 fclose(). Diagnosis by Todd Lyons.
209 PP/17 Update configure.default to handle IPv6 localhost better.
210 Patch by Alain Williams (plus minor tweaks).
213 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
214 This is now consistent with GnuTLS, and is now documented: the
215 previous undocumented portable approach to treating the option as
216 unset was to force an expansion failure. That still works, and
217 an empty string is now equivalent.
219 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
220 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
221 not performing validation itself.
223 PP/20 Added force_command boolean option to pipe transport.
224 Patch from Nick Koston, of cPanel Inc.
226 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
229 TF/04 Added udpsend ACL modifer and hexquote expansion operator
231 PP/21 Fix eximon continuous updating with timestamped log-files.
232 Broken in a format-string cleanup in 4.80, missed when I repaired the
233 other false fix of the same issue.
234 Report and fix from Heiko Schlichting.
237 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
238 Report from Prashanth Katuri.
240 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
241 It's SecureTransport, so affects any MacOS clients which use the
242 system-integrated TLS libraries, including email clients.
244 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
245 using a MIME ACL for non-SMTP local injection.
246 Report and assistance in diagnosis by Warren Baker.
248 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
250 JH/16 Fix comparisons for 64b. Bugzilla 1385.
252 TL/09 Add expansion variable $authenticated_fail_id to keep track of
253 last id that failed so it may be referenced in subsequent ACL's.
255 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
258 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
259 ldap library initialization, allowing self-signed CA's to be
260 used. Also properly sets require_cert option later in code by
261 using NULL (global ldap config) instead of ldap handle (per
262 session). Bug diagnosis and testing by alxgomz.
264 TL/12 Enhanced documentation in the ratelimit.pl script provided in
265 the src/util/ subdirectory.
267 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
268 renamed to Transport Post Delivery Action by Jeremy Harris, as
271 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
272 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
273 redis_servers = needs to be configured which will be used by the redis
274 lookup. Patch from Warren Baker, of The Packet Hub.
276 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
278 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
279 hostname or reverse DNS when processing a host list. Used suggestions
280 from multiple comments on this bug.
282 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
284 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
285 Missed a few lines, added it to make the runtest require no keyboard
288 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
289 contains upper case chars. Make router use caseful_local_part.
291 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
292 support when GnuTLS has been built with p11-kit.
298 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
300 This, or similar/improved, will also be change PP/11 of 4.82.
306 PP/01 Handle short writes when writing local log-files.
307 In practice, only affects FreeBSD (8 onwards).
308 Bugzilla 1053, with thanks to Dmitry Isaikin.
310 NM/01 Bugzilla 949 - Documentation tweak
312 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
315 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
317 PP/02 Implemented gsasl authenticator.
319 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
321 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
322 `pkg-config foo` for cflags/libs.
324 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
325 with rest of GSASL and with heimdal_gssapi.
327 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
328 `pkg-config foo` for cflags/libs for the TLS implementation.
330 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
331 properties get this fed in as external SSF. A number of robustness
332 and debugging improvements to the cyrus_sasl authenticator.
334 PP/08 cyrus_sasl server now expands the server_realm option.
336 PP/09 Bugzilla 1214 - Log authentication information in reject log.
337 Patch by Jeremy Harris.
339 PP/10 Added dbmjz lookup type.
341 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
343 PP/12 MAIL args handles TAB as well as SP, for better interop with
344 non-compliant senders.
345 Analysis and variant patch by Todd Lyons.
347 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
348 Bug report from Lars Müller <lars@samba.org> (via SUSE),
349 Patch from Dirk Mueller <dmueller@suse.com>
351 PP/13 tls_peerdn now print-escaped for spool files.
352 Observed some $tls_peerdn in wild which contained \n, which resulted
353 in spool file corruption.
355 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
356 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
357 or write after TLS renegotiation, which otherwise led to messages
360 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
361 as a tracking header (ie: a signed header comes before the signature).
362 Patch from Wolfgang Breyha.
364 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
365 comma-sep list; embedded commas doubled.
367 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
369 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
371 Report and patch from Dmitry Banschikov.
373 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
374 Removed SSL_clear() after SSL_new() which led to protocol negotiation
375 failures. We appear to now support TLS1.1+ with Exim.
377 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
378 lets Exim select keys and certificates based upon TLS SNI from client.
379 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
380 before an outbound SMTP session. New log_selector, +tls_sni.
382 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
383 NULL dereference. Report and patch from Alun Jones.
385 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
386 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
387 Not seeing resolver debug output on NetBSD, but suspect this is a
388 resolver implementation change.
390 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
391 Left warnings. Added "eximon gdb" invocation mode.
393 PP/21 Defaulting "accept_8bitmime" to true, not false.
395 PP/22 Added -bw for inetd wait mode support.
397 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
398 locate the relevant includes and libraries. Made this the default.
400 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
401 Bugzilla 1246, report and most of solution from Tomasz Kusy.
403 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
404 This may cause build issues on older platforms.
406 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
407 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
408 gnutls_require_mac & gnutls_require_protocols (no longer supported).
409 Added SNI support via GnuTLS too.
410 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
412 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
414 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
415 Applied second patch from Janne, implementing suggestion to default
416 multiple-strings-in-record handling to match SPF spec.
418 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
420 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
421 read-only, out of scope).
422 Patch from Wolfgang Breyha, report from Stuart Northfield.
424 PP/29 Fix three issues highlighted by clang analyser static analysis.
425 Only crash-plausible issue would require the Cambridge-specific
426 iplookup router and a misconfiguration.
427 Report from Marcin Mirosław.
429 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
431 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
432 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
433 As part of this, removing so much warning spew let me fix some minor
434 real issues in debug logging.
436 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
437 assignment on my part. Fixed.
439 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
440 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
441 Janne Snabb (who went above and beyond: thank you).
443 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
444 string otherwise requires a connection and a bunch more work and it's
445 relatively easy to get wrong. Should also expose TLS library linkage
448 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
449 64-bit ${eval} (JH/03).
451 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
452 GNU libc to support some of the 64-bit stuff, should not lead to
453 conflicts. Defined before os.h is pulled in, so if a given platform
454 needs to override this, it can.
456 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
457 protection layer was required, which is not implemented.
458 Bugzilla 1254, patch from Wolfgang Breyha.
460 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
461 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
462 tls_dhparam take prime identifiers. Also unbreak combination of
463 OpenSSL+DH_params+TLSSNI.
465 PP/39 Disable SSLv2 by default in OpenSSL support.
471 PP/01 Solaris build fix for Oracle's LDAP libraries.
472 Bugzilla 1109, patch from Stephen Usher.
474 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
476 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
479 TF/02 Fix a couple more cases where we did not log the error message
480 when unlink() failed. See also change 4.74-TF/03.
482 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
483 lock up or crash if it happened to be inside a call to libc when it
484 got a SIGUSR1 from exiwhat.
486 The SIGUSR1 handler appends the current process status to the process
487 log which is later printed by exiwhat. It used to use the general
488 purpose logging code to do this, but several functions it calls are
489 not safe for signals.
491 The new output code in the SIGUSR1 handler is specific to the process
492 log, and simple enough that it's easy to inspect for signal safety.
493 Removing some special cases also simplifies the general logging code.
494 Removing the spurious timestamps from the process log simplifies
497 TF/04 Improved ratelimit ACL condition.
499 The /noupdate option has been deprecated in favour of /readonly which
500 has clearer semantics. The /leaky, /strict, and /readonly update modes
501 are mutually exclusive. The update mode is no longer included in the
502 database key; it just determines when the database is updated. (This
503 means that when you upgrde Exim will forget old rate measurements.)
505 Exim now checks that the per_* options are used with an update mode that
506 makes sense for the current ACL. For example, when Exim is processing a
507 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
508 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
509 must specify per_mail/readonly. If you omit the update mode it defaults to
510 /leaky where that makes sense (as before) or /readonly where required.
512 The /noupdate option is now undocumented but still supported for
513 backwards compatibility. It is equivalent to /readonly except that in
514 ACLs where /readonly is required you may specify /leaky/noupdate or
515 /strict/noupdate which are treated the same as /readonly.
517 A useful new feature is the /count= option. This is a generalization
518 of the per_byte option, so that you can measure the throughput of other
519 aggregate values. For example, the per_byte option is now equivalent
520 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
522 The per_rcpt option has been generalized using the /count= mechanism
523 (though it's more complicated than the per_byte equivalence). When it is
524 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
525 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
526 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
527 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
528 Note that using per_rcpt with a non-readonly update mode in more than
529 one ACL will cause the recipients to be double-counted. (The per_mail
530 and per_byte options don't have this problem.)
532 The handling of very low rates has changed slightly. If the computed rate
533 is less than the event's count (usually one) then this event is the first
534 after a long gap. In this case the rate is set to the same as this event's
535 count, so that the first message of a spam run is counted properly.
537 The major new feature is a mechanism for counting the rate of unique
538 events. The new per_addr option counts the number of different
539 recipients that someone has sent messages to in the last time period. It
540 behaves like per_rcpt if all the recipient addresses are different, but
541 duplicate recipient addresses do not increase the measured rate. Like
542 the /count= option this is a general mechanism, so the per_addr option
543 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
544 example, measure the rate that a client uses different sender addresses
545 with the options per_mail/unique=$sender_address. There are further
546 details in the main documentation.
548 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
550 TF/06 Removed a few PCRE remnants.
552 TF/07 Automatically extract Exim's version number from tags in the git
553 repository when doing development or release builds.
555 PP/02 Raise smtp_cmd_buffer_size to 16kB.
556 Bugzilla 879. Patch from Paul Fisher.
558 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
559 Heavily based on revision 40f9a89a from Simon Arlott's tree.
562 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
564 PP/05 Variable $av_failed, true if the AV scanner deferred.
565 Bugzilla 1078. Patch from John Horne.
567 PP/06 Stop make process more reliably on build failure.
568 Bugzilla 1087. Patch from Heiko Schlittermann.
570 PP/07 Make maildir_use_size_file an _expandable_ boolean.
571 Bugzilla 1089. Patch from Heiko Schlittermann.
573 PP/08 Handle ${run} returning more data than OS pipe buffer size.
574 Bugzilla 1131. Patch from Holger Weiß.
576 PP/09 Handle IPv6 addresses with SPF.
577 Bugzilla 860. Patch from Wolfgang Breyha.
579 PP/10 GnuTLS: support TLS 1.2 & 1.1.
581 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
584 PP/11 match_* no longer expand right-hand-side by default.
585 New compile-time build option, EXPAND_LISTMATCH_RHS.
586 New expansion conditions, "inlist", "inlisti".
588 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
590 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
592 PP/14 fix log_write() format string regression from TF/03.
593 Bugzilla 1152. Patch from Dmitry Isaikin.
599 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
601 PP/02 Harmonised TLS library version reporting; only show if debugging.
602 Layout now matches that introduced for other libraries in 4.74 PP/03.
604 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
606 PP/04 New "dns_use_edns0" global option.
608 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
611 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
612 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
614 TK/01 Updated PolarSSL code to 0.14.2.
615 Bugzilla 1097. Patch from Andreas Metzler.
617 PP/07 Catch divide-by-zero in ${eval:...}.
620 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
623 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
624 format-string attack -- SECURITY: remote arbitrary code execution.
626 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
627 time unintentionally subject to list matching rules, letting the header
628 cause arbitrary Exim lookups (of items which can occur in lists, *not*
629 arbitrary string expansion). This allowed for information disclosure.
631 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
632 INT_MIN/-1 -- value coerced to INT_MAX.
638 NM/01 Workround for PCRE version dependency in version reporting
641 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
642 This fixes portability to compilers other than gcc, notably
643 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
645 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
646 makefiles for portability to HP-UX and POSIX correctness.
648 PP/01 Permit LOOKUP_foo enabling on the make command-line.
649 Also via indented variable definition in the Makefile.
650 (Debugging by Oliver Heesakkers).
652 PP/02 Restore caching of spamd results with expanded spamd_address.
653 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
655 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
656 Improves build reliability. Fix from: Frank Elsner
658 NM/02 Fix wide character breakage in the rfc2047 coding
659 Fixes bug 1064. Patch from Andrey N. Oktyabrski
661 NM/03 Allow underscore in dnslist lookups
662 Fixes bug 1026. Patch from Graeme Fowler
664 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
665 Code patches from Adam Ciarcinski of NetBSD.
667 NM/04 Fixed exiqgrep to cope with mailq missing size issue
670 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
671 is logged, to avoid truncation. Patch from John Horne.
673 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
674 Patch from Jakob Hirsch.
676 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
677 SQL string expansion failure details.
678 Patch from Andrey Oktyabrski.
680 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
681 Patch from Simon Arlott.
683 PP/09 New lookups functionality failed to compile on old gcc which rejects
684 extern declarations in function scope.
685 Patch from Oliver Fleischmann
687 PP/10 Use sig_atomic_t for flags set from signal handlers.
688 Check getgroups() return and improve debugging.
689 Fixed developed for diagnosis in bug 927 (which turned out to be
692 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
693 Patch from Mark Zealey.
695 PP/12 Bugzilla 1056: Improved spamd server selection.
696 Patch from Mark Zealey.
698 PP/13 Bugzilla 1086: Deal with maildir quota file races.
699 Based on patch from Heiko Schlittermann.
701 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
702 Patch from Uwe Doering, sign-off by Michael Haardt.
704 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
705 variable declaration deep within a block. Bug and patch from
708 PP/15 lookups-Makefile IRIX compatibilty coercion.
710 PP/16 Make DISABLE_DKIM build knob functional.
712 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
713 Patch by Simon Arlott
715 TF/03 Fix valgrind.h portability to C89 compilers that do not support
716 variable argument macros. Our copy now differs from upstream.
722 TF/01 Failure to get a lock on a hints database can have serious
723 consequences so log it to the panic log.
725 TF/02 Log LMTP confirmation messages in the same way as SMTP,
726 controlled using the smtp_confirmation log selector.
728 TF/03 Include the error message when we fail to unlink a spool file.
730 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
731 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
732 for maintaining out-of-tree patches for some time.
734 PP/01 Bugzilla 139: Documentation and portability issues.
735 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
736 Handle per-OS dynamic-module compilation flags.
738 PP/02 Let /dev/null have normal permissions.
739 The 4.73 fixes were a little too stringent and complained about the
740 permissions on /dev/null. Exempt it from some checks.
741 Reported by Andreas M. Kirchwitz.
743 PP/03 Report version information for many libraries, including
744 Exim version information for dynamically loaded libraries. Created
745 version.h, now support a version extension string for distributors
746 who patch heavily. Dynamic module ABI change.
748 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
749 privilege escalation vulnerability whereby the Exim run-time user
750 can cause root to append content of the attacker's choosing to
753 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
756 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
757 If dropping privileges for untrusted macros, we disabled normal logging
758 on the basis that it would fail; for the Exim run-time user, this is not
759 the case, and it resulted in successful deliveries going unlogged.
760 Fixed. Reported by Andreas Metzler.
766 PP/01 Date: & Message-Id: revert to normally being appended to a message,
767 only prepend for the Resent-* case. Fixes regression introduced in
768 Exim 4.70 by NM/22 for Bugzilla 607.
770 PP/02 Include check_rfc2047_length in configure.default because we're seeing
771 increasing numbers of administrators be bitten by this.
773 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
775 PP/03 Bugzilla 994: added openssl_options main configuration option.
777 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
779 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
781 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
783 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
784 without a peer certificate, leading to a segfault because of an
785 assumption that peers always have certificates. Be a little more
786 paranoid. Problem reported by Martin Tscholak.
788 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
789 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
790 NB: ClamAV planning to remove STREAM in "middle of 2010".
791 CL also introduces -bmalware, various -d+acl logging additions and
792 more caution in buffer sizes.
794 PP/09 Implemented reverse_ip expansion operator.
796 PP/10 Bugzilla 937: provide a "debug" ACL control.
798 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
800 PP/12 Bugzilla 973: Implement --version.
802 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
804 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
806 PP/15 Bugzilla 816: support multiple condition rules on Routers.
808 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
809 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
810 ignore trailing whitespace.
812 JJ/02 prevent non-panic DKIM error from being sent to paniclog
814 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
817 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
818 Notification from Dr Andrew Aitchison.
820 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
821 ExtendedDetectionInfo response format.
822 Notification from John Horne.
824 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
827 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
828 XSL and documented dependency on system catalogs, with examples of how
831 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
834 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
835 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
836 configuration file which is writeable by the Exim user or group.
838 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
839 of configuration files to cover files specified with the -C option if
840 they are going to be used with root privileges, not just the default
843 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
844 option (effectively making it always true).
846 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
847 files to be used while preserving root privileges.
849 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
850 that rogue child processes cannot use them.
852 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
853 run-time user, instead of root.
855 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
856 Exim run-time user without dropping privileges.
858 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
859 result string, instead of calling string_vformat() twice with the same
862 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
863 for other users. Others should always drop root privileges if they use
864 -C on the command line, even for a whitelisted configure file.
866 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
868 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
874 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
875 $data_path, and $header_path variables; fixed documentation bugs and
878 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
879 exipick to access non-standard spools, including the "frozen" queue
882 NM/01 Bugzilla 965: Support mysql stored procedures.
883 Patch from Alain Williams
885 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
887 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
888 Patch from Andreas Metzler
890 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
891 Patch from Kirill Miazine
893 NM/05 Bugzilla 671: Added umask to procmail example.
895 JJ/03 installed exipick 20100323.0, fixing doc bug
897 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
898 directory. Notification and patch from Dan Rosenberg.
900 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
902 TK/02 Improve log output when DKIM signing operation fails.
904 MH/01 Treat the transport option dkim_domain as a colon separated
905 list, not as a single string, and sign the message with each element,
906 omitting multiple occurences of the same signer.
908 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
909 Bugzilla 985, 986. Patch by Simon Arlott
911 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
912 Patch by Simon Arlott
914 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
915 MBX locking. Notification from Dan Rosenberg.
921 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
923 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
925 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
927 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
929 NM/04 Bugzilla 915: Flag broken perl installation during build.
935 TK/01 Added patch by Johannes Berg that expands the main option
936 "spamd_address" if it starts with a dollar sign.
938 TK/02 Write list of recipients to X-Envelope-Sender header when building
939 the mbox-format spool file for content scanning (suggested by Jakob
942 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
943 (http://www.dcc-servers.net/) support via dccifd. Activated by
944 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
946 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
947 by Mark Daniel Reidel <mr@df.eu>.
949 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
950 When building exim an external PCRE library is now needed -
951 PCRE is a system library on the majority of modern systems.
952 See entry on PCRE_LIBS in EDITME file.
954 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
955 conversation. Added nologin parameter to request.
956 Patch contributed by Kirill Miazine.
958 TF/01 Do not log submission mode rewrites if they do not change the address.
960 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
962 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
963 log files in place. Contributed by Roberto Lima.
965 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
967 TF/03 Bugzilla 615: When checking the local_parts router precondition
968 after a local_part_suffix or local_part_prefix option, Exim now
969 does not use the address's named list lookup cache, since this
970 contains cached lookups for the whole local part.
972 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
973 Robert Millan. Documentation is in experimental-spec.txt.
975 TF/04 Bugzilla 668: Fix parallel build (make -j).
977 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
979 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
980 Patch provided by Jan Srzednicki.
982 TF/05 Leading white space used to be stripped from $spam_report which
983 wrecked the formatting. Now it is preserved.
985 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
986 that they are available at delivery time.
988 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
990 TF/08 TLS error reporting now respects the incoming_interface and
991 incoming_port log selectors.
993 TF/09 Produce a more useful error message if an SMTP transport's hosts
994 setting expands to an empty string.
996 NM/06 Bugzilla 744: EXPN did not work under TLS.
997 Patch provided by Phil Pennock.
999 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1000 Patch provided by Richard Godbee.
1002 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1003 acl_smtp_notquit, added index entry.
1005 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1006 Patch provided by Eugene Bujak.
1008 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1009 accept(). Patch provided by Maxim Dounin.
1011 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1012 Patch provided by Phil Pennock.
1014 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1016 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1017 Patch provided by Brad "anomie" Jorsch.
1019 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1020 Patch provided by Dean Brooks.
1022 TK/05 Add native DKIM support (does not depend on external libraries).
1024 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1025 Patch provided by Graeme Fowler.
1027 NM/16 Bugzilla 851: Documentation example syntax fix.
1029 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1031 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1034 NM/19 Bugzilla 745: TLS version reporting.
1035 Patch provided by Phil Pennock.
1037 NM/20 Bugzilla 167: bool: condition support.
1038 Patch provided by Phil Pennock.
1040 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1041 clients. Patch provided by Phil Pennock.
1043 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1044 Patch provided by Brad "anomie" Jorsch.
1046 NM/23 Bugzilla 687: Fix misparses in eximstats.
1047 Patch provided by Heiko Schlittermann.
1049 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1050 Patch provided by Heiko Schlittermann.
1052 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1053 plus update to original patch.
1055 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1057 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1058 Patch provided by David Brownlee.
1060 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1062 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1064 NM/30 Bugzilla 888: TLS documentation bugfixes.
1066 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1068 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1069 Unlike the original bugzilla I have changed all shell scripts in src tree.
1071 NM/33 Bugzilla 898: Transport filter timeout fix.
1072 Patch by Todd Rinaldo.
1074 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1075 Patch by Serge Demonchaux.
1077 NM/35 Bugzilla 39: Base64 decode bug fixes.
1078 Patch by Jakob Hirsch.
1080 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1082 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1084 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1086 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1092 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1093 ALT-N's libdkim that I have put here:
1094 http://duncanthrax.net/exim-experimental/
1096 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1097 were called 'true' and it seems that C99 defines that as a reserved
1098 keyword to be used with 'bool' variable types. That means you could
1099 not include C99-style headers which use bools without triggering
1100 build errors in sieve.c.
1102 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1103 as mailq or other aliases. Changed the --help handling significantly
1104 to do whats expected. exim_usage() emits usage/help information.
1106 SC/01 Added the -bylocaldomain option to eximstats.
1108 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1110 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1112 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1118 PH/01 Another patch from the Sieve maintainer.
1120 PH/02 When an IPv6 address is converted to a string for single-key lookup
1121 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1122 dots are used instead of colons so that keys in lsearch files need not
1123 contain colons. This was done some time before quoting was made available
1124 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1125 (notated using the quote facility) so as to distinguish them from IPv4
1126 keys. This meant that lookups for IP addresses in host lists did not work
1127 for iplsearch lookups.
1129 This has been fixed by arranging for IPv6 addresses to be expressed with
1130 colons if the lookup type is iplsearch. This is not incompatible, because
1131 previously such lookups could never work.
1133 The situation is now rather anomolous, since one *can* have colons in
1134 ordinary lsearch keys. However, making the change in all cases is
1135 incompatible and would probably break a number of configurations.
1137 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1140 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1141 conversion specification without a maximum field width, thereby enabling
1142 a rogue spamd server to cause a buffer overflow. While nobody in their
1143 right mind would setup Exim to query an untrusted spamd server, an
1144 attacker that gains access to a server running spamd could potentially
1145 exploit this vulnerability to run arbitrary code as the Exim user.
1147 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1148 $primary_hostname instead of what libspf2 thinks the hosts name is.
1150 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1151 a directory entry by the name of the lookup key. Previously, if a
1152 symlink pointed to a non-existing file or a file in a directory that
1153 Exim lacked permissions to read, a lookup for a key matching that
1154 symlink would fail. Now it is enough that a matching directory entry
1155 exists, symlink or not. (Bugzilla 503.)
1157 PH/03 The body_linecount and body_zerocount variables are now exported in the
1160 PH/04 Added the $dnslist_matched variable.
1162 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1163 This means they are set thereafter only if the connection becomes
1166 PH/06 Added the client_condition to authenticators so that some can be skipped
1167 by clients under certain conditions.
1169 PH/07 The error message for a badly-placed control=no_multiline_responses left
1170 "_responses" off the end of the name.
1172 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1174 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1175 (without spaces) instead of just copying the configuration text.
1177 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1179 PH/11 Added $max_received_linelength.
1181 PH/12 Added +ignore_defer and +include_defer to host lists.
1183 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1184 way in which PCRE > 7.0 is built.
1186 PH/14 Implemented queue_only_load_latch.
1188 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1189 MAIL command. The effect was to mangle the value on 64-bit systems.
1191 PH/16 Another patch from the Sieve maintainer.
1193 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1195 PH/18 If a system quota error occurred while trying to create the file for
1196 a maildir delivery, the message "Mailbox is full" was not appended to the
1197 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1198 only to a quota excession during the actual writing of the file.
1200 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1201 characters?) which causes problems in log lines. The DN values are now
1202 passed through string_printing() before being added to log lines.
1204 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1205 and InterBase are left for another time.)
1207 PH/21 Added message_body_newlines option.
1209 PH/22 Guard against possible overflow in moan_check_errorcopy().
1211 PH/23 POSIX allows open() to be a macro; guard against that.
1213 PH/24 If the recipient of an error message contained an @ in the local part
1214 (suitably quoted, of course), incorrect values were put in $domain and
1215 $local_part during the evaluation of errors_copy.
1221 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1222 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1225 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1226 issue a MAIL command.
1228 PH/02 In an ACL statement such as
1230 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1232 if a client was not listed at all, or was listed with a value other than
1233 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1234 the condition was not true (as it should be), so access was not denied.
1235 The bug was that the ! inversion was incorrectly passed on to the second
1236 item. This has been fixed.
1238 PH/03 Added additional dnslists conditions == and =& which are different from
1239 = and & when the dns lookup returns more than one IP address.
1241 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1242 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1244 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1245 FSYNC, which compiles an option called disable_fsync that allows for
1246 bypassing fsync(). The documentation is heavily laced with warnings.
1248 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1250 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1251 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1252 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1253 including adding "make clean"; (3) Added -fPIC when compiling the test
1254 dynamically loaded module, to get rid of a warning.
1256 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1257 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1258 = 0s. The bug is otherwise harmless.
1260 PH/07 There was a bug in the dovecot authenticator such that the value of
1261 $auth1 could be overwritten, and so not correctly preserved, after a
1262 successful authentication. This usually meant that the value preserved by
1263 the server_setid option was incorrect.
1265 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1267 PH/09 Installed PCRE release 7.0.
1269 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1270 run for batched SMTP input. It is now run at the start of every message
1271 in the batch. While fixing this I discovered that the process information
1272 (output by running exiwhat) was not always getting set for -bs and -bS
1273 input. This is fixed, and it now also says "batched" for BSMTP.
1275 PH/11 Added control=no_pipelining.
1277 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1278 patch, slightly modified), and move the expansion of helo_data till after
1279 the connection is made in the smtp transport (so it can use these
1282 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1284 PH/14 Added log_selector = +pid.
1286 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1288 PH/16 Add ${if forany and ${if forall.
1290 PH/17 Added dsn_from option to vary the From: line in DSNs.
1292 PH/18 Flush SMTP output before performing a callout, unless control =
1293 no_callout_flush is set.
1295 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1296 was true (the default) a successful delivery failed to delete the retry
1297 item, thus causing premature timeout of the address. The bug is now
1300 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1302 PH/21 Long custom messages for fakedefer and fakereject are now split up
1303 into multiline reponses in the same way that messages for "deny" and
1304 other ACL rejections are.
1306 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1307 with slight modification.
1309 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1310 draft, changing the syntax and factoring some duplicate code".
1312 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1313 for deliveries of the second and subsequent messages over the same SMTP
1316 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1317 ${reduce, with only minor "tidies".
1319 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1321 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1322 expansion side effects.
1324 PH/27 When a message times out after an over-quota error from an Exim-imposed
1325 quota, the bounce message says "mailbox is full". This message was not
1326 being given when it was a system quota that was exceeded. It now should
1329 MH/03 Made $recipients available in local_scan(). local_scan() already has
1330 better access to the recipient list through recipients_list[], but
1331 $recipients can be useful in postmaster-provided expansion strings.
1333 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1334 in the case of a MAIL command with additional options following the
1335 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1336 were accidentally chopped off.
1338 PH/29 SMTP synchronization checks are implemented when a command is read -
1339 there is a check that no more input is waiting when there shouldn't be
1340 any. However, for some commands, a delay in an ACL can mean that it is
1341 some time before the response is written. In this time, more input might
1342 arrive, invalidly. So now there are extra checks after an ACL has run for
1343 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1344 pipelining has not been advertised.
1346 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1348 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1349 returned in the SMTP response when smtp_return_error_details was false.
1350 This has been fixed.
1352 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1353 instead of the C I/O that was originally supplied, because problems were
1354 reported on Solaris.
1356 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1357 Exim which did not show up earlier: it was assuming that a call to
1358 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1359 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1360 was a macro that became an assignment, so it seemed to work. This has
1361 changed to a proper function call with a void return, hence the compile
1362 error. Exim's code has been fixed.
1364 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1367 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1369 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1371 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1372 because it turns out to be too big or there is a timeout) there is no
1373 "Completed" line in the log. When some messages of this type were
1374 selected by exigrep, they were listed as "not completed". Others were
1375 picked up by some special patterns. I have improved the selection
1376 criteria to be more general.
1378 PH/38 The host_find_failed option in the manualroute router can now be set
1379 to "ignore", to completely ignore a host whose IP address cannot be
1380 found. If all hosts are ignored, the behaviour is controlled by the new
1381 host_all_ignored option.
1383 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1384 homing or because of multiple MX records with /mx) generated more than
1385 one IP address, and the following item turned out to be the local host,
1386 all the secondary addresses of the first item were incorrectly removed
1387 from the list, along with the local host and any following hosts (which
1388 is what is supposed to happen).
1390 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1391 whoever called Exim into the -H file. In the case of the daemon it was
1392 behaving confusingly. When first started, it used values for whoever
1393 started the daemon, but after a SIGHUP it used the Exim user (because it
1394 calls itself on a restart). I have changed the code so that it now always
1397 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1398 message are rejected with the same error (e.g. no authentication or bad
1399 sender address), and a DATA command is nevertheless sent (as can happen
1400 with PIPELINING or a stupid MUA), the error message that was given to the
1401 RCPT commands is included in the rejection of the DATA command. This is
1402 intended to be helpful for MUAs that show only the final error to their
1405 PH/42 Another patch from the Sieve maintainer.
1407 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1408 Eximstats - Fixed some broken HTML links and added missing column headers
1410 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1411 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1413 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1415 PH/43 Yet another patch from the Sieve maintainer.
1417 PH/44 I found a way to check for a TCP/IP connection going away before sending
1418 the response to the final '.' that terminates a message, but only in the
1419 case where the client has not sent further data following the '.'
1420 (unfortunately, this is allowed). However, in many cases there won't be
1421 any further data because there won't be any more messages to send. A call
1422 to select() can be used: if it shows that the input is "ready", there is
1423 either input waiting, or the socket has been closed. An attempt to read
1424 the next input character can distinguish the two cases. Previously, Exim
1425 would have sent an OK response which the client would never have see.
1426 This could lead to message repetition. This fix should cure that, at
1427 least in a lot of common cases.
1429 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1430 advertised in response to EHLO.
1436 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1437 fixed by 4.65/MH/01 (is this a record?) are fixed:
1439 (i) An empty string was always treated as zero by the numeric comparison
1440 operators. This behaviour has been restored.
1442 (ii) It is documented that the numeric comparison operators always treat
1443 their arguments as decimal numbers. This was broken in that numbers
1444 starting with 0 were being interpreted as octal.
1446 While fixing these problems I realized that there was another issue that
1447 hadn't been noticed. Values of message_size_limit (both the global option
1448 and the transport option) were treated as octal if they started with 0.
1449 The documentation was vague. These values are now always treated as
1450 decimal, and I will make that clear in the documentation.
1456 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1457 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1460 MH/01 Don't check that the operands of numeric comparison operators are
1461 integers when their expansion is in "skipping" mode (fixes bug
1462 introduced by 4.64-PH/07).
1464 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1465 child addresses, Exim now panics and dies. Previously, because the count
1466 is held in a short int, deliveries were likely to be lost. As such a
1467 large number of recipients for a single message is ridiculous
1468 (performance will be very, very poor), I have chosen to impose a limit
1469 rather than extend the field.
1475 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1476 leftover -K file (the existence of which was triggered by #402).
1477 While we were at it, introduced process PID as part of the -K
1478 filename. This should rule out race conditions when creating
1481 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1482 processing considerably. Previous code took too long for large mails,
1483 triggering a timeout which in turn triggers #401.
1485 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1486 in the DK code in transports.c. sendfile() is not really portable,
1487 hence the _LINUX specificness.
1489 TF/01 In the add_headers option to the mail command in an Exim filter,
1490 there was a bug that Exim would claim a syntax error in any
1491 header after the first one which had an odd number of characters
1494 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1495 callout verification, Exim cached a "reject" for the entire domain. This
1496 is correct for most verifications, but it is not correct for a recipient
1497 verification with use_sender or use_postmaster set, because in that case
1498 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1499 case of MAIL FROM:<> rejection from other early rejections (e.g.
1500 rejection of HELO). When verifying a recipient using a non-null MAIL
1501 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1502 Whatever the result of the callout, the value of the domain cache is
1503 left unchanged (for any other kind of callout, getting as far as trying
1504 RCPT means that the domain itself is ok).
1506 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1509 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1510 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1511 session, the daemon ignores ECONNECT errors and logs others; it now
1512 ignores EPIPE as well.
1514 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1515 (quoted-printable decoding).
1517 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1518 later the small subsequent patch to fix an introduced bug.
1520 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1522 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1524 PH/08 An error is now given if message_size_limit is specified negative.
1526 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1527 to be given (somewhat) arbitrary names.
1529 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1532 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1533 miscellaneous code fixes
1535 PH/10 Added the log_reject_target ACL modifier to specify where to log
1538 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1539 hostname. This is wrong, because it relates to the incoming message (and
1540 probably the interface on which it is arriving) and not to the outgoing
1541 callout (which could be using a different interface). This has been
1542 changed to use the value of the helo_data option from the smtp transport
1543 instead - this is what is used when a message is actually being sent. If
1544 there is no remote transport (possible with a router that sets up host
1545 addresses), $smtp_active_hostname is used.
1547 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1548 tweaks were necessary in order to get it to work (see also 21 below):
1549 (a) The code assumed that strncpy() returns a negative number on buffer
1550 overflow, which isn't the case. Replaced with Exim's string_format()
1552 (b) There were several signed/unsigned issues. I just did the minimum
1553 hacking in of casts. There is scope for a larger refactoring.
1554 (c) The code used strcasecmp() which is not a standard C function.
1555 Replaced with Exim's strcmpic() function.
1556 (d) The code set only $1; it now sets $auth1 as well.
1557 (e) A simple test gave the error "authentication client didn't specify
1558 service in request". It would seem that Dovecot has changed its
1559 interface. Fortunately there's a specification; I followed it and
1560 changed what the client sends and it appears to be working now.
1562 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1565 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1566 address A is aliased to B and C, where B exists and C does not. Without
1567 -v the output is "A verified" because verification stops after a
1568 successful redirection if more than one address is generated. However,
1569 with -v the child addresses are also verified. Exim was outputting "A
1570 failed to verify" and then showing the successful verification for C,
1571 with its parentage. It now outputs "B failed to verify", showing B's
1572 parentage before showing the successful verification of C.
1574 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1575 look up a TXT record in a specific list after matching in a combined
1578 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1579 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1580 they consult the DNS. I had assumed they would set it the way they
1581 wanted; and indeed my experiments on Linux seem to show that in some
1582 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1583 To be on the safe side, however, I have now made the interface to
1584 host_find_byname() similar to host_find_bydns(), with an argument
1585 containing the DNS resolver options. The host_find_byname() function now
1586 sets these options at its start, just as host_find_bydns() does. The smtp
1587 transport options dns_qualify_single and dns_search_parents are passed to
1588 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1589 of host_find_byname() use the default settings of RES_DEFNAMES
1590 (qualify_single) but not RES_DNSRCH (search_parents).
1592 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1593 spool_read_header() do less string testing, by means of a preliminary
1594 switch on the second character of optional "-foo" lines. (This is
1595 overdue, caused by the large number of possibilities that now exist.
1596 Originally there were few.) While I was there, I also converted the
1597 str(n)cmp tests so they don't re-test the leading "-" and the first
1598 character, in the hope this might squeeze out yet more improvement.
1600 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1601 flag allowing group syntax was set by the header_syntax check but not
1602 turned off, possible causing trouble later; (2) The flag was not being
1603 set at all for the header_verify test, causing "group"-style headers to
1604 be rejected. I have now set it in this case, and also caused header_
1605 verify to ignore an empty address taken from a group. While doing this, I
1606 came across some other cases where the code for allowing group syntax
1607 while scanning a header line wasn't quite right (mostly, not resetting
1608 the flag correctly in the right place). These bugs could have caused
1609 trouble for malformed header lines. I hope it is now all correct.
1611 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1612 with the "reply" argument non-NULL. The code, however (which originally
1613 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1614 but it didn't always do it. This confused somebody who was copying the
1615 code for some other use. I have removed all the tests.
1617 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1618 feature that was used to support insecure browsers during the U.S. crypto
1619 embargo. It requires special client support, and Exim is probably the
1620 only MTA that supported it -- and would never use it because real RSA is
1621 always available. This code has been removed, because it had the bad
1622 effect of slowing Exim down by computing (never used) parameters for the
1623 RSA_EXPORT functionality.
1625 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1626 authenticator to fail if there's a tab character in the incoming data
1627 (there should never be unless someone is messing about, as it's supposed
1628 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1629 if the connection is using TLS or if the remote IP is the same as the
1630 local IP, and the "valid-client-cert option" if a client certificate has
1633 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1634 authenticators. This can be used for authorization after authentication
1635 succeeds. (In the case of plaintext, it servers for both authentication
1638 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1639 if any retry times were supplied.
1641 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1642 connection, where there is no client IP address to check. In this
1643 situation, the verify now always succeeds.
1645 PH/25 Applied John Jetmore's -Mset patch.
1647 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1649 PH/27 In a string expansion for a processed (not raw) header when multiple
1650 headers of the same name were present, leading whitespace was being
1651 removed from all of them, but trailing whitespace was being removed only
1652 from the last one. Now trailing whitespace is removed from each header
1653 before concatenation. Completely empty headers in a concatenation (as
1654 before) are ignored.
1656 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1657 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1659 PH/29 [Removed. This was a change that I later backed out, and forgot to
1660 correct the ChangeLog entry (that I had efficiently created) before
1661 committing the later change.]
1663 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1664 address errors (4xx response to RCPT) over the same connection as other
1665 messages routed to the same hosts. Such deliveries are always "forced",
1666 so retry times are not inspected. This resulted in far too many retries
1667 for the affected addresses. The effect occurred only when there were more
1668 hosts than the hosts_max_try setting in the smtp transport when it had
1669 the 4xx errors. Those hosts that it had tried were not added to the list
1670 of hosts for which the message was waiting, so if all were tried, there
1671 was no problem. Two fixes have been applied:
1673 (i) If there are any address or message errors in an SMTP delivery, none
1674 of the hosts (tried or untried) are now added to the list of hosts
1675 for which the message is waiting, so the message should not be a
1676 candidate for sending over the same connection that was used for a
1677 successful delivery of some other message. This seems entirely
1678 reasonable: after all the message is NOT "waiting for some host".
1679 This is so "obvious" that I'm not sure why it wasn't done
1680 previously. Hope I haven't missed anything, but it can't do any
1681 harm, as the worst effect is to miss an optimization.
1683 (ii) If, despite (i), such a delivery is accidentally attempted, the
1684 routing retry time is respected, so at least it doesn't keep
1685 hammering the server.
1687 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1688 in ${readsocket because some servers need this prod.
1690 PH/32 Added some extra debug output when updating a wait-xxx database.
1692 PH/33 The hint "could be header name not terminated by colon", which has been
1693 given for certain expansion errors for a long time, was not being given
1694 for the ${if def:h_colon_omitted{... case.
1696 PH/34 The spec says: "With one important exception, whenever a domain list is
1697 being scanned, $domain contains the subject domain." There was at least
1698 one case where this was not true.
1700 PH/35 The error "getsockname() failed: connection reset by peer" was being
1701 written to the panic log as well as the main log, but it isn't really
1702 panic-worthy as it just means the connection died rather early on. I have
1703 removed the panic log writing for the ECONNRESET error when getsockname()
1706 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1707 runs only) independently of the message's sender address. This meant
1708 that, if the 4xx error was in fact related to the sender, a different
1709 message to the same recipient with a different sender could confuse
1710 things. In particualar, this can happen when sending to a greylisting
1711 server, but other circumstances could also provoke similar problems.
1712 I have changed the default so that the retry time for these errors is now
1713 based a combination of the sender and recipient addresses. This change
1714 can be overridden by setting address_retry_include_sender=false in the
1717 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1718 remote server are returned as part of bounce messages. This was not
1719 happening for LMTP over a pipe (the lmtp transport), but now it is the
1720 same for both kinds of LMTP.
1722 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1723 in header lines that were in fact CNAMEs. This is no longer the case.
1725 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1726 and queue runs started by the daemon processed all messages. This has
1727 been fixed so that -R and -S can now usefully be given with -q<time>.
1729 PH/40 Import PCRE release 6.7 (fixes some bugs).
1731 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1733 PH/42 Give an error if -q is specified more than once.
1735 PH/43 Renamed the variables $interface_address and $interface_port as
1736 $received_ip_address and $received_port, to make it clear that these
1737 values apply to message reception, and not to the outgoing interface when
1738 a message is delivered. (The old names remain recognized, of course.)
1740 PH/44 There was no timeout on the connect() call when using a Unix domain
1741 socket in the ${readsocket expansion. There now is.
1743 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1744 be meaningful with "accept".
1746 SC/01 Eximstats V1.43
1747 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1749 SC/02 Eximstats V1.44
1750 Use a glob alias rather than an array ref in the generated
1751 parser. This improves both readability and performance.
1753 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1754 Collect SpamAssassin and rejection statistics.
1755 Don't display local sender or destination tables unless
1756 there is data to show.
1757 Added average volumes into the top table text output.
1759 SC/04 Eximstats V1.46
1760 Collect data on the number of addresses (recipients)
1761 as well as the number of messages.
1763 SC/05 Eximstats V1.47
1764 Added 'Message too big' to the list of mail rejection
1765 reasons (thanks to Marco Gaiarin).
1767 SC/06 Eximstats V1.48
1768 Mainlog lines which have GMT offsets and are too short to
1769 have a flag are now skipped.
1771 SC/07 Eximstats V1.49 (Alain Williams)
1772 Added the -emptyok flag.
1774 SC/08 Eximstats V1.50
1775 Fixes for obtaining the IP address from reject messages.
1777 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1778 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1779 whitesspace changes from 4.64-PH/27
1781 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1784 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1785 are found, allow negative numbers in numeric criteria)
1787 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1789 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1790 to match changes made in 4.64-PH/43
1792 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1794 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1795 "domain" is actually an IP address, and force a failure. This locks out
1796 those revolvers/nameservers that support "A-for-A" lookups, in
1797 contravention of the specifications.
1799 PH/48 When a host name was looked up from an IP address, and the subsequent
1800 forward lookup of the name timed out, the host name was left in
1801 $sender_host_name, contrary to the specification.
1803 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1804 restricted to single-key lookups, Exim was not diagnosing an error if
1805 * or *@ was used with a query-style lookup.
1807 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1809 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1810 long ago, but noone interested enough thought of it. Let's just say that
1811 the "1.1" means that there are some new functions that weren't there at
1812 some point in the past.
1814 PH/51 Error processing for expansion failure of helo_data from an smtp
1815 transport during callout processing was broken.
1817 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1818 tested/used via the -bh/-bhc/-bs options.
1820 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1821 bug, fixed in subsequent PCRE releases).
1823 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1824 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1826 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1832 SC/01 Use a glob alias rather than an array ref in eximstats generated
1833 parser. This improves both readability and performance.
1835 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1836 Don't display local sender or destination tables in eximstats unless
1837 there is data to show.
1838 Added average volumes into the eximstats top table text output.
1840 SC/03 Collect data on the number of addresses (recipients) as well
1841 as the number of messages in eximstats.
1843 TF/01 Correct an error in the documentation for the redirect router. Exim
1844 does (usually) call initgroups() when daemonizing.
1846 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1847 with consistent privilege compared to when running as a daemon.
1849 TF/03 Note in the spec that $authenticated_id is not set for local
1850 submissions from trusted users.
1852 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1853 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1855 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1856 by adding some example configuration directives to the default
1857 configuration file. A little bit of work is required to uncomment the
1858 directives and define how usernames and passwords are checked, but
1859 there is now a framework to start from.
1861 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1862 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1863 without this. I don't know how relevant this is to other LDAP libraries.
1865 PH/02 Add the verb name to the "unknown ACL verb" error.
1867 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1869 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1871 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1872 directory not expanded when it should be if an expanded home directory
1873 was set for the address (which is overridden by the transport).
1875 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1878 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1879 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1880 because it is too late at that time, and has no effect.
1882 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1883 security issue with \' (bugzilla #107). I could not use the
1884 PQescapeStringConn() function, because it needs a PGconn value as one of
1887 PH/08 When testing addresses using -bt, indicate those final addresses that
1888 are duplicates that would not cause an additional delivery. At least one
1889 person was confused, thinking that -bt output corresponded to deliveries.
1890 (Suppressing duplicates isn't a good idea as you lose the information
1891 about possibly different redirections that led to the duplicates.)
1893 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1894 systems where poll() doesn't work, in particular OS X.
1896 PH/10 Added more information to debugging output for retry time not reached.
1898 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1899 operations in malware.c.
1901 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1904 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1905 syslog_duplication set false, log lines that would normally be written
1906 both the the main log and to the reject log were not written to syslog at
1909 PH/14 In the default configuration, change the use of "message" in ACL warn
1910 statements to "add_header".
1912 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1913 not followed by a command (e.g. "seen endif").
1915 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1916 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1919 PH/17 Added extra conditions to the default value of delay_warning_condition
1923 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1924 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1925 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1928 The Auto-Submitted: and various List- headers are standardised, whereas I
1929 don't think Precedence: ever was.
1931 PH/18 Refactored debugging code in route_finduser() to show more information,
1932 in particular, the error code if getpwnam() issues one.
1934 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1935 This is apparently needed in addition to the PH/07 change above to avoid
1936 any possible encoding problems.
1938 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1939 but not after initializing Perl.
1941 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1942 output them only if debugging. By default they are written stderr,
1943 apparently, which is not desirable.
1945 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1948 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1951 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1953 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1954 authenticated or an ident call has been made. Suppress the default
1955 values for $authenticated_id and $authenticated_sender (but permit -oMai
1956 and -oMas) when testing with -bh.
1958 PH/24 Re-jigged the order of the tests in the default configuration so that the
1959 tests for valid domains and recipients precede the DNS black list and CSA
1960 tests, on the grounds that those ones are more expensive.
1962 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1963 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1964 This bug exists in every version of Exim that I still have, right back to
1967 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1968 However, an attempt to turn on case-sensitivity in a regex key by
1969 including (?-i) didn't work because the subject string was already
1970 lowercased, and the effects were non-intuitive. It turns out that a
1971 one-line patch can be used to allow (?-i) to work as expected.
1977 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1978 other effects) broke the use of negated acl sub-conditions.
1980 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1983 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1984 "Deny" causes Exim to reject the incoming connection with a 554 error.
1985 Unfortunately, if there is a major crisis, such as a disk failure,
1986 tcp-wrappers gives "deny", whereas what one would like would be some
1987 kind of temporary error. A kludge has been added to help with this.
1988 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1989 554 error is used if errno is still zero or contains ENOENT (which occurs
1990 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1993 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1995 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1996 errors. Otherwise a message that provokes a temporary error (when other
1997 messages do not) can cause a whole host to time out.
1999 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2000 addresses were routed directly to files or pipes from a redirect router.
2001 File deliveries just didn't batch; pipe deliveries might have suffered
2004 PH/06 A failure to get a lock for a hints database would erroneously always say
2005 "Failed to get write lock", even when it was really a read lock.
2007 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2008 of 0600. This has been changed to use the value of the lockfile_mode
2009 option (which defaults to 0600).
2011 PH/08 Applied small patch from the Sieve maintainer.
2013 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2014 folder from quota calculations, a direct delivery into this folder messed
2015 up the contents of the maildirsize file. This was because the regex was
2016 used only to exclude .Trash (or whatever) when the size of the mailbox
2017 was calculated. There was no check that a delivery was happening into an
2018 excluded directory. This bug has been fixed by ignoring all quota
2019 processing for deliveries into excluded directories.
2021 PH/10 Added the maildirfolder_create_regex option to appendfile.
2027 PH/01 The code for finding all the local interface addresses on a FreeBSD
2028 system running IPv6 was broken. This may well have applied to all BSD
2029 systems, as well as to others that have similar system calls. The broken
2030 code found IPv4 interfaces correctly, but gave incorrect values for the
2031 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2032 that it would not match correctly against @[] and not recognize the IPv6
2035 PH/02 The ipliteral router was not recognizing addresses of the form user@
2036 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2038 PH/03 Added disable_ipv6.
2040 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2041 decoded form, because it is most often used to construct To: headers
2042 lines in autoreplies, and the decoded form may well be syntactically
2043 invalid. However, $reply_address has leading white space removed, and all
2044 newlines turned into spaces so that the autoreply transport does not
2047 PH/05 If group was specified without a user on a router, and no group or user
2048 was specified on a transport, the group from the router was ignored.
2050 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2051 for visible compile-time settings that can be used to change these
2052 numbers, for those that want even more. Backwards compatibility with old
2053 spool files has been maintained. However, going back to a previous Exim
2054 release will lost any variables that are in spool files.
2056 PH/07 Two small changes when running in the test harness: increase delay when
2057 passing a TCP/IP connection to a new process, in case the original
2058 process has to generate a bounce, and remove special handling of
2059 127.0.0.2 (sic), which is no longer necessary.
2061 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2062 be the same on different OS.
2064 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2067 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2068 whether --show-vars was specified or not
2070 JJ/02 exipick: Added support for new ACL variable spool format introduced
2073 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2074 syntactically invalid From: or Reply-to: line, and a filter used this to
2075 generate an autoreply, and therefore failed to obtain an address for the
2076 autoreply, Exim could try to deliver to a non-existent relative file
2077 name, causing unrelated and misleading errors. What now happens is that
2078 it logs this as a hard delivery error, but does not attempt to create a
2081 PH/11 The exinext utility has a -C option for testing purposes, but although
2082 the given file was scanned by exinext itself; it wasn't being passed on
2083 when Exim was called.
2085 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2086 an end-of-file indication when reading a command response.
2088 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2089 compiled. In many other places in Exim, IPv6 addresses are always
2090 recognized, so I have changed this. It also means that IPv4 domain
2091 literals of the form [IPV4:n.n.n.n] are now always recognized.
2093 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2094 used if the router is not running as root, for example, when verifying at
2095 ACL time, or when using -bh. The debugging output from this situation was
2096 non-existent - all you got was a failure to exec. I have made two
2099 (a) Failures to set uid/gid, the current directory, or a process leader
2100 in a subprocess such as that created by queryprogram now generate
2101 suitable debugging ouput when -d is set.
2103 (b) The queryprogram router detects when it is not running as root,
2104 outputs suitable debugging information if -d is set, and then runs
2105 the subprocess without attempting to change uid/gid.
2107 PH/15 Minor change to Makefile for building test_host (undocumented testing
2110 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2111 additional section of a DNS packet that returns MX or SRV records.
2112 Instead, it always explicitly searches for A/AAAA records. This avoids
2113 major problems that occur when a DNS server includes only records of one
2114 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2115 fixed another bug: if SRV records were looked up and the corresponding
2116 address records were *not* found in the additional section, the port
2117 values from the SRV records were lost.
2119 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2120 using the correct key (the original address) when searching the retry
2121 rules in order to find which one to use for generating the retry hint.
2123 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2124 adding the default one. Similarly, if it contains a Reply-To: header, the
2125 errors_reply_to option, if set, is not used.
2127 PH/19 When calculating a retry time, Exim used to measure the "time since
2128 failure" by looking at the "first failed" field in the retry record. Now
2129 it does not use this if it is later than than the arrival time of the
2130 message. Instead it uses the arrival time. This makes for better
2131 behaviour in cases where some deliveries succeed, thus re-setting the
2132 "first failed" field. An example is a quota failure for a huge message
2133 when small messages continue to be delivered. Without this change, the
2134 "time since failure" will always be short, possible causing more frequent
2135 delivery attempts for the huge message than are intended.
2136 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2138 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2139 $1, $2, $3) because the numerical variables can be reset during some
2140 expansion items (e.g. "match"), thereby losing the authentication data.
2142 PH/21 Make -bV show the size of off_t variables so that the test suite can
2143 decide whether to run tests for quotas > 2G.
2145 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2146 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2147 filecount value is greater than 2G or if a quota value is greater than 2G
2148 on a system where the size of off_t is not greater than 4, a panic error
2151 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2152 never match. The debug and -bh output now contains an explicit error
2153 message indicating a malformed IPv4 address or mask.
2155 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2156 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2157 PH/23 above applies.
2159 PH/25 Do not write to syslog when running in the test harness. The only
2160 occasion when this arises is a failure to open the main or panic logs
2161 (for which there is an explicit test).
2163 PH/26 Added the /no_tell option to "control=freeze".
2165 PH/27 If a host name lookup failed very early in a connection, for example, if
2166 the IP address matched host_lookup and the reverse lookup yielded a name
2167 that did not have a forward lookup, an error message of the form "no IP
2168 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2169 could be logged. Now it outputs the IP address instead of "NULL".
2171 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2172 allows the sender and the authenticated sender to be set when
2173 submitting a message from within Exim. Since child_open_exim() is
2174 documented for local_scan(), the new function should be too.
2176 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2177 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2178 results in an empty string is now treated as unset.
2180 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2182 PH/31 Added sender_verify_fail logging option.
2184 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2185 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2186 tidied the source and removed it altogether.
2188 PH/33 When a queue run was abandoned because the load average was too high, a
2189 log line was always written; now it is written only if the queue_run log
2190 selector is set. In addition, the log line for abandonment now contains
2191 information about the queue run such as the pid. This is always present
2192 in "start" and "stop" lines but was omitted from the "abandon" line.
2194 PH/34 Omit spaces between a header name and the colon in the error message that
2195 is given when verify = headers_syntax fails (if there are lots of them,
2196 the message gets confusing).
2198 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2199 names, as there are now some PTR records that contain slashes. This check
2200 is only to protect against broken name servers that fall over on strange
2201 characters, so the fact that it applies to all lookups doesn't matter.
2203 PH/36 Now that the new test suite is complete, we can remove some of the
2204 special code in Exim that was needed for the old test suite. For example,
2205 sorting DNS records because real resolvers return them in an arbitrary
2206 order. The new test suite's fake resolver always returns records in the
2209 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2210 bounces) except when queue_only is set, to avoid logging races between
2211 the different processes.
2213 PH/38 Panic-die if .include specifies a non-absolute path.
2215 PH/39 A tweak to the "H" retry rule from its user.
2217 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2218 a label. They prevented compilation on older perls.
2220 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2221 a warning to be raised on newish perls.
2223 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2224 on queue. Changes to match documented behaviour of showing count of
2225 messages matching specified criteria.
2227 PH/40 Changed the default ident timeout from 30s to 5s.
2229 PH/41 Added support for the use of login_cap features, on those BSD systems
2230 that have them, for controlling the resources used by pipe deliveries.
2232 PH/42 The content-scanning code uses fopen() to create files in which to put
2233 message data. Previously it was not paying any attention to the mode of
2234 the files. Exim runs with umask(0) because the rest of the code creates
2235 files with open(), and sets the required mode explicitly. Thus, these
2236 files were ending up world-writeable. This was not a big issue, because,
2237 being within the spool directory, they were not world-accessible. I have
2238 created a function called modefopen, which takes an additional mode
2239 argument. It sets umask(777), creates the file, chmods it to the required
2240 mode, then resets the umask. All the relevant calls to fopen() in the
2241 content scanning code have been changed to use this function.
2243 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2244 to 24 hours. This avoids potential overflow problems when processing G
2245 and H retry rules. I suspect nobody ever tinkers with this value.
2247 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2249 PH/45 When the plaintext authenticator is running as a client, the server's
2250 challenges are checked to ensure they are valid base64 strings. By
2251 default, the authentication attempt is cancelled if an invalid string is
2252 received. Setting client_ignore_invalid_base64 true ignores these errors.
2253 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2254 they are received. Thus, the responses can be made to depend on the
2255 challenges. If an invalid string is ignored, an empty string is placed in
2258 PH/46 Messages that are created by the autoreply transport now contains a
2259 References: header, in accordance with RFCs 2822 and 3834.
2261 PH/47 Added authenticated_sender_force to the smtp transport.
2263 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2265 PH/49 Installed latest patch from the Sieve maintainer.
2267 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2268 was also set, the appendfile transport was unnecessarily scanning a
2269 directory of message files (e.g. for maildir delivery) to find the count
2270 of files (along with the size), even though it did not need this
2271 information. It now does the scan only if it needs to find either the
2272 size of the count of files.
2274 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2276 PH/52 Two bugs concerned with error handling when the smtp transport is
2279 (i) Exim was not creating retry information for temporary errors given
2280 for individual recipients after the DATA command when the smtp transport
2281 was used in LMTP mode. This meant that they could be retried too
2282 frequently, and not timed out correctly.
2284 (ii) Exim was setting the flag that allows error details to be returned
2285 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2286 recipients that were returned after the DATA command.
2288 PH/53 This is related to PH/52, but is more general: for any failing address,
2289 when detailed error information was permitted to be returned to the
2290 sender, but the error was temporary, then after the final timeout, only
2291 "retry timeout exceeded" was returned. Now it returns the full error as
2292 well as "retry timeout exceeded".
2294 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2295 do this, and (what is worse) MTAs that accept it.
2297 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2298 will now be deprecated.
2300 PH/56 New os.c-cygwin from the Cygwin maintainer.
2302 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2303 formats (previously only available in exim formats via -bpr, -bpru,
2304 and -bpra. Now also available in native and exiqgrep formats)
2306 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2307 with very large, slow to parse queues
2309 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2311 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2313 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2314 responses to authentication challenges, though it was showing the
2315 challenges; (ii) I've removed the CR characters from the debug output for
2318 PH/58 Allow for the insertion of a newline as well as a space when a string
2319 is turned into more than one encoded-word during RFC 2047 encoding. The
2320 Sieve code now uses this.
2322 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2323 data_4xx, lost_connection, tls_required.
2325 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2326 message was being sent as an SMTP response.
2328 PH/61 Add -l and -k options to exicyclog.
2330 PH/62 When verifying, if an address was redirected to one new address, so that
2331 verification continued, and the new address failed or deferred after
2332 having set something in $address_data, the value of $address_data was not
2333 passed back to the ACL. This was different to the case when no
2334 redirection occurred. The value is now passed back in both cases.
2336 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2337 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2338 use login_cap.h, so on its own it isn't the distinguishing feature. The
2339 new name refers directly to the setclassresources() function.
2341 PH/65 Added configuration files for NetBSD3.
2343 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2345 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2346 is preferred over IPv4.
2348 PH/68 The bounce_return_message and bounce_return_body options were not being
2349 honoured for bounces generated during the reception of non-SMTP messages.
2350 In particular, this applied to messages rejected by the ACL. This bug has
2351 been fixed. However, if bounce_return_message is true and bounce_return_
2352 body is false, the headers that are returned for a non-SMTP message
2353 include only those that have been read before the error was detected.
2354 (In the case of an ACL rejection, they have all been read.)
2356 PH/69 The HTML version of the specification is now built in a directory called
2357 spec_html instead of spec.html, because the latter looks like a path with
2358 a MIME-type, and this confuses some software.
2360 PH/70 Catch two compiler warnings in sieve.c.
2362 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2363 function verify_get_ident() calls ip_connect() to connect a socket, but
2364 if the "connect()" function timed out, ip_connect() used to close the
2365 socket. However, verify_get_ident() also closes the socket later, and in
2366 between Exim writes to the log, which may get opened at this point. When
2367 the socket was closed in ip_connect(), the log could get the same file
2368 descriptor number as the socket. This naturally causes chaos. The fix is
2369 not to close the socket in ip_connect(); the socket should be closed by
2370 the function that creates it. There was only one place in the code where
2371 this was missing, in the iplookup router, which I don't think anybody now
2372 uses, but I've fixed it anyway.
2374 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2375 well as to direct DNS lookups. Otherwise the handling of names in host
2376 lists is inconsistent and therefore confusing.
2382 PH/01 Two changes to the default runtime configuration:
2384 (1) Move the checks for relay_from_hosts and authenticated clients from
2385 after to before the (commented out) DNS black list checks.
2387 (2) Add control=submission to the relay_from_hosts and authenticated
2388 clients checks, on the grounds that messages accepted by these
2389 statements are most likely to be submissions.
2391 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2393 (1) Generate an error if the third argument for the ${prvs expansion is
2396 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2399 (3) Reset the variables that are obtained from the first argument of
2400 ${prvscheck and used in the second argument before leaving the code,
2401 because their memory is reclaimed, so using them afterwards may do
2404 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2405 one (it's much easier than Tom thought :-).
2407 (5) Because of (4), we can now allow for the use of $prvscheck_result
2408 inside the third argument.
2410 PH/03 For some reason, the default setting of PATH when running a command from
2411 a pipe transport was just "/usr/bin". I have changed it to
2414 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2415 anything to be listed in the output from -bV.
2417 PH/05 When a filter generated an autoreply, the entire To: header line was
2418 quoted in the delivery log line, like this:
2420 => >A.N.Other <ano@some.domain> <original@ddress> ...
2422 This has been changed so that it extracts the operative address. There
2423 may be more than one such address. If so, they are comma-separated, like
2426 => >ano@some.domain,ona@other.domain <original@ddress> ...
2428 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2429 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2430 not being looked up in the rDNS to get a host name, Exim was showing the
2431 IP address twice in Received: lines, even though the IP addresses were
2432 identical. For example:
2434 Received: from [1.2.3.4] (helo=[1.2.3.4])
2436 However, if the real host name was known, it was omitting the HELO data
2437 if it matched the actual IP address. This has been tidied up so that it
2438 doesn't show the same IP address twice.
2440 PH/07 When both +timestamp and +memory debugging was on, the value given by
2441 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2442 called by the debug printing, thereby overwriting the timestamp buffer.
2443 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2445 PH/08 When the original message was included in an autoreply transport, it
2446 always said "this is a copy of the message, including all the headers",
2447 even if body_only or headers_only was set. It now gives an appropriate
2450 PH/09 Applied a patch from the Sieve maintainer which:
2452 o fixes some comments
2453 o adds the (disabled) notify extension core
2454 o adds some debug output for the result of if/elsif tests
2455 o points to the current vacation draft in the documentation
2456 and documents the missing references header update
2460 o fixes a bug in processing the envelope test (when testing
2461 multiple envelope elements, the last element determinted the
2464 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2465 Electronic Mail") by including:
2467 Auto-submitted: auto-generated
2469 in the messages that it generates (bounce messages and others, such as
2470 warnings). In the case of bounce messages for non-SMTP mesages, there was
2471 also a typo: it was using "Auto_submitted" (underscore instead of
2472 hyphen). Since every message generated by Exim is necessarily in response
2473 to another message, thes have all been changed to:
2475 Auto-Submitted: auto-replied
2477 in accordance with these statements in the RFC:
2479 The auto-replied keyword:
2481 - SHOULD be used on messages sent in direct response to another
2482 message by an automatic process,
2484 - MUST NOT be used on manually-generated messages,
2486 - MAY be used on Delivery Status Notifications (DSNs) and Message
2487 Disposition Notifications (MDNs),
2489 - MUST NOT be used on messages generated by automatic or periodic
2490 processes, except for messages which are automatic responses to
2493 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2494 to the default Received: header definition.
2496 PH/12 Added log selector acl_warn_skipped (default on).
2498 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2499 variables because (a) they are in the wrong storage pool and (b) even if
2500 they were copied, it wouldn't work properly because of the caching.
2502 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2503 checking when decoding. Apparently there are clients that generate
2504 overlong encoded strings. Why am I not surprised?
2506 PH/15 If the first argument of "${if match_address" was not empty, but did not
2507 contain an "@" character, Exim crashed. Now it writes a panic log message
2508 and treats the condition as false.
2510 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2512 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2513 extension "envelope-auth". The code is finished and in agreement with
2514 other implementations, but there is no documentation so far and in fact,
2515 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2516 not changing the active code.
2518 Print executed "if" and "elsif" statements when debugging is used. This
2519 helps a great deal to understand what a filter does.
2521 Document more things not specified clearly in RFC3028. I had all this
2522 sorted out, when out of a sudden new issues came to my mind. Oops."
2524 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2527 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2528 canonical form (as documented). However, after a host name lookup from
2529 the IP address, check_host() was doing a simple string comparison with
2530 addresses acquired from the DNS when checking that the found name did
2531 have the original IP as one of its addresses. Since any found IPv6
2532 addresses are likely to be in abbreviated form, the comparison could
2533 fail. Luckily, there already exists a function for doing the comparison
2534 by converting both addresses to binary, so now that is used instead of
2535 the text comparison.
2537 PH/20 There was another similar case to PH/19, when a complete host name was
2538 given in a host list; looking up its IP address could give an abbreviated
2539 form, whereas the current host's name might or might not be abbreviated.
2540 The same fix has been applied.
2546 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2547 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2550 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2551 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2553 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2555 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2556 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2557 Also, for "Auto-Submitted", treat anything other than "no" as
2558 non-personal, in accordance with RFC 3834. (Previously it treated
2559 anything starting "auto-" as non-personal.)
2561 TF/01 The control=submission/name=... option had a problem with syntax
2562 errors if the name included a slash character. The /name= option
2563 now slurps the rest of the string, so it can include any characters
2564 but it must come last in the list of options (after /sender_retain
2567 PH/05 Some modifications to the interface to the fake nameserver for the new
2575 TK/01 Added the "success_on_redirect" address verification option. See
2576 NewStuff for rationale and an example.
2578 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2580 PH/02 Patch to exigrep to allow it to work on syslog lines.
2582 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2583 fread() to skip over the body file's header line, because in Cygwin the
2584 header line is locked and is inaccessible.
2586 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2587 co-exist for some time) to make it clear that it is the Exim ID that is
2588 referenced, not the Message-ID: header line.
2590 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2591 string_format() function, because snprintf() does not exist on all
2594 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2595 ${stat: expansion item.
2597 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2598 protocol synchronization error", to keep the pedants happy.
2600 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2601 well as for IRIX systems, when gcc is being used. See the host.c source
2604 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2606 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2609 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2610 transport and to the smtp transport in LMTP mode.
2612 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2614 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2615 run a filter in a subprocess. This could lead to confusion in subsequent
2616 lookups in the parent process. There should also be a search_tidyup() at
2617 the end of the subprocess.
2619 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2620 only if the host matched helo_try_verify_hosts, which caused the
2621 verification to occur when the EHLO/HELO command was issued. The ACL just
2622 tested the remembered result. Now, if a previous verification attempt has
2623 not happened, "verify = helo" does it there and then.
2625 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2627 TK/03 Fix log output including CR from clamd.
2629 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2630 exist provoked a memory error which could cause a segfault.
2632 PH/15 Installed PCRE 6.2
2634 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2636 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2637 of the problem. Specifically, suggested +O2 rather than +O1 for the
2640 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2642 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2643 record, the port was not correctly logged when the outgoing_port log
2644 selector was set (it logged the transort's default port).
2646 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2647 fallback_hosts, and "hosts" in the smtp transport.
2649 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2650 host errors such as "Connection refused".
2652 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2653 authentication with radiusclient 0.4.9:
2655 - Error returned from rc_read_config was caught wrongly
2656 - Username/password not passed on to radius server due to wrong length.
2658 The presumption is that some radiusclient API changes for 4.51/PH/17
2659 were not taken care of correctly. The code is still untested by me (my
2660 Linux distribution still has 0.3.2 of radiusclient), but it was
2661 contributed by a Radius user.
2663 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2664 expanding the "port" option of the smtp transport.
2666 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2667 while reading a MIME header. Thanks to Tom Hughes for a patch.
2669 PH/24 Include config.h inside local_scan.h so that configuration settings are
2672 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2673 that in an ACL for RCPT (for example), you can examine exactly what was
2676 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2677 commands, but it was not correctly comparing the address with the actual
2678 client host address. Thus, it would show the EHLO address in Received:
2679 header lines when this was not necessary.
2681 PH/27 Added the % operator to ${eval:}.
2683 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2684 it should be ignoring failures (because with -C, for example, it has lost
2685 privilege). It wasn't ignoring creation failures other than "already
2688 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2691 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2692 because an input line was too long, either on its own, or by virtue of
2693 too many continuations, the temporary file was not being removed, and the
2694 return code was incorrect.
2696 PH/31 Missing "BOOL" in function definition in filtertest.c.
2698 PH/32 Applied Sieve patches from the maintainer.
2700 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2702 PH/33 Added "verify = not_blind".
2704 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2705 Local/Makefile (with some defaults set). These are used in built scripts
2706 such as exicyclog, but they have never been used in the exim_install
2707 script (though there are many overriding facilities there). I have
2708 arranged that the exim_install script now takes note of these two
2711 PH/35 Installed configuration files for Dragonfly.
2713 PH/36 When a locally submitted message by a trusted user did not contain a
2714 From: header, and the sender address was obtained from -f or from an SMTP
2715 MAIL command, and the trusted user did not use -F to supply a sender
2716 name, $originator_name was incorrectly used when constructing a From:
2717 header. Furthermore, $originator_name was used for submission mode
2718 messages from external hosts without From: headers in a similar way,
2719 which is clearly wrong.
2721 PH/37 Added control=suppress_local_fixups.
2723 PH/38 When log_selector = +received_sender was set, and the addition of the
2724 sender made the log line's construction buffer exactly full, or one byte
2725 less than full, an overflow happened when the terminating "\n" was
2728 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2729 when the result of a list match is failure because a DNS lookup failed.
2731 PH/40 RM_COMMAND is now used in the building process.
2733 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2734 the "build-* directories that it finds.
2736 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2737 address in a domain literal was a prefix of an interface address.
2739 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2740 when verifying a sender address, unless rewrite_headers is false.
2742 PH/44 Wrote a long comment about why errors_to addresses are verified as
2743 recipients, not senders.
2745 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2746 the ratelimit ACL was added.
2748 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2750 PH/46 Added extra information about PostgreSQL errors to the error string.
2752 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2753 avoiding the need to install special zones in a real server. This is
2754 backwards compatible; if it can't find the fake resolver, it drops back.
2755 Thus, both old and new test suites can be run.
2757 TF/02 Added util/ratelimit.pl
2759 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2760 clock is set back in time.
2762 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2763 Candler <B.Candler@pobox.com>.
2765 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2766 OK for addresses that are the result of redirections.
2768 PH/48 A number of further additions for the benefit of the new test suite,
2769 including a fake gethostbyname() that interfaces to the fake DNS resolver
2772 TF/06 The fix for widen_domains has also been applied to qualify_single and
2773 search_parents which are the other dnslookup options that can cause
2776 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2779 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2781 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2782 service" if it encountered an MX record with an empty target hostname.
2783 The message is now "an MX or SRV record indicated no SMTP service".
2785 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2786 if the DNS of the sending site is misconfigured. This is quite a
2787 common situation. This change restores the behaviour of treating a
2788 helo verification defer as a failure.
2790 PH/51 If self=fail was set on a router, the bounce message did not include the
2791 actual error message.
2797 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2799 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2800 command itself ended in error, the underlying message about the transport
2801 filter timeout was being overwritten with the pipe command error. Now the
2802 underlying error message should be appended to the second error message.
2804 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2806 PH/02 Reduce the amount of output that "make" produces by default. Full output
2807 can still be requested.
2809 PH/03 The warning log line about a condition test deferring for a "warn" verb
2810 was being output only once per connection, rather than after each
2811 occurrence (because it was using the same function as for successful
2812 "warn" verbs). This seems wrong, so I have changed it.
2814 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2815 it should not have, which might have caused a crash in the right
2816 circumstances, but probably never did.
2818 PH/04 Installed a modified version of Tony Finch's patch to make submission
2819 mode fix the return path as well as the Sender: header line, and to
2820 add a /name= option so that you can make the user's friendly name appear
2823 TF/03 Added the control = fakedefer ACL modifier.
2825 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2826 Mark Lowes for thorough testing.
2828 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2830 TK/03 Merged latest SRS patch from Miles Wilton.
2832 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2833 with the definition in sysexits.h (which is #included earlier).
2834 Fortunately, Exim does not actually use EX_OK. The code used to try to
2835 preserve the sysexits.h value, by assumimg that macro definitions were
2836 scanned for macro replacements. I have been disabused of this notion,
2837 so now the code just undefines EX_OK before #including unistd.h.
2839 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2840 in the smtp transport. When a block could not be written in a single
2841 write() function, the timeout was being re-applied to each part-write.
2842 This seems wrong - if the receiver was accepting one byte at a time it
2843 would take for ever. The timeout is now adjusted when this happens. It
2844 doesn't have to be particularly precise.
2846 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2847 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2849 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2850 without a domain if the check to <postmaster@domain> fails.
2852 SC/01 Eximstats: added -xls and the ability to specify output files
2853 (patch written by Frank Heydlauf).
2855 SC/02 Eximstats: use FileHandles for outputing results.
2857 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2859 SC/04 Eximstats: fixed display of large numbers with -nvr option
2861 SC/05 Eximstats: fixed merging of reports with empty tables.
2863 SC/06 Eximstats: added the -include_original_destination flag
2865 SC/07 Eximstats: removed tabs and trailing whitespace.
2867 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2869 TK/06 MBOX spool code: Add real "From " MBOX separator line
2870 so the .eml file is really in mbox format (even though
2871 most programs do not really care). Patch from Alex Miller.
2873 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2874 The latter is generated from $received_to and is only set if the
2875 message has one envelope recipient. SA can use these headers,
2876 obviously out-of-the-box. Patch from Alex Miller.
2878 PH/08 The ${def test on a variable was returning false if the variable's
2879 value was "0", contrary to what the specification has always said!
2880 The result should be true unless the variable is empty.
2882 PH/09 The syntax error of a character other than { following "${if
2883 def:variable_name" (after optional whitespace) was not being diagnosed.
2884 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2885 accidental colon was present, for example, could give incorrect results.
2887 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2888 result is used (not including appendfile, where other changes are about
2891 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2892 This involved changing a lot of size variables from int to off_t. It
2893 should work with maildirs and everything.
2895 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2896 spamd dying while we are connected to it.
2898 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2901 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2902 function for BDB 4.3.
2904 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2906 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2907 the sources for PCRE have been re-arranged and more files are now
2910 PH/15 The code I had for printing potentially long long variables in PH/11
2911 above was not the best (it lost precision). The length of off_t variables
2912 is now inspected at build time, and an appropriate printing format (%ld
2913 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2914 to be "long long int" or "long int". This is needed for the internal
2915 formatting function string_vformat().
2917 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2918 the configuration file to be ":syslog", then the script "guesses" where
2919 the logs files are, rather than using the compiled in default. In our
2920 case the guess is not the same as the compiled default, so the script
2921 suddenly stopped working when I started to use syslog. The patch checks
2922 to see if log_file_path is "". If so, it attempts to read it from exim
2923 with no configuration file to get the compiled in version, before it
2924 falls back to the previous guessing code."
2926 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2927 implementing BATV in an Exim configuration. See NewStuff for the gory
2930 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2931 Makefile that are specific to HP-UX.
2933 PH/18 If the "use_postmaster" option was set for a recipient callout together
2934 with the "random" option, the postmaster address was used as the MAIL
2935 FROM address for the random test, but not for the subsequent recipient
2936 test. It is now used for both.
2938 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2939 patch removes a few documentation additions to RFC 3028, because the
2940 latest draft now contains them. It adds the new en;ascii-case comparator
2941 and a new error check for 8bit text in MIME parts. Comparator and
2942 require names are now matched exactly. I enabled the subaddress
2943 extension, but it is not well tested yet (read: it works for me)."
2945 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2946 rework some of the code of TK/09 above to avoid the hardwired use of
2947 "%lld" and "long long". Replaced the call to snprintf() with a call to
2950 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2951 records point to non-existent hosts", "retry timeout exceeded", and
2952 "retry time not reached for any host after a long failure period".
2954 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2955 experimental DomainKeys support:
2957 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2958 (2) On an error such as an illegally used "control", the wrong name for
2959 the control was given.
2961 These problems did NOT occur unless DomainKeys support was compiled.
2963 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2965 PH/24 Added ${if match_ip condition.
2967 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2968 databases so that it will be absolutely obvious if a crash occurs in the
2969 DB library. This is a regular occurrence (often caused by mis-matched
2972 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2973 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2974 picked up on a user's system that detects such things. There doesn't seem
2975 to be a gcc warning option for this - only an attribute that has to be
2976 put on the function's prototype. It seems that in Fedora Core 4 they have
2977 set this on a number of new functions. No doubt there will be more in due
2980 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2981 specify a transport. However, if an address that was verified by such a
2982 router was the subject of a callout, Exim crashed because it tried to
2983 read the rcpt_include_affixes from the non-existent transport. Now it
2984 just assumes that the setting of that option is false. This bug was
2985 introduced by 4.51/PH/31.
2987 PH/28 Changed -d+all to exclude +memory, because that information is very
2988 rarely of interest, but it makes the output a lot bigger. People tend to
2989 do -d+all out of habit.
2991 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2992 code in os-type was giving problems when libc.so lives in lib64, like on
2995 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2996 aren't the modern standard, and it seems that some systems' include files
2997 don't always have them. Exim was already checking for some of the newer
2998 ones like T_AAAA, and defining it itself. I've added checks for all the
2999 record types that Exim uses.
3001 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3002 not automatically generating a new one, as it is supposed to. This
3003 prevented TLS from working. If the file did exist, but contained invalid
3004 data, a new version was generated, as expected. It was only the case of a
3005 non-existent file that was broken.
3007 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3008 with a change in libdomainkeys > 0.64.
3010 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3011 from DNS. If the selector record carries the flag, it now has
3012 precedence over the domain-wide flag.
3014 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3016 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3017 the use of a port name that isn't defined in /etc/services) occurred, the
3018 message was deferred as in a normal delivery, and thus remained on the
3019 spool, instead of being failed because of the mua_wrapper setting. This
3020 is now fixed, and I tidied up some of the mua_wrapper messages at the
3023 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3024 the messages in a hash of arrays rather than using individual hashes.
3025 This is a bit cleaner and results in dramatic memory savings, albeit
3026 at a slight CPU cost.
3028 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3029 as requested by Marc Sherman.
3031 SC/10 Eximstats: added histograms for user specified patterns as requested
3034 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3036 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3037 fopen() in the content-scanning modules that did not already have it.
3043 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3044 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3046 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3048 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3050 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3051 to test Sieve filters that use "vacation".
3053 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3054 that changes the way the GnuTLS parameters are stored in the cache file.
3055 The new format can be generated externally. For backward compatibility,
3056 if the data in the cache doesn't make sense, Exim assumes it has read an
3057 old-format file, and it generates new data and writes a new file. This
3058 means that you can't go back to an older release without removing the
3061 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3062 work if there are any delivery delays because "one_time" forces the
3063 parent to be marked "delivered", so its unseen clone is never tried
3064 again. For this reason, Exim now forbids the simultaneous setting of
3067 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3068 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3069 entry if you want to know the details. The fix, however, neglected to
3070 consider the case when local delivery batching is involved. The test for
3071 "previously delivered" was not happening when checking to see if an
3072 address could be batched with a previous (undelivered) one; under
3073 certain circumstances this could lead to multiple deliveries to the same
3076 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3077 in its include files, and this causes problems building Exim.
3079 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3080 header_syntax) but Exim was just ignoring anything given after a slash.
3081 In particular, this caused confusion with an attempt to use "verify =
3082 reverse_host_lookup/defer_ok". An error is now given when options are
3083 supplied for verify items that do not have them. (Maybe reverse_host_
3084 lookup should have a defer_ok option, but that's a different point.)
3086 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3087 defined by RFC 821) to 2048, because there were problems with some AUTH
3088 commands, and RFC 1869 says the size should be increased for extended
3089 SMTP commands that take arguments.
3091 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3094 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3095 "unknown" error; now it says that the functionality isn't in the binary.
3097 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3098 an address' error message when a string expansion fails (syntax or
3099 whatever). Otherwise the password may appear in the log. Following change
3100 PH/42 below, there is no longer a chance of it appearing in a bounce
3103 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3105 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3106 of its addresses was ever tried. (Bugzilla bug #2.)
3108 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3109 the result incorrectly in the debug output. (It correctly added a newline
3110 to what was transported.)
3112 TF/01 Added $received_time.
3114 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3115 commented out examples of how to interface to a virus scanner and to
3116 SpamAssassin. Also added commented examples of av_scanner and
3117 spamd_address settings.
3119 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3120 and controls are allowed in which ACLs. There were a couple of minor
3121 errors. Some of the entries in the conditions table (which is a table of
3122 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3123 negation of where the condition IS allowed.
3125 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3127 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3128 header file does not have a version number, so I've had to invent a new
3129 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3130 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3131 radiusclient), but it was contributed by a Radius user.
3133 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3134 files or directories, for OpenSSL.
3136 PH/19 When an Exim process that is running as root has to create an Exim log
3137 file, it does so in a subprocess that runs as exim:exim so as to get the
3138 ownership right at creation (otherwise, other Exim processes might see
3139 the file with the wrong ownership). There was no test for failure of this
3140 fork() call, which would lead to the process getting stuck as it waited
3141 for a non-existent subprocess. Forks do occasionally fail when resources
3142 run out. I reviewed all the other calls to fork(); they all seem to check
3145 PH/20 When checking for unexpected SMTP input at connect time (before writing
3146 the banner), Exim was not dealing correctly with a non-positive return
3147 from the read() function. If the client had disconnected by this time,
3148 the result was a log entry for a synchronization error with an empty
3149 string after "input=" when read() returned zero. If read() returned -1
3150 (an event I could not check), uninitialized data bytes were printed.
3151 There were reports of junk text (parts of files, etc) appearing after
3154 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3156 PH/22 Added support for macro redefinition, and (re)definition in between
3157 driver and ACL definitions.
3159 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3160 forgetting to use the resulting value; it was using the unexpanded value.
3162 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3163 hadn't been configured. The fix is from Juergen Kreileder, who
3164 understands it better than I do:
3166 "Here's what I see happening with three configured cyrus_sasl
3167 authenticators configured (plain, login, cram-md5):
3169 On startup auth_cyrus_sasl_init() gets called for each of these.
3170 This means three calls to sasl_listmech() without a specified mech_list.
3171 => SASL tests which mechs of all available mechs actually work
3172 => three warnings about OTP not working
3173 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3175 With the patch, sasl_listmech() also gets called three times. But now
3176 SASL's mech_list option is set to the server_mech specified in the the
3177 authenticator. Or in other words, the answer from sasl_listmech()
3178 gets limited to just the mech you're testing for (which is different
3180 => the return list contains just 'plain' or 'login', 'cram-md5' or
3181 nothing depending on the value of ob->server_mech.
3183 I've just tested the patch: Authentication still works fine,
3184 unavailable mechs specified in the exim configuration are still
3185 caught, and the auth.log warnings about OTP are gone."
3187 PH/25 When debugging is enabled, the contents of the command line are added
3188 to the debugging output, even when log_selector=+arguments is not
3191 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3192 answer is "GNU", and only if the return is "GNU/something" is the answer
3195 PH/27 $acl_verify_message is now set immediately after the failure of a
3196 verification in an ACL, and so is available in subsequent modifiers. In
3197 particular, the message can be preserved by coding like this:
3199 warn !verify = sender
3200 set acl_m0 = $acl_verify_message
3202 Previously, $acl_verify_message was set only while expanding "message"
3203 and "log_message" when a very denied access.
3205 PH/28 Modified OS/os.c-Linux with
3207 -#ifndef OS_LOAD_AVERAGE
3208 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3210 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3211 nomenclature these days.)
3213 PH/29 Installed patch from the Sieve maintainer that adds the options
3214 sieve_useraddress and sieve_subaddress to the redirect router.
3216 PH/30 In these circumstances:
3217 . Two addresses routed to the same list of hosts;
3218 . First host does not offer TLS;
3219 . First host accepts first address;
3220 . First host gives temporary error to second address;
3221 . Second host offers TLS and a TLS session is established;
3222 . Second host accepts second address.
3223 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3224 and peerdn, if requested) that were in fact used only for the second
3227 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3228 attention to any local part prefix or suffix that was matched by the
3229 router that accepted the address. It now behaves in the same way as it
3230 does for delivery: the affixes are removed from the local part unless
3231 rcpt_include_affixes is set on the transport.
3233 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3234 timeout during the DATA phase of an incoming message.
3236 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3237 applied a patch sanctioned by the Sieve maintainer.
3239 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3240 the uid or gid is negative. A case of a negative gid caused this to be
3241 noticed. The fix allows for either to be negative.
3243 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3244 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3247 PH/36 The change PH/12 above was broken. Fixed it.
3249 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3250 the grounds that routing the same address twice would always produce the
3251 same answer. This might have been true once, but it is certainly no
3252 longer true now. Routing a child address may depend on the previous
3253 routing that produced that child. Some complicated redirection strategies
3254 went wrong when messages had multiple recipients, and made Exim's
3255 behaviour dependent on the order in which the addresses were given.
3257 I have moved the duplicate checking until after the routing is complete.
3258 Exim scans the addresses that are assigned to local and remote
3259 transports, and removes any duplicates. This means that more work will be
3260 done, as duplicates will always all be routed, but duplicates are
3261 presumably rare, so I don't expect this is of any significance.
3263 For deliveries to pipes, files, and autoreplies, the duplicate checking
3264 still happens during the routing process, since they are not going to be
3267 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3268 It corrects a timeout issue with spamd. This is Ian's comment: "The
3269 background is that sometimes spamd either never reads data from a
3270 connection it has accepted, or it never writes response data. The exiscan
3271 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3272 blindly assumes that writes won't block so it may never time out."
3274 PH/39 Allow G after quota size as well as K and M.
3276 PH/40 The value set for $authenticated_id in an authenticator may not contain
3277 binary zeroes or newlines because the value is written to log lines and
3278 to spool files. There was no check on this. Now the value is run through
3279 the string_printing() function so that such characters are converted to
3280 printable escape sequences.
3282 PH/41 $message_linecount is a new variable that contains the total number of
3283 lines in the message. Compare $body_linecount, which is the count for the
3286 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3287 bounce and delay warning messages, except in certain special cases, which
3290 (a) An SMTP error message from a remote host;
3291 (b) A message specified in a :fail: redirection;
3292 (c) A message specified in a "fail" command in a system filter;
3293 (d) A message specified in a FAIL return from the queryprogram router;
3294 (e) A message specified by the cannot_route_message router option.
3296 In these cases only, Exim does include the error details in bounce and
3297 warning messages. There are also a few cases where bland messages such
3298 as "unrouteable address" or "local delivery error" are given.
3300 PH/43 $value is now also set for the "else" part of a ${run expansion.
3302 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3303 being worked on, but at least Exim now implements the latest version to
3306 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3307 process to complete was treated as a delivery failure, a timeout while
3308 writing the message to the pipe was logged, but erroneously treated as a
3309 successful delivery. Such timeouts include transport filter timeouts. For
3310 consistency with the overall process timeout, these timeouts are now
3311 treated as errors, giving rise to delivery failures by default. However,
3312 there is now a new Boolean option for the pipe transport called
3313 timeout_defer, which, if set TRUE, converts the failures into defers for
3314 both kinds of timeout. A transport filter timeout is now identified in
3317 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3318 systems where "make" and "gmake" are different, calling "gmake" at top
3319 level broke things. I've arranged for the value of $(MAKE) to be passed
3320 from the Makefile to this script so that it can call the same version of
3324 A note about Exim versions 4.44 and 4.50
3325 ----------------------------------------
3327 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3328 changes of various kinds. As a consequence, a big documentation update was
3329 needed. This delayed the release for rather longer than seemed good, especially
3330 in the light of a couple of (minor) security issues. Therefore, the changes
3331 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3332 release. So 4.44 and 4.50 are in effect two different branches that both start
3335 I have left the 4.50 change log unchanged; it contains all the changes since
3336 4.43. The change log for 4.44 is below; many of its items are identical to
3337 those for 4.50. This seems to be the most sensible way to preserve the
3338 historical information.
3344 1. Minor wording change to the doc/README.SIEVE file.
3346 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3347 computation of the current number of files was incorrect.
3349 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3350 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3353 4. Give more explanation in the error message when the command for a transport
3354 filter fails to execute.
3356 5. There are several places where Exim runs a non-Exim command in a
3357 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3358 was being done only for the command run by the queryprogram router. It is
3359 now done for all such subprocesses. The other cases are: ${run, transport
3360 filters, and the commands run by the lmtp and pipe transports.
3362 6. Added CONFIGURE_GROUP build-time option.
3364 7. Some older OS have a limit of 256 on the maximum number of file
3365 descriptors. Exim was using setrlimit() to set 1000 as a large value
3366 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3367 systems. I've change it so that if it can't get 1000, it tries for 256.
3369 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3370 was an oversight, and furthermore, ever since the addition of extra
3371 controls (e.g. 4.43/32), the checks on when to allow different forms of
3372 "control" were broken. There should now be diagnostics for all cases when a
3373 control that does not make sense is encountered.
3375 9. Added the /retain_sender option to "control=submission".
3377 10. $recipients is now available in the predata ACL (oversight).
3379 11. Tidy the search cache before the fork to do a delivery from a message
3380 received from the command line. Otherwise the child will trigger a lookup
3381 failure and thereby defer the delivery if it tries to use (for example) a
3382 cached ldap connection that the parent has called unbind on.
3384 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3385 of $address_data from the recipient verification was clobbered by the
3386 sender verification.
3388 13. The value of address_data from a sender verification is now available in
3389 $sender_address_data in subsequent conditions in the ACL statement.
3391 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3393 15. Added a new option "connect=<time>" to callout options, to set a different
3396 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3397 was its contents. (It was OK if the option was not defined at all.)
3399 17. A "Completed" log line is now written for messages that are removed from
3400 the spool by the -Mrm option.
3402 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3403 information about exactly what failed.
3405 19. Added -dd to debug only the daemon process.
3407 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3408 handles timeouts, both on the server side and network timeouts. Renamed the
3409 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3411 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3412 It is now set to "smtps".
3414 22. $host_address is now set to the target address during the checking of
3415 ignore_target_hosts.
3417 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3418 being passed; this would have caused $sender_host_name to have been used if
3419 matching the list had actually called for a host name (not very likely,
3420 since this list is usually IP addresses). A host name is now passed as
3423 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3424 code that specifies a non-restarting handler (typically sigaction() in
3425 modern systems) in an attempt to fix a rare and obscure crash bug.
3427 25. Narrowed the window for a race in the daemon that could cause it to ignore
3428 SIGCHLD signals. This is not a major problem, because they are used only to
3429 wake it up if nothing else does.
3431 26. A malformed maildirsize file could cause Exim to calculate negative values
3432 for the mailbox size or file count. Odd effects could occur as a result.
3433 The maildirsize information is now recalculated if the size or filecount
3436 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3437 support for a long time. Removed HAVE_SYS_VFS_H.
3439 28. Installed the latest version of exipick from John Jetmore.
3441 29. In an address list, if the pattern was not a regular expression, an empty
3442 subject address (from a bounce message) matched only if the pattern was an
3443 empty string. Non-empty patterns were not even tested. This was the wrong
3444 because it is perfectly reasonable to use an empty address as part of a
3445 database query. An empty address is now tested by patterns that are
3446 lookups. However, all the other forms of pattern expect the subject to
3447 contain a local part and a domain, and therefore, for them, an empty
3448 address still always fails if the pattern is not itself empty.
3450 30. Exim went into a mad DNS loop when attempting to do a callout where the
3451 host was specified on an smtp transport, and looking it up yielded more
3452 than one IP address.
3454 31. Re-factored the code for checking spool and log partition space into a
3455 function that finds that data and another that does the check. The former
3456 is then used to implement four new variables: $spool_space, $log_space,
3457 $spool_inodes, and $log_inodes.
3459 32. The RFC2047 encoding function was originally intended for short strings
3460 such as real names; it was not keeping to the 75-character limit for
3461 encoded words that the RFC imposes. It now respects the limit, and
3462 generates multiple encoded words if necessary. To be on the safe side, I
3463 have increased the buffer size for the ${rfc2047: expansion operator from
3466 33. It is now permitted to omit both strings after an "if" condition; if the
3467 condition is true, the result is "true". As before, when the second string
3468 is omitted, a false condition yields an empty string. This makes it less
3469 cumbersome to write custom ACL and router conditions.
3471 34. Failure to deliver a bounce message always caused it to be frozen, even if
3472 there was an errors_to setting on the router. The errors_to setting is now
3475 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3476 canonical form (fully expanded) before being placed in
3477 $sender_host_address.
3479 36. The table in the code that translates DNS record types into text (T_A to
3480 "A" for instance) was missing entries for NS and CNAME. It is just possible
3481 that this could have caused confusion if both these types were looked up
3482 for the same domain, because the text type is used as part of Exim's
3483 per-process caching. But the chance of anyone hitting this buglet seems
3486 37. The dnsdb lookup has been extended in a number of ways.
3488 (1) There is a new type, "zns", which walks up the domain tree until it
3489 finds some nameserver records. It should be used with care.
3491 (2) There is a new type, "mxh", which is like "mx" except that it returns
3492 just the host names, not the priorities.
3494 (3) It is now possible to give a list of domains (or IP addresses) to be
3495 looked up. The behaviour when one of the lookups defers can be
3496 controlled by a keyword.
3498 (4) It is now possible to specify the separator character for use when
3499 multiple records are returned.
3501 38. The dnslists ACL condition has been extended: it is now possible to supply
3502 a list of IP addresses and/or domains to be looked up in a particular DNS
3505 39. Added log_selector=+queue_time_overall.
3507 40. When running the queue in the test harness, wait just a tad after forking a
3508 delivery process, to get repeatability of debugging output.
3510 41. Include certificate and key file names in error message when GnuTLS fails
3511 to set them up, because the GnuTLS error message doesn't include the name
3512 of the failing file when there is a problem reading it.
3514 42. Allow both -bf and -bF in the same test run.
3516 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3518 44. The "Exiscan patch" is now merged into the mainline Exim source.
3520 45. Sometimes the final signoff response after QUIT could fail to get
3521 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3522 < 0 before doing a fflush(). This bug looks as though it goes back to the
3523 introduction of TLS in release 3.20, but "sometimes" must have been rare
3524 because the tests only now provoked it.
3526 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3527 (this can affect the format of dates).
3529 47. exim_tidydb, when checking for the continued existence of a message for
3530 which it has found a message-specific retry record, was not finding
3531 messages that were in split spool directories. Consequently, it was
3532 deleting retry records that should have stayed in existence.
3534 48. Steve fixed some bugs in eximstats.
3536 49. The SPA authentication driver was not abandoning authentication and moving
3537 on to the next authenticator when an expansion was forced to fail,
3538 contradicting the general specification for all authenticators. Instead it
3539 was generating a temporary error. It now behaves as specified.
3541 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3542 (the order specifies the preference for clients). The order is now AES256,
3543 AES128, 3DES, ARCFOUR128.
3545 51. Small patch to Sieve code - explicitly set From: when generating an
3548 52. Exim crashed if a remote delivery caused a very long error message to be
3549 recorded - for instance if somebody sent an entire SpamAssassin report back
3550 as a large number of 550 error lines. This bug was coincidentally fixed by
3551 increasing the size of one of Exim's internal buffers (big_buffer) that
3552 happened as part of the Exiscan merge. However, to be on the safe side, I
3553 have made the code more robust (and fixed the comments that describe what
3556 53. Now that there can be additional text after "Completed" in log lines (if
3557 the queue_time_overall log selector is set), a one-byte patch to exigrep
3558 was needed to allow it to recognize "Completed" as not the last thing in
3561 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3562 patch that reportedly fixes this has been added. I am not expert enough to
3563 create a test for it. This is what the patch creator wrote:
3565 "I found a little strange behaviour of ldap code when working with
3566 Windows 2003 AD Domain, where users was placed in more than one
3567 Organization Units. When I tried to give exim partial DN, the exit code
3568 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3569 But simultaneously result of request was absolutely normal ldap result,
3570 so I produce this patch..."
3572 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3573 REFERENCE, so I have modified the code to exclude the patch when that macro
3576 55. Some experimental protocols are using DNS PTR records for new purposes. The
3577 keys for these records are domain names, not reversed IP addresses. The
3578 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3579 leaves it alone. Component reversal etc. now happens only for IP addresses.
3582 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3584 57. Double the size of the debug message buffer (to 2048) so that more of very
3585 long debug lines gets shown.
3587 58. The exicyclog utility now does better if the number of log files to keep
3588 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3590 59. Two changes related to the smtp_active_hostname option:
3592 (1) $smtp_active_hostname is now available as a variable.
3593 (2) The default for smtp_banner uses $smtp_active_hostname instead
3594 of $primary_hostname.
3596 60. The host_aton() function is supposed to be passed a string that is known
3597 to be a valid IP address. However, in the case of IPv6 addresses, it was
3598 not checking this. This is a hostage to fortune. Exim now panics and dies
3599 if the condition is not met. A case was found where this could be provoked
3600 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3601 components; fortuitously, this particular loophole had already been fixed
3602 by change 4.50/55 above.
3604 If there are any other similar loopholes, the new check in host_aton()
3605 itself should stop them being exploited. The report I received stated that
3606 data on the command line could provoke the exploit when Exim was running as
3607 exim, but did not say which command line option was involved. All I could
3608 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3609 running as the user.
3612 61. There was a buffer overflow vulnerability in the SPA authentication code
3613 (which came originally from the Samba project). I have added a test to the
3614 spa_base64_to_bits() function which I hope fixes it.
3617 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3618 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3620 63. The daemon start-up calls getloadavg() while still root for those OS that
3621 need the first call to be done as root, but it missed one case: when
3622 deliver_queue_load_max is set with deliver_drop_privilege. This is
3623 necessary for the benefit of the queue runner, because there is no re-exec
3624 when deliver_drop_privilege is set.
3626 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3627 This has been fixed.
3629 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3630 was in use, was not putting the data itself into the right store pool;
3631 consequently, it could be overwritten for a subsequent message in the same
3632 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3635 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3637 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3638 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3639 the function were treating the return as a boolean value, which happened to
3640 work because 0=false and not-0=true, but is not correct code.
3642 68. The host_aton() function was not handling scoped IPv6 addresses (those
3643 with, for example, "%eth0" on the end) correctly.
3645 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3646 negated items (that is, ~something) in unsigned ints. Some compilers
3647 apparently mutter when there is no cast.
3649 70. If an address verification called from an ACL failed, and did not produce a
3650 user-specific message (i.e. there was only a "system" message), nothing was
3651 put in $acl_verify_message. In this situation, it now puts the system
3654 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3655 session; change 4.31/43 added the unwanted input to the log line - except
3656 that it did not do this in the start of session case. It now does.
3658 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3659 This is wrong and can cause the other end to generate a synchronization
3660 error if it is another Exim or anything else that does the synchronization
3661 check. A QUIT command is no longer sent after a timeout.
3663 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3664 during host lookups.
3666 74. The defer_ok option of callout verification was not working if it was used
3667 when verifying addresses in header lines, that is, for this case:
3669 verify = header_sender/callout=defer_ok
3671 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3672 those file descriptors could be used for SMTP connections. If anything
3673 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3674 could be sent to the SMTP client, causing chaos. The daemon now opens
3675 stdin, stdout, and stderr to /dev/null when it puts itself into the
3678 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3679 log by default. The user can override this with suitable Perl magic.
3681 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3682 the log message when discard triggers, was not working for the DATA ACL or
3683 for the non-SMTP ACL.
3685 78. Error message wording change in sieve.c.
3687 79. If smtp_accept_max_per_host was set, the number of connections could be
3688 restricted to fewer than expected, because the daemon was trying to set up
3689 a new connection before checking whether the processes handling previous
3690 connections had finished. The check for completed processes is now done
3691 earlier. On busy systems, this bug wouldn't be noticed because something
3692 else would have woken the daemon, and it would have reaped the completed
3695 80. If a message was submitted locally by a user whose login name contained one
3696 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3697 It caused a spool format error. I have fixed the spool reading code. A
3698 related problem was that the "from" clause in the Received: line became
3699 illegal because of the space(s). It is now covered by ${quote_local_part.
3701 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3704 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3706 83. Patch from Sieve maintainer to fix unterminated string problem in
3707 "vacation" handling.
3709 84. Some minor changes to the Linux configuration files to help with other
3710 OS variants using glibc.
3712 85. One more patch for Sieve to update vacation handling to latest spec.
3715 ----------------------------------------------------
3716 See the note above about the 4.44 and 4.50 releases.
3717 ----------------------------------------------------
3723 1. Change 4.43/35 introduced a bug that caused file counts to be
3724 incorrectly computed when quota_filecount was set in an appendfile
3727 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3728 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3731 3. Give more explanation in the error message when the command for a transport
3732 filter fails to execute.
3734 4. There are several places where Exim runs a non-Exim command in a
3735 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3736 was being done only for the command run by the queryprogram router. It is
3737 now done for all such subprocesses. The other cases are: ${run, transport
3738 filters, and the commands run by the lmtp and pipe transports.
3740 5. Some older OS have a limit of 256 on the maximum number of file
3741 descriptors. Exim was using setrlimit() to set 1000 as a large value
3742 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3743 systems. I've change it so that if it can't get 1000, it tries for 256.
3745 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3746 was an oversight, and furthermore, ever since the addition of extra
3747 controls (e.g. 4.43/32), the checks on when to allow different forms of
3748 "control" were broken. There should now be diagnostics for all cases when a
3749 control that does not make sense is encountered.
3751 7. $recipients is now available in the predata ACL (oversight).
3753 8. Tidy the search cache before the fork to do a delivery from a message
3754 received from the command line. Otherwise the child will trigger a lookup
3755 failure and thereby defer the delivery if it tries to use (for example) a
3756 cached ldap connection that the parent has called unbind on.
3758 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3759 of $address_data from the recipient verification was clobbered by the
3760 sender verification.
3762 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3763 was its contents. (It was OK if the option was not defined at all.)
3765 11. A "Completed" log line is now written for messages that are removed from
3766 the spool by the -Mrm option.
3768 12. $host_address is now set to the target address during the checking of
3769 ignore_target_hosts.
3771 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3772 being passed; this would have caused $sender_host_name to have been used if
3773 matching the list had actually called for a host name (not very likely,
3774 since this list is usually IP addresses). A host name is now passed as
3777 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3778 code that specifies a non-restarting handler (typically sigaction() in
3779 modern systems) in an attempt to fix a rare and obscure crash bug.
3781 15. Narrowed the window for a race in the daemon that could cause it to ignore
3782 SIGCHLD signals. This is not a major problem, because they are used only to
3783 wake it up if nothing else does.
3785 16. A malformed maildirsize file could cause Exim to calculate negative values
3786 for the mailbox size or file count. Odd effects could occur as a result.
3787 The maildirsize information is now recalculated if the size or filecount
3790 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3791 support for a long time. Removed HAVE_SYS_VFS_H.
3793 18. Updated exipick to current release from John Jetmore.
3795 19. Allow an empty sender to be matched against a lookup in an address list.
3796 Previously the only cases considered were a regular expression, or an
3799 20. Exim went into a mad DNS lookup loop when doing a callout where the
3800 host was specified on the transport, if the DNS lookup yielded more than
3803 21. The RFC2047 encoding function was originally intended for short strings
3804 such as real names; it was not keeping to the 75-character limit for
3805 encoded words that the RFC imposes. It now respects the limit, and
3806 generates multiple encoded words if necessary. To be on the safe side, I
3807 have increased the buffer size for the ${rfc2047: expansion operator from
3810 22. Failure to deliver a bounce message always caused it to be frozen, even if
3811 there was an errors_to setting on the router. The errors_to setting is now
3814 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3815 canonical form (fully expanded) before being placed in
3816 $sender_host_address.
3818 24. Updated eximstats to version 1.33
3820 25. Include certificate and key file names in error message when GnuTLS fails
3821 to set them up, because the GnuTLS error message doesn't include the name
3822 of the failing file when there is a problem reading it.
3824 26. Expand error message when OpenSSL has problems setting up cert/key files.
3827 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3828 (this can affect the format of dates).
3830 28. exim_tidydb, when checking for the continued existence of a message for
3831 which it has found a message-specific retry record, was not finding
3832 messages that were in split spool directories. Consequently, it was
3833 deleting retry records that should have stayed in existence.
3835 29. eximstats updated to version 1.35
3836 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3837 1.35 - bugfix such that pie charts by volume are generated correctly
3839 30. The SPA authentication driver was not abandoning authentication and moving
3840 on to the next authenticator when an expansion was forced to fail,
3841 contradicting the general specification for all authenticators. Instead it
3842 was generating a temporary error. It now behaves as specified.
3844 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3845 (the order specifies the preference for clients). The order is now AES256,
3846 AES128, 3DES, ARCFOUR128.
3848 31. Small patch to Sieve code - explicitly set From: when generating an
3851 32. Exim crashed if a remote delivery caused a very long error message to be
3852 recorded - for instance if somebody sent an entire SpamAssassin report back
3853 as a large number of 550 error lines. This bug was coincidentally fixed by
3854 increasing the size of one of Exim's internal buffers (big_buffer) that
3855 happened as part of the Exiscan merge. However, to be on the safe side, I
3856 have made the code more robust (and fixed the comments that describe what
3859 33. Some experimental protocols are using DNS PTR records for new purposes. The
3860 keys for these records are domain names, not reversed IP addresses. The
3861 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3862 leaves it alone. Component reversal etc. now happens only for IP addresses.
3865 34. The host_aton() function is supposed to be passed a string that is known
3866 to be a valid IP address. However, in the case of IPv6 addresses, it was
3867 not checking this. This is a hostage to fortune. Exim now panics and dies
3868 if the condition is not met. A case was found where this could be provoked
3869 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3870 components; fortuitously, this particular loophole had already been fixed
3871 by change 4.50/55 or 4.44/33 above.
3873 If there are any other similar loopholes, the new check in host_aton()
3874 itself should stop them being exploited. The report I received stated that
3875 data on the command line could provoke the exploit when Exim was running as
3876 exim, but did not say which command line option was involved. All I could
3877 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3878 running as the user.
3881 35. There was a buffer overflow vulnerability in the SPA authentication code
3882 (which came originally from the Samba project). I have added a test to the
3883 spa_base64_to_bits() function which I hope fixes it.
3886 36. The daemon start-up calls getloadavg() while still root for those OS that
3887 need the first call to be done as root, but it missed one case: when
3888 deliver_queue_load_max is set with deliver_drop_privilege. This is
3889 necessary for the benefit of the queue runner, because there is no re-exec
3890 when deliver_drop_privilege is set.
3892 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3893 was in use, was not putting the data itself into the right store pool;
3894 consequently, it could be overwritten for a subsequent message in the same
3895 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3898 38. Sometimes the final signoff response after QUIT could fail to get
3899 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3900 < 0 before doing a fflush(). This bug looks as though it goes back to the
3901 introduction of TLS in release 3.20, but "sometimes" must have been rare
3902 because the tests only now provoked it.
3908 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3909 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3910 Now it returns an integer. A number of calls were still expecting a T/F
3911 return. Fortuitously, in all cases, the tests worked in OK situations,
3912 which is the norm. However, things would have gone wrong on any write
3913 failures on the smtp file descriptor. This function is used when sending
3914 messages over SMTP and also when doing verify callouts.
3916 2. When Exim is called to do synchronous delivery of a locally submitted
3917 message (the -odf or -odi options), it no longer closes stderr before doing
3920 3. Implemented the mua_wrapper option.
3922 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3924 5. Implemented the functions header_remove(), header_testname(),
3925 header_add_at_position(), and receive_remove_recipient(), and exported them
3928 6. If an ACL "warn" statement specified the addition of headers, Exim already
3929 inserted X-ACL-Warn: at the start if there was no header name. However, it
3930 was not making this test for the second and subsequent header lines if
3931 there were newlines in the string. This meant that an invalid header could
3932 be inserted if Exim was badly configured.
3934 7. Allow an ACL "warn" statement to add header lines at the start or after all
3935 the Received: headers, as well as at the end.
3937 8. Added the rcpt_4xx retry error code.
3939 9. Added postmaster_mailfrom=xxx to callout verification option.
3941 10. Added mailfrom=xxxx to the callout verification option, for verify=
3944 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3945 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3947 12. Inserted some casts to stop certain compilers warning when using pointer
3948 differences as field lengths or precisions in printf-type calls (mostly
3949 affecting debugging statements).
3951 13. Added optional readline() support for -be (dynamically loaded).
3953 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3954 same clock tick as a message's arrival, so that its received time was the
3955 same as the "first fail" time on the retry record, and that message
3956 remained on the queue past the ultimate address timeout, every queue runner
3957 would try a delivery (because it was past the ultimate address timeout) but
3958 after another failure, the ultimate address timeout, which should have then
3959 bounced the address, did not kick in. This was a "< instead of <=" error;
3960 in most cases the first failure would have been in the next clock tick
3961 after the received time, and all would be well.
3963 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3964 being recognized when the domain list was tested by the match_domain
3965 condition in an expansion string.
3967 16. Added the ${str2b64: operator.
3969 17. Exim was always calling setrlimit() to set a large limit for the number of
3970 processes, without checking whether the existing limit was already
3971 adequate. (It did check for the limit on file descriptors.) Furthermore,
3972 errors from getrlimit() and setrlimit() were being ignored. Now they are
3973 logged to the main and panic logs, but Exim does carry on, to try to do its
3974 job under whatever limits there are.
3976 18. Imported PCRE 5.0.
3978 19. Trivial typo in log message " temporarily refused connection" (the leading
3981 20. If the log selector return_path_on_delivery was set and an address was
3982 redirected to /dev/null, the delivery process crashed because it assumed
3983 that a return path would always be set for a "successful" delivery. In this
3984 case, the whole delivery is bypassed as an optimization, and therefore no
3987 21. Internal re-arrangement: the function for sending a challenge and reading
3988 a response while authentication was assuming a zero-terminated challenge
3989 string. It's now changed to take a pointer and a length, to allow for
3990 binary data in such strings.
3992 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3994 23. Exim was not respecting finduser_retries when seeking the login of the
3995 uid under which it was called; it was always trying 10 times. (The default
3996 setting of finduser_retries is zero.) Also, it was sleeping after the final
3997 failure, which is pointless.
3999 24. Implemented tls_on_connect_ports.
4001 25. Implemented acl_smtp_predata.
4003 26. If the domain in control=submission is set empty, Exim assumes that the
4004 authenticated id is a complete email address when it generates From: or
4005 Sender: header lines.
4007 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4008 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4009 chown and chgrp in /bin and hostname in /usr/bin.
4011 28. Exim was keeping the "process log" file open after each use, just as it
4012 does for the main log. This opens the possibility of it remaining open for
4013 long periods when the USR1 signal hits a daemon. Occasional processlog
4014 errors were reported, that could have been caused by this. Anyway, it seems
4015 much more sensible not to leave this file open at all, so that is what now
4018 29. The long-running daemon process does not normally write to the log once it
4019 has entered its main loop, and it closes the log before doing so. This is
4020 so that log files can straightforwardly be renamed and moved. However,
4021 there are a couple of unusual error situations where the daemon does write
4022 log entries, and I had neglected to close the log afterwards.
4024 30. The text of an SMTP error response that was received during a remote
4025 delivery was being truncated at 512 bytes. This is too short for some of
4026 the long messages that one sometimes sees. I've increased the limit to
4029 31. It is now possible to make retry rules that apply only when a message has a
4030 specific sender, in particular, an empty sender.
4032 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4033 it possible to be selective about when SMTP synchronization is enforced.
4035 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4037 32. Implemented hosts_connection_nolog.
4039 33. Added an ACL for QUIT.
4041 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4044 35. Added mailbox_size and mailbox_filecount to appendfile.
4046 36. Added control = no_multiline_responses to ACLs.
4048 37. There was a bug in the logic of the code that waits for the clock to tick
4049 in the case where the clock went backwards by a substantial amount such
4050 that the microsecond fraction of "now" was more than the microsecond
4051 fraction of "then" (but the whole seconds number was less).
4053 38. Added support for the libradius Radius client library this is found on
4054 FreeBSD (previously only the radiusclient library was supported).
4060 1. When certain lookups returned multiple values in the form name=value, the
4061 quoting of the values was not always being done properly. Specifically:
4062 (a) If the value started with a double quote, but contained no whitespace,
4064 (b) If the value contained whitespace other than a space character (i.e.
4065 tabs or newlines or carriage returns) it was not quoted.
4066 This fix has been applied to the mysql and pgsql lookups by writing a
4067 separate quoting function and calling it from the lookup code. The fix
4068 should probably also be applied to nisplus, ibase and oracle lookups, but
4069 since I cannot test any of those, I have not disturbed their existing code.
4071 2. A hit in the callout cache for a specific address caused a log line with no
4072 reason for rejecting RCPT. Now it says "Previous (cached) callout
4073 verification failure".
4075 3. There was an off-by-one bug in the queryprogram router. An over-long
4076 return line was truncated at 256 instead of 255 characters, thereby
4077 overflowing its buffer with the terminating zero. As well as fixing this, I
4078 have increased the buffer size to 1024 (and made a note to document this).
4080 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4081 when Exim is waiting for an SMTP response from a remote server, Exim
4082 restarts its select() call on the socket, thereby resetting its timeout.
4083 This is not a problem when such interrupts are rare. Somebody set up a cron
4084 job to run exiwhat every 2 minutes, which is less than the normal select()
4085 timeout (5 or 10 minutes). This meant that the select() timeout never
4086 kicked in because it was always reset. I have fixed this by comparing the
4087 time when an interrupt arrives with the time at the start of the first call
4088 to select(). If more time than the timeout has elapsed, the interrupt is
4089 treated as a timeout.
4091 5. Some internal re-factoring in preparation for the addition of Sieve
4092 extensions (by MH). In particular, the "personal" test is moved to a
4093 separate function, and given an option for scanning Cc: and Bcc: (which is
4094 not set for Exim filters).
4096 6. When Exim created an email address using the login of the caller as the
4097 local part (e.g. when creating a From: or Sender: header line), it was not
4098 quoting the local part when it contained special characters such as @.
4100 7. Installed new OpenBSD configuration files.
4102 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4103 try to make them clearer.
4105 9. Callout options, other than the timeout value, were being ignored when
4106 verifying sender addresses in header lines. For example, when using
4108 verify = header_sender/callout=no_cache
4110 the cache was (incorrectly) being used.
4112 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4113 only the Cygwin environment.
4115 11. When return_path_on_delivery was set as a log selector, if different remote
4116 addresses in the same message used different return paths and parallel
4117 remote delivery occurred, the wrong values would sometimes be logged.
4118 (Whenever a remote delivery process finished, the return path value from
4119 the most recently started remote delivery process was logged.)
4121 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4122 lines when AUTH and/or TLS are in use. This is the "received protocol"
4123 field. Exim used to use "asmtp" for authenticated SMTP, without any
4124 indication (in the protocol name) for TLS use. Now it follows the RFC and
4125 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4126 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4127 names appear in log lines as well as in Received: header lines.
4129 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4130 extensions, and comparison tests, and to fix some bugs.
4132 14. Changes to the "personal" filter test:
4134 (1) The test was buggy in that it was just doing the equivalent of
4135 "contains" tests on header lines. For example, if a user's address was
4136 anne@some.where, the "personal" test would incorrectly be true for
4138 To: susanne@some.where
4140 This test is now done by extracting each address from the header in turn,
4141 and checking the entire address. Other tests that are part of "personal"
4142 are now done using regular expressions (for example, to check local parts
4143 of addresses in From: header lines).
4145 (2) The list of non-personal local parts in From: addresses has been
4146 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4147 taken from the Sieve specification recommendations.
4149 (3) If the message contains any header line starting with "List-" it is
4150 treated as non-personal.
4152 (4) The test for "circular" in the Subject: header line has been removed
4153 because it now seems ill-conceived.
4155 15. Minor typos in src/EDITME comments corrected.
4157 16. Installed latest exipick from John Jetmore.
4159 17. If headers_add on a router specified a text string that was too long for
4160 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4161 of string_sprintf() is now avoided.
4163 18. $message_body_size was not set (it was always zero) when running the DATA
4164 ACL and the local_scan() function.
4166 19. For the "mail" command in an Exim filter, no default was being set for
4167 the once_repeat time, causing a random time value to be used if "once" was
4168 specified. (If the value happened to be <= 0, no repeat happened.) The
4169 default is now 0s, meaning "never repeat". The "vacation" command was OK
4170 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4171 (I found it when inspecting the code).
4173 20. There is now an overall timeout for performing a callout verification. It
4174 defaults to 4 times the callout timeout, which applies to individual SMTP
4175 commands during the callout. The overall timeout applies when there is more
4176 than one host that can be tried. The timeout is checked before trying the
4177 next host. This prevents very long delays if there are a large number of
4178 hosts and all are timing out (e.g. when the network connections are timing
4179 out). The value of the overall timeout can be changed by specifying an
4180 additional sub-option for "callout", called "maxwait". For example:
4182 verify = sender/callout=5s,maxwait=20s
4184 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4185 to the end before writing, but this should make it even safer).
4187 22. Exim was forgetting that it had advertised PIPELINING for the second and
4188 subsequent messages on an SMTP connection. It was also not resetting its
4189 memory on STARTTLS and an internal HELO.
4191 23. When Exim logs an SMTP synchronization error within a session, it now
4192 records whether PIPELINING has been advertised or not.
4194 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4195 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4196 rather than long int.
4198 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4200 26. Added the never_mail option to autoreply.
4206 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4207 crash if the getsockname() call failed; this can happen if a connection is
4208 closed very soon after it is established. The problem was simply in the
4209 order in which certain operations were done, causing Exim to try to write
4210 to the SMTP stream before it had set up the file descriptor. The bug has
4211 been fixed by making things happen in the correct order.
4217 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4218 before the connection was closed, thus losing the rejection response.
4220 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4221 some early Solaris releases, but causes trouble in current releases where
4222 socklen_t is defined.
4224 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4227 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4230 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4231 that releasing the top of it at the end releases what was used for sub-
4232 expansions (unless the block got too big). However, discard this block if
4233 the first thing is a variable or header, so that we can use its block when
4234 it is dynamic (useful for very large $message_headers, for example).
4236 6. Lookups now cache *every* query, not just the most recent. A new, separate
4237 store pool is used for this. It can be recovered when all lookup caches are
4238 flushed. Lookups now release memory at the end of their result strings.
4239 This has involved some general refactoring of the lookup sources.
4241 7. Some code has been added to the store_xxx() functions to reduce the amount
4242 of flapping under certain conditions.
4244 8. log_incoming_interface used to affect only the <= reception log lines. Now
4245 it causes the local interface and port to be added to several more SMTP log
4246 lines, for example "SMTP connection from", and rejection lines.
4248 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4250 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4252 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4253 could be overwritten at the end of the current message (or the start of a
4254 new message if it was set in a HELO ACL). The value is now preserved for
4255 the duration of the SMTP connection.
4257 12. If a transport had a headers_rewrite setting, and a matching header line
4258 contained an unqualified address, that address was qualified, even if it
4259 did not match any rewriting rules. The underlying bug was that the values
4260 of the flags that permit the existence of unqualified sender and recipient
4261 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4262 non-local messages, and by -bnq for local messages) were not being
4263 preserved with the message after it was received.
4265 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4266 "next input=" as part of the text comprising the host identity instead of
4267 the correct text. The code was using the same buffer for two different
4268 strings. However, depending on which order the printing function evaluated
4269 its arguments, the bug did not always show up. Under Linux, for example, my
4270 test suite worked just fine.
4272 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4273 doesn't work with some older versions of Perl. It has been changed to "my",
4274 which in any case is probably the better facility to use.
4276 15. A really picky compiler found some instances of statements for creating
4277 error messages that either had too many or two few arguments for the format
4280 16. The size of the buffer for calls to the DNS resolver has been increased
4281 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4282 for addresses that have a lot of PTR records. This alleviates a problem; it
4283 does not fully solve it.
4285 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4286 buffer now truncates the list and logs the incident, which is the same
4287 action as happens when Exim is looking up a host name and its aliases.
4288 Previously in this situation something unpredictable would happen;
4289 sometimes it was "internal error: store_reset failed".
4291 18. If a server dropped the connection unexpectedly when an Exim client was
4292 using GnuTLS and trying to read a response, the client delivery process
4293 crashed while trying to generate an error log message.
4295 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4296 string, for example:
4298 warn message = H1: something\nH2: something
4300 the text was added as a single header line from Exim's point of view
4301 though it ended up OK in the delivered message. However, searching for the
4302 second and subsequent header lines using $h_h2: did not work. This has been
4303 fixed. Similarly, if a system filter added multiple headers in this way,
4304 the routers could not see them.
4306 20. Expanded the error message when iplsearch is called with an invalid key to
4307 suggest using net-iplsearch in a host list.
4309 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4310 ACLs are no longer actually imposed (and a message to that effect is
4313 22. If a "gecos" field in a passwd entry contained escaped characters, in
4314 particular, if it contained a \" sequence, Exim got it wrong when building
4315 a From: or a Sender: header from that name. A second bug also caused
4316 incorrect handling when an unquoted " was present following a character
4317 that needed quoting.
4319 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4320 was not being matched caselessly.
4322 24. Arranged for all hyphens in the exim.8 source to be escaped with
4325 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4326 properly. Recipient callout cache records were still being keyed to include
4327 the sender, even when use_sender was set false. This led to far more
4328 callouts that were necessary. The sender is no longer included in the key
4329 when use_sender is false.
4331 26. Added "control = submission" modifier to ACLs.
4333 27. Added the ${base62d: operator to decode base 62 numbers.
4335 28. dnsdb lookups can now access SRV records.
4337 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4338 the configuration file.
4340 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4341 (-v) mode. This makes the output for a verbose queue run more intelligible.
4343 31. Added a use_postmaster feature to recipient callouts.
4345 32. Added the $body_zerocount variable, containing the number of binary zero
4346 bytes in the message body.
4348 33. The time of last modification of the "new" subdirectory is now used as the
4349 "mailbox time last read" when there is a quota error for a maildir
4352 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4354 35. Added +ignore_unknown as a special item in host lists.
4356 36. Code for decoding IPv6 addresses in host lists is now included, even if
4357 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4358 address was recognized as an IP address, but was then not correctly decoded
4359 into binary, causing unexpected and incorrect effects when compared with
4366 1. Very minor rewording of debugging text in manualroute to say "list of
4367 hosts" instead of "hostlist".
4369 2. If verify=header_syntax was set, and a header line with an unqualified
4370 address (no domain) and a large number of spaces between the end of the
4371 name and the colon was received, the reception process suffered a buffer
4372 overflow, and (when I tested it) crashed. This was caused by some obsolete
4373 code that should have been removed. The fix is to remove it!
4375 3. When running in the test harness, delay a bit after writing a bounce
4376 message to get a bit more predictability in the log output.
4378 4. Added a call to search_tidyup() just before forking a reception process. In
4379 theory, someone could use a lookup in the expansion of smtp_accept_max_
4380 per_host which, without the tidyup, could leave open a database connection.
4382 5. Added the variables $recipient_data and $sender_data which get set from a
4383 lookup success in an ACL "recipients" or "senders" condition, or a router
4384 "senders" option, similar to $domain_data and $local_part_data.
4386 6. Moved the writing of debug_print from before to after the "senders" test
4389 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4390 problems for message scanning, either using a data ACL, or using
4391 local_scan() because the Received: header was not generated till after they
4392 were called (in order to set the time as the time of reception completion).
4393 I have revised the way this works. The header is now generated after the
4394 body is received, but before the ACL or local_scan() are called. After they
4395 are run, the timestamp in the header is updated.
4401 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4402 before starting a queue runner without re-exec. This happened only when
4403 deliver_drop_privilege was set or when the Exim user was set to root. The
4404 effect of the bug was that timeouts during subsequent deliveries caused
4405 crashes instead of being properly handled. The handler is now left at its
4406 default (and expected) setting.
4408 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4409 message, again when deliver_drop_privilege is set or Exim is run as root.
4410 The bug described in (1) was not present in this case, but the tidying up
4411 of the other signals was missing. I have made the two cases consistent.
4413 3. The ignore_target_hosts setting on a manualroute router was being ignored
4414 for hosts that were looked up using the /MX notation.
4416 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4419 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4420 operated on the sender address. After changing the $sender_address to <>
4421 for the sender address verify, Exim was re-instated it as the original
4422 (before rewriting) address, but remembering that it had rewritten it, so it
4423 wasn't rewriting it again. This bug also had the effect of breaking the
4424 sender address verification caching when the sender address was rewritten.
4426 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4427 This has been changed so that if the ip literal address matches
4428 ignore_target_hosts, the router declines.
4430 7. Added expansion conditions match_domain, match_address, and match_local_
4431 part (NOT match_host).
4433 8. The placeholder for the Received: header didn't have a length field set.
4435 9. Added code to Exim itself and to exim_lock to test for a specific race
4436 condition that could lead to file corruption when using MBX delivery. The
4437 issue is with the lockfile that is created in /tmp. If this file is removed
4438 after a process has opened it but before that process has acquired a lock,
4439 there is the potential for a second process to recreate the file and also
4440 acquire a lock. This could lead to two Exim processes writing to the file
4441 at the same time. The added code performs the same test as UW imapd; it
4442 checks after acquiring the lock that its file descriptor still refers to
4443 the same named file.
4445 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4446 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4449 11. Added the smtp_active_hostname option. If used, this will typically be made
4450 to depend on the incoming interface address. Because $interface_address is
4451 not set up until the daemon has forked a reception process, error responses
4452 that can happen earlier (such as "too many connections") no longer contain
4455 12. If an expansion in a condition on a "warn" statement fails because a lookup
4456 defers, the "warn" statement is abandoned, and the next ACL statement is
4457 processed. Previously this caused the whole ACL to be aborted.
4459 13. Added the iplsearch lookup type.
4461 14. Added ident_timeout as a log selector.
4463 15. Added tls_certificate_verified as a log selector.
4465 16. Added a global option tls_require_ciphers (compare the smtp transport
4466 option of the same name). This controls incoming TLS connections.
4468 17. I finally figured out how to make tls_require_ciphers do a similar thing
4469 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4470 before starting the TLS session.
4472 18. Tabs are now shown as \t in -bP output.
4474 19. If the log selector return_path_on_delivery was set, Exim crashed when
4475 bouncing a message because it had too many Received: header lines.
4477 20. If two routers both had headers_remove settings, and the first one included
4478 a superfluous trailing colon, the final name in the first list and the
4479 first name in the second list were incorrectly joined into one item (with a
4480 colon in the middle).
4486 1. Added -C and -D options to the exinext utility, mainly to make it easier
4487 to include in the automated testing, but these could be helpful when
4488 multiple configurations are in use.
4490 2. The exinext utility was not formatting the output nicely when there was
4491 an alternate port involved in the retry record key, nor when there was a
4492 message id as well (for retries that were specific to a specific message
4493 and a specific host). It was also confused by IPv6 addresses, because of
4494 the additional colons they contain. I have fixed the IPv4 problem, and
4495 patched it up to do a reasonable job for IPv6.
4497 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4498 delivery, the log line now contains "pipelined" if PIPELINING was used.
4500 4. An SMTP transport process used to panic and die if the bind() call to set
4501 an explicit outgoing interface failed. This has been changed; it is now
4502 treated in the same way as a connect() failure.
4504 5. A reference to $sender_host_name in the part of a conditional expansion
4505 that was being skipped was still causing a DNS lookup. This no longer
4508 6. The def: expansion condition was not recognizing references to header lines
4509 that used bh_ and bheader_.
4511 7. Added the _cache feature to named lists.
4513 8. The code for checking quota_filecount in the appendfile transport was
4514 allowing one more file than it should have been.
4516 9. For compatibility with Sendmail, the command line option
4524 and sets the incoming protocol and host name (for trusted callers). The
4525 host name and its colon can be omitted when only the protocol is to be set.
4526 Note the Exim already has two private options, -pd and -ps, that refer to
4527 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4528 "s", but I don't think that's a major issue.
4530 10. A number of refactoring changes to the code, none of which should affect
4533 (a) The number of logging options was getting close to filling up the
4534 32-bit word that was used as a bit map. I have split them into two classes:
4535 those that are passed in the argument to log_write(), and those that are
4536 only ever tested independently outside of that function. These are now in
4537 separate 32-bit words, so there is plenty of room for expansion again.
4538 There is no change in the user interface or the logging behaviour.
4540 (b) When building, for example, log lines, the code previously used a
4541 macro that called string_cat() twice, in order to add two strings. This is
4542 not really sufficiently general. Furthermore, there was one instance where
4543 it was actually wrong because one of the argument was used twice, and in
4544 one call a function was used. (As it happened, calling the function twice
4545 did not affect the overall behaviour.) The macro has been replaced by a
4546 function that can join an arbitrary number of extra strings onto a growing
4549 (c) The code for expansion conditions now uses a table and a binary chop
4550 instead of a serial search (which was left over from when there were very
4551 few conditions). Also, it now recognizes conditions like "pam" even when
4552 the relevant support is not compiled in: a suitably worded error message is
4553 given if an attempt is made to use such a condition.
4555 11. Added ${time_interval:xxxxx}.
4557 12. A bug was causing one of the ddress fields not to be passed back correctly
4558 from remote delivery subprocesses. The field in question was not being
4559 subsequently used, so this caused to problems in practice.
4561 13. Added new log selectors queue_time and deliver_time.
4563 14. Might have fixed a bug in maildirsizefile handling that threw up
4564 "unexpected character" debug warnings, and recalculated the data
4565 unnecessarily. In any case, I expanded the warning message to give more
4568 15. Added the message "Restricted characters in address" to the statements in
4569 the default ACL that block characters like @ and % in local parts.
4571 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4572 Three changes have been made:
4574 (a) There was a serious bug; a negative response to MAIL caused the whole
4575 recipient domain to be cached as invalid, thereby blocking all messages
4576 to all local parts at the same domain, from all senders. This bug has
4577 been fixed. The domain is no longer cached after a negative response to
4578 MAIL if the sender used is not empty.
4580 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4583 (c) A new callout option, "use_sender" has been added for people who want
4584 the modified behaviour.
4590 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4593 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4594 indeed breaks things for older releases.
4596 3. Added additional logging to the case where there is a problem reading data
4597 from a filter that is running in a subprocess using a pipe, in order to
4598 try to track down a specific problem.
4600 4. Testing facility fudge: when running in the test harness and attempting
4601 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4602 getting "No route to host". Convert this to a timeout.
4604 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4607 6. Some OS don't have socklen_t but use size_t instead. This affects the
4608 fifth argument of getsockopt() amongst other things. This is now
4609 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4610 can be set for individual OS. I have set it for SunOS5, OSF1, and
4611 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4612 some earlier ones do not.
4614 7. Change 4.30/15 was not doing the test caselessly.
4616 8. The standard form for an IPv6 address literal was being rejected by address
4617 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4618 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4619 this, as well as the form without the "IPv6" on the front (but only when
4620 address literals are enabled, of course).
4622 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4624 10. Exim crashed if a message with an empty sender address specified by -f
4625 encountered a router with an errors_to setting. This could be provoked only
4626 by a command such as
4630 where an empty string was supplied; "<>" did not hit this bug.
4632 11. Installed PCRE release 4.5.
4634 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4635 remained set. It is now erased.
4637 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4638 times from message ids (which are base 36 rather than the normal 62).
4640 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4641 were being counted as actual protocol errors, and logged if the log
4642 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4643 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4644 and DATA following a set of rejected RCPTs do not count as protocol errors.
4645 In other words, Exim assumes they were pipelined, though this may not
4646 actually be the case. Of course, in all cases the client gets an
4647 appropriate error code.
4649 15. If a lookup fails in an ACL condition, a message about the failure may
4650 be available; it is used if testing the ACL cannot continue, because most
4651 such messages specify what the cause of the deferral is. However, some
4652 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4653 that caused an old message to be retained and used if a later statement
4654 caused a defer, replacing the real cause of the deferral.
4656 16. If an IP address had so many PTR records that the DNS lookup buffer
4657 was not large enough to hold them, Exim could crash while trying to process
4658 the truncated data. It now detects and logs this case.
4660 17. Further to 4.21/58, another change has been made: if (and only if) the
4661 first line of a message (the first header line) ends with CRLF, a bare LF
4662 in a subsequent header line has a space inserted after it, so as not to
4663 terminate the header.
4665 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4666 unnecessarily, used atoi() instead of strtol(), and didn't check the
4667 termination when getting file sizes from file names by regex.
4669 19. Completely re-implemented the support for maildirsize files, in the light
4670 of a number of problems with the previous contributed implementation
4671 (4.30/29). In particular:
4673 . If the quota is zero, the maildirsize file is maintained, but no quota is
4676 . If the maildir directory does not exist, it is created before any attempt
4677 to write a maildirsize file.
4679 . The quota value in the file is just a cache; if the quota is changed in
4680 the transport, the new value overrides.
4682 . A regular expression is available for excluding directories from the
4685 20. The autoreply transport checks the characters in options that define the
4686 message's headers; it allows continued headers, but it was checking with
4687 isspace() after an embedded newline instead of explicitly looking for a
4690 21. If all the "regular" hosts to which an address was routed had passed their
4691 expiry times, and had not reached their retry times, the address was
4692 bounced, even if fallback hosts were defined. Now Exim should go on to try
4695 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4696 equivalent code in the SMTP transport. Some hosts send humungous responses
4697 to HELO/EHLO, more than 1024 it seems.
4699 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4700 gives compiler warnings in some environments. I've now done it "properly",
4703 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4704 (because of problems with the built-in one) was declared to return uschar *
4705 instead of char *, causing compiler failure.
4707 25. Fixed a file descriptor leak when processing alias/forward files.
4709 26. Fixed a minor format string issue in dbfn.c.
4711 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4713 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4714 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4717 29. When a sender address is verified, it is cached, to save repeating the test
4718 when there is more than one recipient in a message. However, when the
4719 verification involves a callout, it is possible for different callout
4720 options to be set for different recipients. It is too complicated to keep
4721 track of this in the cache, so now Exim always runs a verification when a
4722 callout is required, relying on the callout cache for the optimization.
4723 The overhead is duplication of the address routing, but this should not be
4726 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4727 to be verified with callout=postmaster, and the main callout worked but the
4728 postmaster check failed, the verification correctly failed. However, if a
4729 subsequent RCPT command asked for sender verification *without* the
4730 postmaster check, incorrect caching caused this verification also to fail,
4733 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4734 it was not caching the DNS options (qualify_single, search_parents) that
4735 were used when the lookup failed. A subsequent lookup with different
4736 options therefore always gave the same answer, though there were cases
4737 where it should not have. (Example: a "domains = !$mx_any" option on a
4738 dnslookup router: the "domains" option is always processed without any
4739 widening, but the router might have qualify_single set.) Now Exim uses the
4740 cached value only when the same options are set.
4742 32. Added John Jetmore's "exipick" utility to the distribution.
4744 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4745 than a timeout (e.g. a certificate is required, and is not provided), an
4746 Exim server now closes the connection immediately. Previously it waited for
4747 the client to close - but if the client is SSL, it seems that they each
4748 wait for each other, leading to a delay before one of them times out.
4750 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4751 maintained 0.8.x compatibility because I don't think many are using it, and
4752 it is clearly obsolete.
4754 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4757 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4758 client certificate was expired. A simple patch fixes this, though I don't
4759 understand the full logic of why the verify callback is called multiple
4762 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4763 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4764 which causes problems with some clients (such as the Certicom SSL Plus
4765 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4766 disables the coutermeasure allowing Eudora to connect."
4768 38. Exim was not checking that a write() to a log file succeeded. This could
4769 lead to Bad Things if a log got too big, in particular if it hit a file
4770 size limit. Exim now panics and dies if it cannot write to a log file, just
4771 as it does if it cannot open a log file.
4773 39. Modified OS/Makefile-Linux so that it now contains
4775 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4777 The two -D definitions ensure that Exim is compiled with large file
4778 support, which makes it possible to handle log files that are bigger than
4781 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4782 instance) a domain was checked against a named list that involved a lookup,
4783 causing $domain_data to be set, then another domain was checked against the
4784 same list, then the first domain was re-checked, the value of $domain_data
4785 after the final check could be wrong. In particular, if the second check
4786 failed, it could be set empty. This bug probably also applied to
4789 41. The strip_trailing_dot option was not being applied to the address given
4790 with the -f command-line option.
4792 42. The code for reading a message's header from the spool was incrementing
4793 $received_count, but never initializing it. This meant that the value was
4794 incorrect (doubled) while delivering a message in the same process in which
4795 it was received. In the most common configuration of Exim, this never
4796 happens - a fresh exec is done - but it can happen when
4797 deliver_drop_privilege is set.
4799 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4800 it now includes up to 150 characters of the unexpected data in the log
4803 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4804 and building data strings. The size of both of these buffers was 10 000
4805 bytes - far larger than anybody would *ever* want, thought I. Needless to
4806 say, somebody hit the limit. I have increased the maximum line length to
4807 20 000 and the maximum data length of concatenated lines to 100 000. I have
4808 also fixed two bugs, because there was no checking on these buffers. Tsk,
4809 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4810 buffer is too small.
4812 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4813 lsearch lookups. Now it does.
4815 46. When parsing a route_list item in a manualroute router, a fixed-length
4816 buffer was used for the list of hosts. I made this 1024 bytes long,
4817 thinking that nobody would ever have a list of hosts that long. Wrong.
4818 Somebody had a whole pile of complicated expansion conditions, and the
4819 string was silently truncated, leading to an expansion error. It turns out
4820 that it is easier to change to an unlimited length (owing to other changes
4821 that have happened since this code was originally written) than to build
4822 structure for giving a limitation error. The length of the item that
4823 expands into the list of hosts is now unlimited.
4825 47. The lsearch lookup could not handle data where the length of text line was
4826 more than 4095 characters. Such lines were truncated, leading to shortened
4827 data being returned. It should now handle lines of any length.
4829 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4830 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4833 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4834 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4835 no longer happens in comment lines. A list of replacements is now placed
4836 at the head of all of the source files, except those whose only change is
4837 to replace PERL_COMMAND in the very first #! line.
4839 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4840 messages on the queue, with a bottom-up merge sort, using code contributed
4841 by Michael Haardt. This should make operations like -bp somewhat faster on
4842 large queues. It won't affect queue runners, except when queue_run_in_order
4845 51. Installed eximstats 1.31 in the distribution.
4847 52. Added support for SRV lookups to the dnslookup router.
4849 53. If an ACL referred to $message_body or $message_body_end, the value was not
4850 reset for any messages that followed in the same SMTP session.
4852 54. The store-handling optimization for building very long strings was not
4853 differentiating between the different store pools. I don't think this
4854 actually made any difference in practice, but I've tidied it.
4856 55. While running the routers to verify a sender address, $sender_address
4857 was still set to the sender address. This is wrong, because when routing to
4858 send a bounce to the sender, it would be empty. Therefore, I have changed
4859 it so that, while verifying a sender address, $sender_address is set to <>.
4860 (There is no change to what happens when verifying a recipient address.)
4862 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4863 target A or AAAA records (if not already returned) without resetting the
4864 qualify_single or search_parents options of the DNS resolver. These are
4865 inappropriate in this case because the targets of MX and SRV records must
4866 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4867 target that, when qualified, matched something in the local domain. These
4868 two options are now turned off when doing these lookups.
4870 57. It seems that at least some releases of Reiserfs (which does not have the
4871 concept of a fixed number of inodes) returns zero and not -1 for the
4872 number of available inodes. This interacted badly with check_spool_inodes,
4873 which assumed that -1 was the "no such thing" setting. What I have done is
4874 to check that the total number of inodes is greater than zero before doing
4875 the test of how many are available.
4877 58. When a "warn" ACL statement has a log_message modifier, the message is
4878 remembered, and not repeated. This is to avoid a lot of repetition when a
4879 message has many recipients that cause the same warning to be written.
4880 Howewer, Exim was preserving the list of already written lines for an
4881 entire SMTP session, which doesn't seem right. The memory is now reset if a
4882 new message is started.
4884 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4885 debugging output unless log_rewrite was also set.
4887 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4888 of (int)(handle) when we know that handle contains (void *)(-1).
4890 61. The Exim daemon panic-logs an error return when it closes the incoming
4891 connection. However "connection reset by peer" seems to be common, and
4892 isn't really an error worthy of noting specially, so that particular error
4895 62. When Exim is trying to find all the local interfaces, it used to panic and
4896 die if the ioctl to get the interface flags failed. However, it seems that
4897 on at least one OS (Solaris 9) it is possible to have an interface that is
4898 included in the list of interfaces, but for which you get a failure error
4899 for this call. This happens when the interface is not "plumbed" into a
4900 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4901 failure of the "get flags" call assumes that the interface is down.
4903 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4904 makes life easier for people who are doing arithmetic on fields extracted
4905 from dates, where you often get leading zeros that should not be
4906 interpreted as octal.
4908 64. Added qualify_domain to the redirect router, to override the global
4911 65. If a pathologically long header line contained very many addresses (the
4912 report of this problem mentioned 10 000) and each of them was rewritten,
4913 Exim could use up a very large amount of memory. (It kept on making new
4914 copies of the header line as it rewrote, and never released the old ones.)
4915 At the expense of a bit more processing, the header rewriting function has
4916 been changed so that it no longer eats memory in this way.
4918 66. The generation of the Received: header has been moved from the time that a
4919 message starts to be received, to the time that it finishes. The timestamp
4920 in the Received: header should now be very close to that of the <= log
4921 line. There are two side-effects of this change:
4923 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4924 logged header lines no longer include the local Received: line, because
4925 it has not yet been created. The same applies to a copy of the message
4926 that is returned to a non-SMTP sender when a message is rejected.
4928 (b) When a filter file is tested using -bf, no additional Received: header
4929 is added to the test message. After some thought, I decided that this
4932 This change does not affect the value of $received_for. It is still set
4933 after address rewriting, but before local_scan() is called.
4935 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4937 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4938 gave an unhelpful panic error message, and a defer error. I have managed to
4939 change this behaviour so that it now rejects any supplied certificate,
4940 which seems right, as the list of acceptable certificates is empty.
4942 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4943 gave an unhelpful defer error. I have not managed to make this reject any
4944 supplied certificates, but the error message it gives is "no certificate
4945 supplied", which is not helpful.
4947 70. exigrep's output now also includes lines that are not associated with any
4948 message, but which match the given pattern. Implemented by a patch from
4949 Martin Sluka, which also tidied up the Perl a bit.
4951 71. Recipient callout verification, like sender verification, was using <> in
4952 the MAIL FROM command. This isn't really the right thing, since the actual
4953 sender may affect whether the remote host accepts the recipient or not. I
4954 have changed it to use the actual sender in the callout; this means that
4955 the cache record is now keyed on a recipient/sender pair, not just the
4956 recipient address. There doesn't seem to be a real danger of callout loops,
4957 since a callout by the remote host to check the sender would use <>.
4958 [SEE ABOVE: changed after hitting problems.]
4960 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4961 temporary errors. However, in the case of such a code being given after
4962 the end of a data transmission (i.e. after ".") Exim was failing to write
4963 a retry record for the message. (Yes, there was some broken host that was
4964 actually sending 8xx at this point.)
4966 73. An unknown lookup type in a host list could cause Exim to panic-die when
4967 the list was checked. (An example that provoked this was putting <; in the
4968 middle of a list instead of at the start.) If this happened during a DATA
4969 ACL check, a -D file could be left lying around. This kind of configuration
4970 error no longer causes Exim to die; instead it causes a defer errror. The
4971 incident is still logged to the main and panic logs.
4973 74. Buglet left over from Exim 3 conversion. The message "too many messages
4974 in one connection" was written to the rejectlog but not the mainlog, except
4975 when address rewriting (yes!) was being logged.
4977 75. Added write_rejectlog option.
4979 76. When a system filter was run not as root (that is, when system_filter_user
4980 was set), the values of the $n variables were not being returned to the
4981 main process; thus, they were not subsequently available in the $sn
4984 77. Added +return_path_on_delivery log selector.
4986 78. A connection timeout was being treated differently from recipients deferred
4987 when testing hosts_max_try with a message that was older than the host's
4988 retry timeout. (The host should not be counted, thus allowing all hosts to
4989 be tried at least once before bouncing.) This may have been the cause of an
4990 occasionally reported bug whereby a message would remain on the queue
4991 longer than the retry timeout, but would be bounced if a delivery was
4992 forced. I say "may" because I never totally pinned down the problem;
4993 setting up timeout/retry tests is difficult. See also the next item.
4995 79. The ultimate address timeout was not being applied to errors that involved
4996 a combination of host plus message (for example, a timeout on a MAIL
4997 command). When an address resolved to a number of possible hosts, and they
4998 were not all tried for each delivery (e.g. because of hosts_max_try), a
4999 message could remain on the queue longer than the retry timeout.
5001 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5004 81. Fixed an obscure SMTP outgoing bug which required at least the following
5005 conditions: (a) there was another message waiting for the same server;
5006 (b) the server returned 5xx to all RCPT commands in the first message so
5007 that the message was not completed; (c) the server dropped the connection
5008 or gave a negative response to the RSET that Exim sends to abort the
5009 transaction. The observed case was a dropped connection after DATA that had
5010 been sent in pipelining mode. That is, the server had advertised PIPELINING
5011 but was not implementing it correctly. The effect of the bug was incorrect
5012 behaviour, such as trying another host, and this could lead to a crash.
5018 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5019 and daemon.c were passed as pointers to ints; they should have been
5020 pointers to socklen_t variables (which are typically unsigned ints).
5022 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5025 3. Fixed a really odd bug that affected only the testing scheme; patching a
5026 certain fixed string in the binary changed the value of another string that
5027 happened to be identical to the end of the original first string.
5029 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5030 name", it returns that address as the IP address. On some operating
5031 systems (e.g. Solaris), it also passes back the IP address string as the
5032 "host name". However, on others (e.g. Linux), it passes back an empty
5033 string. Exim wasn't checking for this, and was changing the host name to an
5034 empty string, assuming it had been canonicized.
5036 5. Although rare, it is permitted to have more than one PTR record for a given
5037 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5038 all the names associated with an address, because they do in Solaris.
5039 However, it seems that they do not in Linux for data that comes from the
5040 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5041 I found this out when I moved to a new Linux workstation and tried to run
5042 the Exim test suite.
5044 To get round this problem I have changed the code so that it now does its
5045 own call to the DNS to look up PTR records when searching for a host name.
5046 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5047 addresses that are only in /etc/hosts are still found.
5049 This behaviour is, however, controlled by an option called host_lookup_
5050 order, which defaults to "bydns:byaddr". If people want to use the other
5051 order, or indeed, just use one or the other means of lookup, they can
5052 specify it in this variable.
5054 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5055 some operating systems, this comes back from gethostbyaddr() as an empty
5056 string, and this is what Exim used to test for. However, it seems that in
5057 other systems, "." is yielded. Exim now tests for this case too.
5059 7. The values of check_spool_space and check_log_space are now held internally
5060 as a number of kilobytes instead of an absolute number of bytes. If a
5061 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5062 kilobyte. This means that much larger values can be stored.
5064 8. Exim monitor: an attempt to get the action menu when not actually pointing
5065 at a message produces an empty menu entitled "No message selected". This
5066 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5067 no entries in it ("Shell widget menu has zero width and/or height"). So I
5068 have added a single, blank menu entry in this case.
5070 9. Added ${quote_local_part.
5072 10. MIME decoding is now applied to the contents of Subject: header lines when
5075 11. Now that a reference to $sender_host_address automatically causes a reverse
5076 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5077 host lookup before query-style lookups in lists that might use this
5078 variable. This has therefore been abolished, and the "net-" prefix is no
5079 longer necessary for query-style lookups.
5081 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5082 have been a typo for LFLAGS, so it has been changed.
5084 13. The install script calls Exim with "-C /dev/null" in order to find the
5085 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5086 to be output. Howeve, since Exim outputs its version number before the
5087 error, it didn't break the script. It just looked ugly. I fixed this by
5088 always allowing "-C /dev/null" if the caller is root.
5090 14. Ignore overlarge ACL variable number when reading spool file - insurance
5091 against a later release with more variables having written the file.
5093 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5094 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5095 this, as well as the form without the "IPv6" on the front.
5097 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5098 OS/Makefile-Darwin file.
5100 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5101 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5103 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5104 variable. However, some OS use ints for inodes. I've added cast to long int
5105 to get rid of the compiler warning.
5107 19. I had forgotten to lock out "/../" in configuration file names when
5108 ALT_CONFIG_PREFIX was set.
5110 20. Routers used for verification do not need to specify transports. However,
5111 if such a router generated a host list, and callout was configured, Exim
5112 crashed, because it could not find a port number from the (non-existent)
5113 transport. It now assumes port 25 in this circumstance.
5115 21. Added the -t option to exigrep.
5117 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5118 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5119 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5120 them is set, LOOKUP_LSEARCH is forced.
5122 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5124 24. Added sender and host information to the "rejected by local_scan()" log
5125 line; previously there was no indication of these.
5127 25. Added .include_if_exists.
5129 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5130 Linux. It turns out that not all file systems support this. Apparently some
5131 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5132 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5133 not supported on the file descriptor, is now ignored when Exim is trying to
5134 sync a directory. This applies only to Linux.
5136 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5138 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5139 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5142 29. Added support for maildirsize files from supplied patch (modified a bit).
5144 30. The use of :fail: followed by an empty string could lead Exim to respond to
5145 sender verification failures with (e.g.):
5147 550 Verification failed for <xxx>
5148 550 Sender verify failed
5150 where the first response line was missing the '-' that indicates it is not
5151 the final line of the response.
5153 31. The loop for finding the name of the user that called Exim had a hardwired
5154 limit of 10; it now uses the value of finduser_retries, which is used for
5155 all other user lookups.
5157 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5160 33. Exim was neglecting to zero errno before one call of strtol() when
5161 expanding a string and expecting an integer value. On some systems this
5162 resulted in spurious "integer overflow" errors. Also, it was casting the
5163 result into an int without checking.
5165 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5166 did not work. The code looks as if it has *never* worked, though it appears
5167 to have been documented since at least releast 1.62. I have made it work.
5169 35. The "timeout_DNS" error in retry rules, also documented since at least
5170 1.62, also never worked. As it isn't clear exactly what this means, and
5171 clearly it isn't a major issue, I have abolished the feature by treating it
5172 as "timeout", and writing a warning to the main and panic logs.
5174 36. The display of retry rules for -brt wasn't always showing the error code
5177 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5178 timeout_connect_A, timeout_connect_MX.
5180 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5181 to the empty sender.
5183 39. The daemon was not analysing the content of -oX till after it had closed
5184 stderr and disconnected from the controlling terminal. This meant that any
5185 syntax errors were only noted on the panic log, and the return code from
5186 the command was 0. By re-arranging the code a little, I've made the
5187 decoding happen first, so such errors now appear on stderr, and the return
5188 code is 1. However, the actual setting up of the sockets still happens in
5189 the disconnected process, so errors there are still only recorded on the
5192 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5193 connections (as happens on some IP stacks) was logged at start up time as
5194 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5195 it from "IPv6 and IPv4", which means that two separate sockets are being
5198 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5199 says whether AF_INET or AF_INET6 was passed as an argument.
5201 42. Exiwhat output was messed up when time zones were included in log
5204 43. Exiwhat now gives more information about the daemon's listening ports,
5205 and whether -tls-on-connect was used.
5207 44. The "port" option of the smtp transport is now expanded.
5209 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5210 silently ignored. Now an error message is written to the main and panic
5213 46. There's a new ACL modifier called "logwrite" which writes to a log file
5214 as soon as it is encountered.
5216 47. Added $local_user_uid and $local_user_gid at routing time.
5218 48. Exim crashed when trying to verify a sender address that was being
5221 49. Exim was recognizing only a space character after ".include". It now also
5222 recognizes a tab character.
5224 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5225 extracting the relevant information from the specification. The man page no
5226 longer contains scrambled data for the -d option, and I've added a section
5227 at the front about calling Exim under different names.
5229 51. Added "extra_headers" argument to the "mail" command in filter files.
5231 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5234 53. Installed eximstats 1.29.
5236 54. Added transport_filter_timeout as a generic transport option.
5238 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5239 Cc: header lines. This was required by RFC 822, but it not required by RFC
5242 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5243 incoming messages that did not have them. Now it does so only if the
5244 message originates locally, that is, if there is no associated remote host
5245 address. When Resent- header lines are present, this applies to the Resent-
5246 lines rather than the non-Resent- lines.
5248 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5249 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5251 58. Messages for configuration errors now include the name of the main
5252 configuration file - useful now that there may be more than one file in a
5253 list (.included file names were always shown).
5255 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5256 for those rare installations that do not start the daemon as root or run it
5257 setuid root. I've cut out the call to initgroups() if the daemon is not
5260 60. The Exim user and group can now be bound into the binary as text strings
5261 that are looked up at the start of Exim's processing.
5263 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5265 62. Added $mailstore_basename variable.
5267 63. Installed patch to sieve.c from Michael Haardt.
5269 64. When Exim failed to open the panic log after failing to open the main log,
5270 the original message it was trying to log was written to stderr and debug
5271 output, but if they were not available (the usual case in production), it
5272 was lost. Now it is written to syslog before the two lines that record the
5273 failures to open the logs.
5275 65. Users' Exim filters run in subprocesses under the user's uid. It is
5276 possible for a "deliver" command or an alias in a "personal" command to
5277 provoke an address rewrite. If logging of address rewriting is configured,
5278 this fails because the process is not running as root or exim. There may be
5279 a better way of dealing with this, but for the moment (because 4.30 needs
5280 to be released), I have disabled address rewrite logging when running a
5281 filter in a non-root, non-exim process.
5287 1. The buildconfig auxiliary program wasn't quoting the value set for
5288 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5289 not defined. This bug was masked in 4.22 by the effect that was fixed in
5292 2. Some messages that were rejected after a message id was allocated were
5293 shown as "incomplete" by exigrep. It no longer does this for messages that
5294 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5296 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5297 have allow_domain_literals set, the ID did not get logged in the <= line.
5298 Domain literals are now always recognized in Message-ID: header lines.
5300 4. The first argument for a ${extract expansion item is the key name or field
5301 number. Leading and trailing spaces in this item were not being ignored,
5302 causing some misleading effects.
5304 5. When deliver_drop_privilege was set, single queue runner processes started
5305 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5306 same command in the process it spins off) were not dropping privilege.
5308 6. When the daemon running as "exim" started a queue runner, it always
5309 re-executed Exim in the spun-off process. This is a waste of effort when
5310 deliver_drop_privilege is set. The new process now just calls the
5311 queue-runner function directly.
5317 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5320 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5321 ignored. Though the use of -oP was forcing the writing of a pid file, it
5322 was always written to the default place.
5324 3. If the message "no IP address found for host xxxx" is generated during
5325 incoming verification, it is now followed by identification of the incoming
5326 connection (so you can more easily find what provoked it).
5328 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5330 5. Added some features to "harden" Exim a bit more against certain attacks:
5332 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5333 be put in Local/Makefile. This is like the never_users runtime option,
5334 but it cannot be overridden. The default setting is "root".
5336 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5337 prefix string with which any file named in a -C command line option
5340 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5341 is retained for -C and -D only if the caller of Exim is root. Without
5342 it, the exim user may also use -C and -D and retain privilege.
5344 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5345 command line option is disabled.
5347 6. Macro names set by the -D option must start with an upper case letter, just
5348 like macro names defined in the configuration file.
5350 7. Added "dereference=" facility to LDAP.
5352 8. Two instances of the typo "uknown" in the source files are fixed.
5354 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5355 the Configure-Makefile script screwed up while processing it.
5357 10. Incorporated PCRE 4.4.
5359 11. The SMTP synchronization check was not operating right at the start of an
5360 SMTP session. For example, it could not catch a HELO sent before the client
5361 waited for the greeting. There is now a check for outstanding input at the
5362 point when the greeting is written. Because of the duplex, asynchronous
5363 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5364 way, but not yet received, when the check is performed.
5366 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5367 on TCP/IP sockets, because this apparently causes some broken clients to
5370 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5371 unchanged) from the Cygwin maintainer.
5373 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5374 was supported instead of testing for mbx. Effectively a typo.
5376 15. The spa authenticator server code was not checking that the input it
5377 received was valid base64.
5379 16. The debug output line for the "set" modifier in ACLs was not showing the
5380 name of the variable that was being set.
5382 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5384 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5385 lookup on its own. Something else was needed to trigger the lookup. For
5386 example, a match in host_lookup or the need for a host name in a host list.
5387 Now, if $sender_host_name is referenced and the host name has not yet been
5388 looked up, a lookup is performed. If the lookup fails, the variable remains
5389 empty, and $host_lookup_failed is set to "1".
5391 19. Added "eqi" as a case-independent comparison operator.
5393 20. The saslauthd authentication condition could segfault if neither service
5394 nor realm was specified.
5396 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5397 error message that was logged was misleading, and incoming SMTP
5398 connections were dropped. The message is now more accurate, and temporary
5399 errors are given to SMTP connections.
5401 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5402 (except RSET) with a 421 error, until QUIT is received. However, it was
5403 failing to send a response to QUIT.
5405 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5406 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5407 ACL is now run after the helo_try_verify_hosts code.
5409 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5410 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5411 case-independent, so other case variants are also recognized). Apparently
5412 some systems use these upper case variants.
5414 25. If more than two messages were waiting for the same host, and a transport
5415 filter was specified for the transport, Exim sent two messages over the
5416 same TCP/IP connection, and then failed with "socket operation on non-
5417 socket" when it tried to send the third.
5419 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5421 27. The extern definition of crypt16() in expand.c was not being excluded when
5422 the OS had its own crypt16() function.
5424 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5425 as a preferred synonym for return_size_limit, both as an option and as an
5428 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5430 30. Changed the default configuration ACL to relax the local part checking rule
5431 for addresses that are not in any local domains. For these addresses,
5432 slashes and pipe symbols are allowed within local parts, but the sequence
5433 /../ is explicitly forbidden.
5435 31. SPA server authentication was not clearing the challenge buffer before
5438 32. log_message in a "warn" ACL statement was writing to the reject log as
5439 well as to the main log, which contradicts the documentation and doesn't
5440 seem right (because no rejection is happening). So I have stopped it.
5442 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5443 However, I am unable to do any testing of this.
5445 34. Fixed an infelicity in the appendfile transport. When checking directories
5446 for a mailbox, to see if any needed to be created, it was accidentally
5447 using path names with one or more superfluous leading slashes; tracing
5448 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5450 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5451 contents are added to the log line that is written for every discarded
5452 recipient. (Previously a log_message setting was ignored.)
5454 36. The ${quote: operator now quotes the string if it is empty.
5456 37. The install script runs exim in order to find its version number. If for
5457 some reason other than non-existence or emptiness, which it checks, it
5458 could not run './exim', it was installing it with an empty version number,
5459 i.e. as "exim-". This error state is now caught, and the installation is
5462 38. An argument was missing from the function that creates an error message
5463 when Exim fails to connect to the socket for saslauthd authentication.
5464 This could cause Exim to crash, or give a corrupted message.
5466 39. Added isip, isip4, and isip6 to ${if conditions.
5468 40. The ACL variables $acl_xx are now saved with the message, and can be
5469 accessed later in routers, transports, and filters.
5471 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5472 strings in the file are not string-expanded.
5474 42. If a MAIL command specified a SIZE value that was too large to fit into an
5475 int variable, the check against message_size_limit failed. Such values are
5476 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5477 day this will have to be increased, but I don't think I want to be around
5478 when emails are that large.
5485 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5486 iconv() is not standard in FreeBSD.
5488 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5489 IPv6 enabled. The observed symptom was a segmentation fault on return from
5490 the function os_common_find_running_interfaces() in src/os.c.
5492 3. In the check_special_case() function in daemon.c I had used "errno" as an
5493 argument name, which causes warnings on some systems. This was basically a
5494 typo, since it was named "eno" in the comments!
5496 4. The code that waits for the clock to tick (at a resolution of some fraction
5497 of a second) so as to ensure message-id uniqueness was always waiting for
5498 at least one whole tick, when it could have waited for less. [This is
5499 almost certainly not relevant at current processor speeds, where it is
5500 unlikely to ever wait at all. But we try to future-proof.]
5502 5. The function that sleeps for a time interval that includes fractions of a
5503 second contained a race. It did not block SIGALRM between setting the
5504 timer, and suspending (a couple of lines later). If the interval was short
5505 and the sigsuspend() was delayed until after it had expired, the suspension
5506 never ended. On busy systems this could lead to processes getting stuck for
5509 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5510 process, before it forks any delivery processes. The open lookup caching
5511 mechanism meant that the open file or database connection was passed into
5512 the delivery process. The problem was that delivery processes always tidy
5513 up cached lookup data. This could cause a problem for the next delivery
5514 process started by the queue runner, because the external queue runner
5515 process does not know about the closure. So the next delivery process
5516 still has data in the lookup cache. In the case of a file lookup, there was
5517 no problem because closing a file descriptor in a subprocess doesn't affect
5518 the parent. However, if the lookup was caching a connection to a database,
5519 the connection was closed, and the second delivery process was likely to
5520 see errors such as "PGSQL: query failed: server closed the connection
5521 unexpectedly". The problem has been fixed by closing all cached lookups
5522 in a queue runner before running a delivery process.
5524 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5525 seem to have the "const" qualifier which it has on other OS. I've
5528 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5529 *of the same type* (client or server) with the same public name that an
5530 error should be diagnosed.
5532 9. When Exim looked up a host name for an IP address, but failed to find the
5533 original IP address when looking up the host name (a safety check), it
5534 output the message "<ip address> does not match any IP for NULL", which was
5535 confusing, to say the least. The bug was that the host name should have
5536 appeared instead of "NULL".
5538 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5539 user that is built into the binary, and the -C or -D options is used, root
5540 privilege is dropped before the configuration file is read. In addition,
5541 logging is switched to stderr instead of the normal log files. If the
5542 configuration then re-defines the Exim user, the unprivileged environment
5543 is probably not what is expected, so Exim logs a panic warning message (but
5546 However, if deliver_drop_privilege is set, the unprivileged state may well
5547 be exactly what is intended, so the warning has been cut out in that case,
5548 and Exim is allowed to try to write to its normal log files.
5554 1. smtp_return_error_details was not giving details for temporary sender
5555 or receiver verification errors.
5557 2. Diagnose a configuration error if two authenticators have the same public
5560 3. Exim used not to create the message log file for a message until the first
5561 delivery attempt. This could be confusing when incoming messages were held
5562 for policy or load reasons. The message log file is now created at the time
5563 the message is received, and an initial "Received" line is written to it.
5565 4. The automatically generated man page for command line options had a minor
5566 bug that caused no ill effects; however, a more serious problem was that
5567 the procedure for building the man page automatically didn't always
5568 operate. Consequently, release 4.20 contains an out-of-date version. This
5569 shouldn't happen again.
5571 5. When building Exim with embedded Perl support, the script that builds the
5572 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5573 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5575 6. The freeze_tell option was not being used for messages that were frozen on
5576 arrival, either by an ACL or by local_scan().
5578 7. Added the smtp_incomplete_transaction log selector.
5580 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5581 was accepting AUTH without a new EHLO.
5583 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5584 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5587 10. Logging of TCP/IP connections (when configured) now happens in the main
5588 daemon process instead of the child process, so that the TCP/IP connection
5589 count is more accurate (but it can never be perfect).
5591 11. The use of "drop" in a nested ACL was not being handled correctly in the
5592 outer ACL. Now, if condition failure induced by the nested "drop" causes
5593 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5594 or "require"), the connection is dropped.
5596 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5597 that yield "discard" can now be used with an "accept" or a "discard" verb,
5598 but an error is generated for any others (because I can't see a useful way
5599 to define what should happen).
5601 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5602 are now processed in the same way as lines in the Exim configuration file.
5603 In particular, continuation lines are supported.
5605 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5607 15. Added -ti meaning -t -i.
5609 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5610 domains, and warn by logging if others are found.
5612 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5613 returned from GIFCONF when Exim is trying to find the list of interfaces on
5614 a host. The code in os.c has been modified to copy each ifreq to an aligned
5615 structure in all cases.
5617 Also, in some cases, the returned ifreq's were being copied to a 'struct
5618 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5619 means the last couple of bytes of an IPv6 address could be chopped if the
5620 ifreq contained only a normal sockaddr (14 bytes storage).
5622 18. Named domain lists were not supported in the hosts_treat_as_local option.
5623 An entry such as +xxxx was not recognized, and was treated as a literal
5626 19. Ensure that header lines added by a DATA ACL are included in the reject log
5627 if the ACL subsequently rejects the message.
5629 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5630 MD5 (which is deprecated).
5632 21. When testing a filter file using -bf, Exim was writing a message when it
5633 took the sender from a "From " line in the message, but it was not doing so
5634 when it took $return_path from a Return-Path: header line. It now does.
5636 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5637 with a valid header line field name (a series of printing characters
5638 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5640 23. Changed "disc" in the source to "disk" to conform to the documentation and
5641 the book and for uniformity.
5643 24. Ignore Sendmail's -Ooption=value command line item.
5645 25. When execve() failed while trying to run a command in a pipe transport,
5646 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5647 could be confused with a return value of 69 from the command itself. This
5648 has been changed to 127, the value the shell returns if it is asked to run
5649 a non-existent command. The wording for the related log line suggests a
5650 non-existent command as the problem.
5652 26. If received_header_text expands to an empty string, do not add a Received:
5653 header line to the message. (Well, it adds a token one on the spool, but
5654 marks it "old" so that it doesn't get used or transmitted.)
5656 27. Installed eximstats 1.28 (addition of -nt option).
5658 28. There was no check for failure on the call to getsockname() in the daemon
5659 code. This can fail if there is a shortage of resources on the system, with
5660 ENOMEM, for example. A temporary error is now given on failure.
5662 29. Contrary to the C standard, it seems that in some environments, the
5663 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5664 program. Exim now does this explicitly; it affects the formatting of
5665 timestamps using strftime().
5667 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5668 complaints. I've now initialized all the variables, to avoid this.
5670 32. Header lines added by a system filter were not being "seen" during
5671 transport-time rewrites.
5673 33. The info_callback() function passed to OpenSSL is set up with type void
5674 (*)(SSL *, int, int), as described somewhere. However, when calling the
5675 function (actually a macro) that sets it up, the type void(*)() is
5676 expected. I've put in a cast to prevent warnings from picky compilers.
5678 34. If a DNS black list lookup found a CNAME record, but there were no A
5679 records associated with the domain it pointed at, Exim crashed.
5681 35. If a DNS black list lookup returned more than one A record, Exim ignored
5682 all but the first. It now scans all returned addresses if a particular IP
5683 value is being sought. In this situation, the contents of the
5684 $dnslist_value variable are a list of all the addresses, separated by a
5687 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5688 to accept a host name and all its aliases if the forward lookup for any of
5689 them yielded the IP address of the incoming connection. Now it accepts only
5690 those names whose forward lookup yields the correct IP address. Any other
5691 names are discarded. This closes a loophole whereby a rogue DNS
5692 administrator could create reverse DNS records to break through a
5693 wildcarded host restriction in an ACL.
5695 37. If a user filter or a system filter that ran in a subprocess used any of
5696 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5697 the wrong values were passed to the pipe command ($thisaddress had the
5698 value of $0, $0 had the value of $1, etc). This bug was introduced by
5699 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5701 38. Improved the line breaking for long SMTP error messages from ACLs.
5702 Previously, if there was no break point between 40 and 75 characters, Exim
5703 left the rest of the message alone. Two changes have been made: (a) I've
5704 reduced the minimum length to 35 characters; (b) if it can't find a break
5705 point between 35 and 75 characters, it looks ahead and uses the first one
5706 that it finds. This may give the occasional overlong line, but at least the
5707 remaining text gets split now.
5709 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5710 file descriptors might be low, and that setting 1000 would always raise it.
5711 It turns out that in some environments, the limit is already over 1000 and
5712 that lowering it causes trouble. So now Exim takes care not to decrease it.
5714 40. When delivering a message, the value of $return_path is set to $sender_
5715 address at the start of routing (routers may change the value). By an
5716 oversight, this default was not being set up when an address was tested by
5717 -bt or -bv, which affected the outcome if any router or filter referred to
5720 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5721 log only when "message" or "log_message" are set. However, if one of the
5722 conditions was an address verification, or a call to a nested ACL, the
5723 messages generated by the underlying test were being passed through. This
5724 no longer happens. The underlying message is available in $acl_verify_
5725 message for both "message" and "log_message" expansions, so it can be
5726 passed through if needed.
5728 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5729 new expansion $bh_ to give the encoded byte string without charset
5730 translation. Translation happens only if iconv() is available; HAVE_ICONV
5731 indicates this at build time. HEADERS_CHARSET gives the charset to
5732 translate to; headers_charset can change it in the configuration, and
5733 "headers charset" can change it in an individual filter file.
5735 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5736 that creates RFC 2047 encoded "words" labels them as that charset instead
5737 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5738 expansion operator; (ii) when Exim creates a From: line for a local
5739 message; (iii) when a header line is rewritten to include a "phrase" part.
5741 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5742 buggy, causing it to skip the first lines of messages whose message ID
5743 ended in 'D'. This would not have bitten before Exim release 4.14, because
5744 message IDs were unlikely to end in 'D' before then. The effect was to have
5745 incorrect size information for certain domains.
5747 45. #include "config.h" was missing at the start of the crypt16.c module. This
5748 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5751 46. If there was a timeout during a "random" callout check, Exim treated it as
5752 a failure of the random address, and carried on sending RSET and the real
5753 address. If the delay was just some slowness somewhere, the response to the
5754 original RCPT would be taken as a response to RSET and so on, causing
5755 mayhem of various kinds.
5757 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5758 when I implemented it. It didn't allow for the fact that some option values
5759 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5760 the right test for positive values.
5762 48. Domain names in DNS records are case-independent. Exim always looks them up
5763 in lower case. Some resolvers return domain names in exactly the case they
5764 appear in the zone file, that is, they may contain uppercase letters. Not
5765 all resolvers do this - some return always lower case. Exim was treating a
5766 change of case by a resolver as a change of domain, similar to a widening
5767 of a domain abbreviation. This triggered its re-routing code and so it was
5768 trying to route what was effectively the same domain again. This normally
5769 caused routing to fail (because the router wouldn't handle the domain
5770 twice). Now Exim checks for this case specially, and just changes the
5771 casing of the domain that it ultimately uses when it transmits the message
5774 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5777 50. If a filter generated a file delivery with a non-absolute name (possible if
5778 no home directory exists for the router), the forbid_file option was not
5781 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5782 the existing equality matching.
5784 52. Exim was using ints instead of ino_t variables in some places where it was
5785 dealing with inode numbers.
5787 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5788 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5789 called TMPDIR, and if it finds it is different, it changes its value.
5791 54. The smtp_printf() function is now made available to local_scan() so
5792 additional output lines can be written before returning. There is also an
5793 smtp_fflush() function to enable the detection of a dropped connection.
5794 The variables smtp_input and smtp_batched_input are exported to
5797 55. Changed the default runtime configuration: the message "Unknown user"
5798 has been removed from the ACL, and instead placed on the localuser router,
5799 using the cannot_route_message feature. This means that any verification
5800 failures that generate their own messages won't get overridden. Similarly,
5801 the "Unrouteable address" message that was in the ACL for unverifiable
5802 relay addresses has also been removed.
5804 56. Added hosts_avoid_esmtp to the smtp transport.
5806 57. The exicyclog script was not checking for the esoteric option
5807 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5808 will work only if exicyclog is run under the appropriate euid.
5810 58. Following a discussion on the list, the rules by which Exim recognises line
5811 endings on incoming messages have been changed. The -dropcr and drop_cr
5812 options are now no-ops, retained only for backwards compatibility. The
5813 following line terminators are recognized: LF CRLF CR. However, special
5814 processing applies to CR:
5816 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5817 nor a local message in the state where . is a terminator.
5819 (ii) If a bare CR is encountered in a header line, an extra space is added
5820 after the line terminator so as not to end the header. The reasoning
5821 behind this is that bare CRs in header lines are most likely either
5822 to be mistakes, or people trying to play silly games.
5824 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5825 was being incorrectly given as 18 bytes larger than it should have been.
5826 This is a VOB (very old bug).
5828 60. This may never have affected anything current, but just in case it has:
5829 When the local host is found other than at the start of a list of hosts,
5830 the local host, those with the same MX, and any that follow, are discarded.
5831 When the list in question was part of a longer list of hosts, the following
5832 hosts (not currently being processed) were also being discarded. This no
5833 longer happens. I'm not sure if this situation could ever has previously
5836 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5839 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5840 header. This is something that is recommended in a new Internet Draft, and
5841 is something that is documented as being done by Sendmail. There are two
5842 possible values. For messages generated by the autoreply transport, Exim
5845 Auto-Submitted: auto-replied
5847 whereas for all other generated messages (e.g. bounces) it adds
5849 Auto-Submitted: auto-generated
5851 63. The "personal" condition in filters now includes a test for the
5852 Auto-Submitted: header. If it contains the string "auto-" the message it
5853 not considered personal.
5855 64. Added rcpt_include_affixes as a generic transport option.
5857 65. Added queue_only_override (default true).
5859 66. Added the syslog_duplication option.
5861 67. If what should have been the first header line of a message consisted of
5862 a space followed by a colon, Exim was mis-interpreting it as a header line.
5863 It isn't of course - it is syntactically invalid and should therefore be
5864 treated as the start of the message body. The misbehaviour could have
5865 caused a number of strange effects, including loss of data in subsequent
5866 header lines, and spool format errors.
5868 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5869 client host had authenticated. This control can now be exercised by an ACL
5870 for more flexibility.
5872 69. By default, callouts do not happen when testing with -bh. There is now a
5873 variant, -bhc, which does actually run the callout code, including
5874 consulting and updating the callout cache.
5876 70. Added support for saslauthd authentication, courtesy of Alexander
5879 71. If statvfs() failed on the spool or log directories while checking their
5880 size for availability, Exim confusingly gave the error "space shortage".
5881 Furthermore, in debugging mode it crashed with a floating point exception.
5882 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5883 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5884 serious problem, Exim now writes to the main and panic logs when this
5885 happens, with details of the failure. It then refuses to accept the
5886 incoming message, giving the message "spool directory problem" or "log
5887 directory problem" with a 421 code for SMTP messages.
5889 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5890 0, 1, and 2 exist, because some OS complain for execs without them (see
5891 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5892 descriptors. However, the code omitted to check that the open succeeded,
5893 causing mysterious errors if for some reason the permissions on /dev/null
5894 got screwed. Now Exim writes a message to the main and panic logs, and
5895 bombs out if it can't open /dev/null.
5897 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5898 interact so that it is all more flexible. It is supposed to remain
5899 backwards compatible. Also added extra_local_interfaces.
5901 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5902 to bomb out with an assertion failure - to the client this appears as a
5903 connection drop. This problem occurs in the part of the code that was taken
5904 from the Samba project. Fortunately, the assertion is in a very simple
5905 function, so I have fixed this by reproducing the function inline in the
5906 one place where it is called, and arranging for authentication to fail
5907 instead of killing the process with assert().
5909 75. The SPA client code was not working when the server requested OEM rather
5910 than Unicode encoding.
5912 76. Added code to make require_files with a specific uid setting more usable in
5913 the case where statting the file as root fails - usually a non-root-mounted
5914 NFS file system. When this happens and the failure is EACCES, Exim now
5915 forks a subprocess and does the per-uid checking as the relevant uid.
5917 77. Added process_log_path.
5919 78. If log_file_path was not explicitly set, a setting of check_log_space or
5920 check_log_inodes was ignored.
5922 79. If a space check for the spool or log partitions fails, the incident is now
5923 logged. Of course, in the latter case the data may get lost...
5925 80. Added the %p formatting code to string_format() so that it can be used to
5926 print addresses in debug_print(). Adjusted all the address printing in the
5927 debugging in store.c to use %p rather than %d.
5929 81. There was a concern that a line of code in smtp_in.c could overflow a
5930 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5931 initially expressed, the concern was not well-founded, because trailing
5932 spaces are removed early. However, if the trailing spaces were followed by
5933 a NULL, they did not get removed, so the overflow was possible. Two fixes
5936 (a) I re-wrote the offending code in a cleaner fashion.
5937 (b) If an incoming SMTP command contains a NULL character, it is rejected
5940 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5941 runs initgroups(), so that if the Exim user is in any additional groups,
5942 they will be used during message reception.
5948 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob